Commit 5925c1088535c46601077c699d19769c1229c706
1 parent
2f50a0df
Browsing optimization using browser cache (with the use of header 'Etag')
Showing
2 changed files
with
3774 additions
and
3811 deletions
Show diff stats
calexium_lib/web/CXM_generic_login.anubis
| 1 | - | |
| 2 | - | |
| 3 | - | |
| 4 | - | |
| 5 | - Rationalisation de la gestion des logins et des mots de passe | |
| 6 | - | |
| 7 | -read CXM_common.anubis | |
| 8 | -read CXM_making_a_web_site.anubis | |
| 9 | -read CXM_generic_form.anubis | |
| 10 | - | |
| 11 | - | |
| 12 | - | |
| 13 | - | |
| 14 | - *** (1) Connection sur site sécurisée | |
| 15 | - | |
| 16 | - ou formulaire de saisie du login et du mot de passe | |
| 17 | - | |
| 18 | - | |
| 19 | - Le formulaire de saisie du login et du mot de passe pour se connecter à un site https | |
| 20 | - se compose : | |
| 21 | - .1. d'un éventuel message pour alerter que la paire (login,passwd) est erronée, | |
| 22 | - .2. du formulaire proprement dit pour lequel il faut donner : | |
| 23 | - - le titre du formulaire | |
| 24 | - - les textes figurant devant les 2 texts input | |
| 25 | - - le text du bouton submit | |
| 26 | - - le nom de l'action, | |
| 27 | - - la couleur de fond | |
| 28 | - | |
| 29 | - | |
| 30 | - | |
| 31 | -public define HTML_Off_Form | |
| 32 | - login_form | |
| 33 | - ( | |
| 34 | - String wrong_message, | |
| 35 | - RGB background_color, | |
| 36 | - String title_text, | |
| 37 | - String pseudo_text, | |
| 38 | - String passwd_text, | |
| 39 | - String submit_text, | |
| 40 | - String login_action | |
| 41 | - ). | |
| 42 | - | |
| 43 | - | |
| 44 | - | |
| 45 | - | |
| 46 | - *** (2) Vérification de la saisie | |
| 47 | - | |
| 48 | - | |
| 49 | -public define Maybe($User) | |
| 50 | - check_login_passwd | |
| 51 | - ( | |
| 52 | - String -> Maybe($User) check_pseudo, | |
| 53 | - $User -> ByteArray get_passwd, | |
| 54 | - List(Web_arg) lwa | |
| 55 | - ). | |
| 56 | - | |
| 57 | - | |
| 58 | - | |
| 59 | - --- That's all for the public part ! -------------------------------------------------- | |
| 60 | - | |
| 61 | - | |
| 62 | - | |
| 63 | - *** [1] Connection sur site sécurisée | |
| 64 | - | |
| 65 | -public define HTML_Off_Form | |
| 66 | - login_form | |
| 67 | - ( | |
| 68 | - String wrong_message, | |
| 69 | - RGB background_color, | |
| 70 | - String title_text, | |
| 71 | - String pseudo_text, | |
| 72 | - String passwd_text, | |
| 73 | - String submit_text, | |
| 74 | - String login_action | |
| 75 | - ) = | |
| 76 | - generic_form | |
| 77 | - ("login_form",background_color,700, | |
| 78 | - [ | |
| 79 | - title (title_text), | |
| 80 | - explain (wrong_message), | |
| 81 | - input ("pseudo",pseudo_text,narrow,"",mandatory), | |
| 82 | - password_input ("passwd",passwd_text,mandatory), | |
| 83 | - submit (login_action,failure,submit_text,[]) | |
| 84 | - ]). | |
| 85 | - | |
| 86 | - | |
| 87 | - | |
| 88 | - *** [2] Vérification de la saisie | |
| 89 | - | |
| 90 | -public define Maybe($User) | |
| 91 | - check_login_passwd | |
| 92 | - ( | |
| 93 | - String -> Maybe($User) check_pseudo, | |
| 94 | - $User -> ByteArray get_passwd, | |
| 95 | - List(Web_arg) lwa | |
| 96 | - ) = | |
| 97 | - if web_arg_value(lwa,"pseudo") is | |
| 98 | - { | |
| 99 | - not_found then failure, | |
| 100 | - found(ps) then | |
| 101 | - if web_arg_value(lwa,"passwd") is | |
| 102 | - { | |
| 103 | - not_found then failure, | |
| 104 | - found(pwd) then | |
| 105 | - if check_pseudo(ps) is | |
| 106 | - { | |
| 107 | - failure then failure, | |
| 108 | - success(user) then | |
| 109 | - if sha1(pwd)=get_passwd(user) | |
| 110 | - then success(user) | |
| 111 | - else failure | |
| 112 | - } | |
| 113 | - }}. | |
| 114 | - | |
| 115 | - | |
| 116 | - | |
| 1 | + | |
| 2 | + | |
| 3 | + | |
| 4 | + | |
| 5 | + Rationalisation de la gestion des logins et des mots de passe | |
| 6 | + | |
| 7 | +read CXM_common.anubis | |
| 8 | +read CXM_making_a_web_site.anubis | |
| 9 | +read CXM_generic_form.anubis | |
| 10 | + | |
| 11 | + | |
| 12 | + | |
| 13 | + | |
| 14 | + *** (1) Connection sur site sécurisée | |
| 15 | + | |
| 16 | + ou formulaire de saisie du login et du mot de passe | |
| 17 | + | |
| 18 | + | |
| 19 | + Le formulaire de saisie du login et du mot de passe pour se connecter à un site https | |
| 20 | + se compose : | |
| 21 | + .1. d'un éventuel message pour alerter que la paire (login,passwd) est erronée, | |
| 22 | + .2. du formulaire proprement dit pour lequel il faut donner : | |
| 23 | + - le titre du formulaire | |
| 24 | + - les textes figurant devant les 2 texts input | |
| 25 | + - le text du bouton submit | |
| 26 | + - le nom de l'action, | |
| 27 | + - la couleur de fond | |
| 28 | + | |
| 29 | + | |
| 30 | + | |
| 31 | +public define HTML_Off_Form | |
| 32 | + login_form | |
| 33 | + ( | |
| 34 | + String wrong_message, | |
| 35 | + RGB background_color, | |
| 36 | + String title_text, | |
| 37 | + String pseudo_text, | |
| 38 | + String passwd_text, | |
| 39 | + String submit_text, | |
| 40 | + String login_action | |
| 41 | + ). | |
| 42 | + | |
| 43 | + | |
| 44 | + | |
| 45 | + | |
| 46 | + *** (2) Vérification de la saisie | |
| 47 | + | |
| 48 | + | |
| 49 | +public define Maybe($User) | |
| 50 | + check_login_passwd | |
| 51 | + ( | |
| 52 | + String -> Maybe($User) check_pseudo, | |
| 53 | + $User -> ByteArray get_passwd, | |
| 54 | + List(Web_arg) lwa | |
| 55 | + ). | |
| 56 | + | |
| 57 | + | |
| 58 | + | |
| 59 | + --- That's all for the public part ! -------------------------------------------------- | |
| 60 | + | |
| 61 | + | |
| 62 | + | |
| 63 | + *** [1] Connection sur site sécurisée | |
| 64 | + | |
| 65 | +public define HTML_Off_Form | |
| 66 | + login_form | |
| 67 | + ( | |
| 68 | + String wrong_message, | |
| 69 | + RGB background_color, | |
| 70 | + String title_text, | |
| 71 | + String pseudo_text, | |
| 72 | + String passwd_text, | |
| 73 | + String submit_text, | |
| 74 | + String login_action | |
| 75 | + ) = | |
| 76 | + generic_form | |
| 77 | + ("login_form",background_color,700, | |
| 78 | + [ | |
| 79 | + title (title_text), | |
| 80 | + explain (wrong_message), | |
| 81 | + input ("pseudo",pseudo_text,narrow,"",mandatory), | |
| 82 | + password_input ("passwd",passwd_text,mandatory), | |
| 83 | + submit (login_action,failure,submit_text,[]) | |
| 84 | + ]). | |
| 85 | + | |
| 86 | + | |
| 87 | + | |
| 88 | + *** [2] Vérification de la saisie | |
| 89 | + | |
| 90 | +public define Maybe($User) | |
| 91 | + check_login_passwd | |
| 92 | + ( | |
| 93 | + String -> Maybe($User) check_pseudo, | |
| 94 | + $User -> ByteArray get_passwd, | |
| 95 | + List(Web_arg) lwa | |
| 96 | + ) = | |
| 97 | + if web_arg_value(lwa,"pseudo") is | |
| 98 | + { | |
| 99 | + not_found then failure, | |
| 100 | + found(ps) then | |
| 101 | + if web_arg_value(lwa,"passwd") is | |
| 102 | + { | |
| 103 | + not_found then failure, | |
| 104 | + found(pwd) then | |
| 105 | + if check_pseudo(ps) is | |
| 106 | + { | |
| 107 | + failure then failure, | |
| 108 | + success(user) then | |
| 109 | + if sha1(to_byte_array(pwd))=get_passwd(user) | |
| 110 | + then success(user) | |
| 111 | + else failure | |
| 112 | + } | |
| 113 | + }}. | |
| 114 | + | |
| 115 | + | |
| 116 | + | ... | ... |
calexium_lib/web/CXM_multihost_http_server.anubis
Changes suppressed. Click to show
| 1 | - | |
| 2 | - *Project* The Anubis Project | |
| 3 | - | |
| 4 | - *Title* A Multi Host HTTP/HTTPS Server | |
| 5 | - | |
| 6 | - *Copyright* Copyright (c) Anubis Team 2003-2007. | |
| 7 | - | |
| 8 | - | |
| 9 | - *Authors* Alain Prouté | |
| 10 | - David René | |
| 11 | - Cédric Ricard | |
| 12 | - | |
| 13 | - | |
| 14 | - *Revised* July 2007. | |
| 15 | - | |
| 16 | - | |
| 17 | - | |
| 18 | - *Overviews* | |
| 19 | - In this file a HTTP/HTTPS server is defined, which is able to handle multiple hosts | |
| 20 | - (virtual hosts). It answers HTTP/HTTPS requests, sends files (images or any other kind | |
| 21 | - of file), constructs HTML pages on the fly using informations received from the client | |
| 22 | - (when the URI ends by '.awp'), handles uploading of files and redirections. It is | |
| 23 | - multitasking by itself, and can handle any number of sites and clients simultaneously. | |
| 24 | - It should better be used in conjunction with 'making_a_web_site.anubis' to be found in | |
| 25 | - the same directory. If you use 'web/making_a_web_site.anubis', you don't need to read | |
| 26 | - this file. | |
| 27 | - | |
| 28 | - | |
| 29 | - ----------------------------------- Table of Contents --------------------------------- | |
| 30 | - | |
| 31 | - *** (1) Multihosting and redirections. | |
| 32 | - *** (2) The incompatibility between SSL and virtual hosts. | |
| 33 | - *** (3) HTTP headers and web arguments. | |
| 34 | - *** (4) Site descriptions. | |
| 35 | - *** (5) Protection against denial of service attacks. | |
| 36 | - *** (6) Starting your HTTP and HTTPS servers. | |
| 37 | - *** (7) Private download. | |
| 38 | - *** (8) About web argument names. | |
| 39 | - *** (9) A web dispatcher. | |
| 40 | - | |
| 41 | - --------------------------------------------------------------------------------------- | |
| 42 | - | |
| 43 | - | |
| 44 | - | |
| 45 | - | |
| 46 | - *** (1) Multihosting and redirections. | |
| 47 | - | |
| 48 | - This HTTP/HTTPS server can handle several host (also called 'virtual hosts'), in other | |
| 49 | - words, you may have several sites on the same server, with the same IP address and same | |
| 50 | - port numbers, but distinct 'host names'. | |
| 51 | - | |
| 52 | - A HTTP request sent by a browser contains the following informations: | |
| 53 | - | |
| 54 | - - a 'host name', | |
| 55 | - - an URI (Uniform Resource Identifier), | |
| 56 | - - HTTP headers, | |
| 57 | - - web arguments (in the form 'name=value'). | |
| 58 | - | |
| 59 | - Actually, the host name is just the value of the HTTP header whose name is 'Host'. The | |
| 60 | - host name indicates which site is requested. Hence, it is the primary information for | |
| 61 | - branching to the right site. If there is no 'Host' HTTP header in the request, the | |
| 62 | - request is denied. | |
| 63 | - | |
| 64 | - From now on, we may assume that the host is determined, and consequently that we are | |
| 65 | - concerned by only one site. Each site has his own directories on the server's | |
| 66 | - disk. | |
| 67 | - | |
| 68 | - Each site also has a list of 'redirections'. A redirection is a triplet, like this one: | |
| 69 | - | |
| 70 | - redirect("/", "www.our-business.com", "/homepage.awp") | |
| 71 | - | |
| 72 | - meaning that if the host is "www.our-business.com", and if the requested URI is "/", | |
| 73 | - then the URI to be served is "/homepage.awp". 'redirect' is a constructor of the type | |
| 74 | - 'Redirection' defined in 'web/common.anubis'. | |
| 75 | - | |
| 76 | - Now, an URI may end by ".awp" (meaning 'Anubis Web Page') or not. If it does, the | |
| 77 | - server understands that an HTML page must be constructed on the fly, and to that end it | |
| 78 | - calls the 'awp handler' of the site. Otherwise, the URI must end by a known extension, | |
| 79 | - like ".jpg", ".png", ".txt", etc... and represents a file path relative to the | |
| 80 | - 'public' directory of the site. If these conditions are satisfied, the file is sent to | |
| 81 | - the client. Known extensions are recorded in 'web/mime.anubis'. | |
| 82 | - | |
| 83 | - | |
| 84 | - | |
| 85 | - | |
| 86 | - *** (2) The incompatibility between SSL and virtual hosts. | |
| 87 | - | |
| 88 | - Handling virtual hosts makes a problem under SSL (i.e. when using HTTPS), which is due | |
| 89 | - to the fact that the guys at Netscape who designed SSL probably did not have the | |
| 90 | - question of virtual hosts in mind. Indeed, the SSL handshake is completed before the | |
| 91 | - server can know about the value of the 'Host' HTTP header, so that it cannot know which | |
| 92 | - server certificate must be sent to the client. This makes a problem, because the | |
| 93 | - browser will not accept a certificate whose common name does not correspond to the name | |
| 94 | - of the requested host. The user will have to accept the certificate manually, which is | |
| 95 | - not good for the security image of the site. This problem has at least two solutions | |
| 96 | - (as far as Anubis is concerned). | |
| 97 | - | |
| 98 | - Solution 1. Arrange so that the network interface on which the server is listening | |
| 99 | - has at least as many different IP addresses as you have virtual hosts. Such | |
| 100 | - supplementary IP addresses are called 'IP Aliases'. In this case, start one HTTPS | |
| 101 | - server for each virtual host, each one listening on a different address. For the time | |
| 102 | - being, this method is applicable under Anubis only if you start as many instances of | |
| 103 | - 'anbexec' as you have virtual hosts, because each instance of 'anbexec' can handle only | |
| 104 | - one server certificate. Of course, getting IP aliases is another problem to be solved | |
| 105 | - with your Internet provider. | |
| 106 | - | |
| 107 | - Solution 2. We propose a simple solution, using only one server certificate (hence | |
| 108 | - only one instance of 'anbexec'). Since, we have only one server certificate, we must | |
| 109 | - introduce a notion of 'main host', i.e. a host containing all other 'virtual | |
| 110 | - hosts'. The unique server certificate belong to the main host, so that only the main | |
| 111 | - host is identified by the client. The client must trust the main host and be confident | |
| 112 | - that the main host redirects him to the right virtual host. Actually, the process will | |
| 113 | - be transparent to the client, except that the client will see the name of the main host | |
| 114 | - instead of the name of the virtual host in the 'location' field of the browser. | |
| 115 | - | |
| 116 | - So, assume that the name of main host is 'www.securedhost.com', and that the names of | |
| 117 | - the virtual hosts are: | |
| 118 | - | |
| 119 | - actual name simplified name | |
| 120 | - ----------------------------------------------------- | |
| 121 | - www.virtual1.com virtual1 | |
| 122 | - www.virtual2.com virtual2 | |
| 123 | - www.virtual3.com virtual3 | |
| 124 | - | |
| 125 | - Then the (confidential) document '/doc/my_document.pdf' on 'www.virtual2.com' will have | |
| 126 | - the URL: | |
| 127 | - | |
| 128 | - https://www.securedhost.com/virtual2/doc/my_document.pdf | |
| 129 | - | |
| 130 | - In order to work transparently, this solution must combine HTTP and HTTPS. Indeed, the | |
| 131 | - vitual host must have a first page reachable under HTTP, through the URL: | |
| 132 | - | |
| 133 | - http://www.virtual2.com/ | |
| 134 | - | |
| 135 | - The HTTP server will redirect this URL to the awp handler of virtual host 'virtual2'. | |
| 136 | - The handler of this virtual host is able to generate a first page containing the | |
| 137 | - following HTML meta: | |
| 138 | - | |
| 139 | - <meta http-equiv="Refresh" content="0;URL=https://www.securedhost.com/virtual2/">, | |
| 140 | - | |
| 141 | - so that the client is immediately redirected to the main host under HTTPS (hence | |
| 142 | - accepting tranparently the server certificate). The awp handler of 'virtual2' then | |
| 143 | - redirects this URL to the home page (maybe a login page) of 'virtual2'. | |
| 144 | - | |
| 145 | - See 'web/making_a_web_site.anubis' for the sequel of this story. | |
| 146 | - | |
| 147 | - | |
| 148 | - | |
| 149 | - | |
| 150 | - | |
| 151 | - *** (3) HTTP headers and web arguments. | |
| 152 | - | |
| 153 | - Each HTTP request which arrives on the server contains a request line followed by a | |
| 154 | - series of HTTP headers. Each HTTP header is a pair '(name,value)' assigning a value to | |
| 155 | - a name. The type 'HTTP_header' is defined in 'web/common.anubis'. | |
| 156 | - | |
| 157 | - The request may also have a 'body'. The body contains either 'web arguments' or | |
| 158 | - uploaded files (or both). The request line itself may also contain web arguments (in a | |
| 159 | - so-called 'query string'). Like HTTP headers, 'web arguments' are pairs | |
| 160 | - '(name,value)', but the difference is that these pairs are generated by the page within | |
| 161 | - which the client clicks, while HTTP headers are generated by the browser itself. The | |
| 162 | - type 'Web_arg' is defined in 'web/common.anubis'. It has two alternatives, one for | |
| 163 | - ordinary web arguments (pairs) and one for uploaded files. | |
| 164 | - | |
| 165 | -read CXM_common.anubis | |
| 166 | -read tools/basis.anubis | |
| 167 | -read system/string.anubis | |
| 168 | -read CXM_mime.anubis | |
| 169 | - | |
| 170 | - | |
| 171 | - | |
| 172 | - *** (4) Site descriptions. | |
| 173 | - | |
| 174 | - The type HTTP_Info gathers informations comming along with the client's request. These | |
| 175 | - informations are rarely used for composing HTML pages. Nevertheless, they are at your | |
| 176 | - disposal. | |
| 177 | - | |
| 178 | -public type HTTP_Info: | |
| 179 | - http_info | |
| 180 | - ( | |
| 181 | - Int32 ip_address, // IP address of the client | |
| 182 | - String uri, // URI requested by the client | |
| 183 | - List(HTTP_header) http_headers, // HTTP headers sent by the client | |
| 184 | - One -> String generate_trust_ticket // may be used against denial of | |
| 185 | - // service attacks | |
| 186 | - ). | |
| 187 | - | |
| 188 | - | |
| 189 | - | |
| 190 | - Each site is described by a 'web site description', which is a datum of type | |
| 191 | - 'Web_Site_Description'. | |
| 192 | - | |
| 193 | -public type Web_Site_Description: | |
| 194 | - web_site_description( | |
| 195 | - List(String) common_names, | |
| 196 | - String site_directory, | |
| 197 | - List(Redirection) redirections, | |
| 198 | - String charset, | |
| 199 | - List(String) journal_extensions, | |
| 200 | - List(String) journal_headers, | |
| 201 | - String authorization_secret, | |
| 202 | - List(MIME) known_mime_types, | |
| 203 | - (String host_name, | |
| 204 | - HTTP_Info http_info, | |
| 205 | - List(Web_arg) lwa, | |
| 206 | - Bool is_https) -> (List(HTTP_header), | |
| 207 | - Printable_tree) awp_handler, | |
| 208 | - (List(Web_arg) lwa) -> One before_send_file). | |
| 209 | - | |
| 210 | - The component 'common_names' is the list of names of the site, like for example | |
| 211 | - "www.our-business.com". The reason why we have a list of common names instead of a | |
| 212 | - single common name, is that it may be useful to have a common name like "192.168.0.1" | |
| 213 | - for testing. | |
| 214 | - | |
| 215 | - 'charset' is a string which will determine the character encoding to be used by the | |
| 216 | - browser. Typically, this string is one of: "UTF-8", "ISO-8859-1", "Windows-1252", | |
| 217 | - etc... | |
| 218 | - | |
| 219 | - 'journal_extensions' is the list of URI extensions for which you want a log in the | |
| 220 | - journal (and on the console). When a request arrives, and if the extension is a member | |
| 221 | - of this list, a message is printed into the journal of the site including the date, the | |
| 222 | - IP address of the client, the complete HTTP request line. The HTTP headers whose name | |
| 223 | - is a member of 'journal_headers' are also printed in the journal. A reasonable minimum | |
| 224 | - for these two components is: | |
| 225 | - | |
| 226 | - [".awp"] for journal_extensions | |
| 227 | - ["user-agent"] for journal_headers | |
| 228 | - | |
| 229 | - 'authorization_secret' is a string which should just be unguessable. You may choose | |
| 230 | - something like (but don't choose this one !): | |
| 231 | - | |
| 232 | - "Hg8kJe42gCML9jNH-74" | |
| 233 | - | |
| 234 | - i.e. a sequence of characters typed at random, long enough to be unguessable. This is | |
| 235 | - used by the 'private download' mecanism, which is discussed later in this file. | |
| 236 | - | |
| 237 | - The component 'awp_handler' is a function of type: | |
| 238 | - | |
| 239 | - (String host_name, | |
| 240 | - HTTP_Info http_info, | |
| 241 | - List(Web_arg) web_args, | |
| 242 | - Bool is_https) -> Printable_tree | |
| 243 | - | |
| 244 | - ('Printable_tree' is a substitute for 'String' and is defined in | |
| 245 | - 'tools/basis.anubis'). This function is the 'awp handler' for the site. When the URI | |
| 246 | - ends by ".awp", this function is called, and the result (an HTML page) is sent to the | |
| 247 | - client over the connection. The last operand to this function is a boolean which is | |
| 248 | - 'true' when the requests arrives through the HTTPS channel, and 'false' when it arrives | |
| 249 | - through the HTTP channel. | |
| 250 | - | |
| 251 | - | |
| 252 | - | |
| 253 | - | |
| 254 | - | |
| 255 | - | |
| 256 | - | |
| 257 | - *** (5) Protection against denial of service attacks. | |
| 258 | - | |
| 259 | - We need to protect our servers against 'denial of service' attacks. The attack may be | |
| 260 | - send automatically from machines which are infested by viruses. In that case, our | |
| 261 | - server is saturated of connections (all virtual machines at work), but nothing is | |
| 262 | - comming on the connections. In order to avoid this problem, we propose the following: | |
| 263 | - | |
| 264 | - (1) Limit the number of simultaneous connections (say to 100). | |
| 265 | - (2) Close a connection if the request is not complete after say 10 seconds. | |
| 266 | - (3) Close the connection if the request is bigger than a given size (normal requests | |
| 267 | - are small except when there are uploaded files. | |
| 268 | - (4) Close the connection during the sending of the answer if the client is waiting | |
| 269 | - too much. | |
| 270 | - (5) Record all IP addresses with which we have encountered one of the problems above. | |
| 271 | - (6) Immediately close the connections if the IP address is in our list. | |
| 272 | - (7) Remove an address from the list only after 5 minutes of inactivity of this | |
| 273 | - address. | |
| 274 | - (8) Maintain a list of reliable IP addresses. | |
| 275 | - | |
| 276 | - Of course, all the above are approximative solutions which may in some circumstances | |
| 277 | - become either cumbersome or also partially block the system. So, it is needed to have a | |
| 278 | - set of dynamically modifiable parameters in order to master the behavior of this | |
| 279 | - mecanism. | |
| 280 | - | |
| 281 | - | |
| 282 | - Each dubious IP address is recorded together with its last activity time. | |
| 283 | - | |
| 284 | -public type DubiousIP: | |
| 285 | - dubious_ip (Int32 address, | |
| 286 | - Int32 last_activity). | |
| 287 | - | |
| 288 | - | |
| 289 | -public type DenialOfService: | |
| 290 | - denial_of_service(Var(Int32) max_connections, | |
| 291 | - Var(Int32) request_line_delay, // seconds | |
| 292 | - Var(Int32) headers_delay, | |
| 293 | - Var(Int32) answer_delay, | |
| 294 | - Var(List(DubiousIP)) list_of_dubious, | |
| 295 | - Var(List(Int32)) reliable_addresses). | |
| 296 | - | |
| 297 | - The informations in this set of variables are stored serialized into the file | |
| 298 | - 'my_anubis/web_sites/dos_info'. If this file does not exist a set if variables with | |
| 299 | - default values is created. The values are saved on the disk each time they are | |
| 300 | - modified. | |
| 301 | - | |
| 302 | -public define DenialOfService load_denial_of_service_info. | |
| 303 | - | |
| 304 | - | |
| 305 | - | |
| 306 | - *** (6) Starting your HTTP and HTTPS servers. | |
| 307 | - | |
| 308 | - When your web site descriptions are ready, you can start a pair of servers (a HTTP | |
| 309 | - server and a HTTPS server) for serving your web sites. Notice that there are always | |
| 310 | - two servers, regardless of the number of web sites, and that each web sites normally | |
| 311 | - uses the two servers. | |
| 312 | - | |
| 313 | - | |
| 314 | -public define StartServerResult | |
| 315 | - start_http_server | |
| 316 | - ( | |
| 317 | - Int32 ip_address, | |
| 318 | - Int32 http_port, | |
| 319 | - List(Web_Site_Description) web_sites, | |
| 320 | - DenialOfService dos | |
| 321 | - ). | |
| 322 | - | |
| 323 | -public define StartServerResult | |
| 324 | - start_https_server | |
| 325 | - ( | |
| 326 | - Int32 ip_address, | |
| 327 | - Int32 https_port, | |
| 328 | - String certificate_common_name, | |
| 329 | - List(Web_Site_Description) web_sites, | |
| 330 | - DenialOfService dos | |
| 331 | - ). | |
| 332 | - | |
| 333 | - The first argument 'ip_address' is the IP address on which the servers listen. If you | |
| 334 | - put 0, the servers listen on all adresses of the machine (which is useful if the | |
| 335 | - machine has several network interfaces). Otherwise, use the function 'ip_address' | |
| 336 | - defined in 'tools/basis.anubis' for composing a particular IP address. | |
| 337 | - | |
| 338 | - The next arguments are the port numbers for HTTP and HTTPS. The usual values are 80 and | |
| 339 | - 443, but you may have reasons to choose other values. | |
| 340 | - | |
| 341 | - The next argument is the list of your web site descriptions. All the sites described in | |
| 342 | - this list will be accessible on the server. | |
| 343 | - | |
| 344 | - The argument 'dos' is a set of dynamic variables containing the informations for | |
| 345 | - protecting the servers against denial of service attacks. | |
| 346 | - | |
| 347 | - | |
| 348 | - | |
| 349 | - | |
| 350 | - | |
| 351 | - | |
| 352 | - *** (7) Private download. | |
| 353 | - | |
| 354 | - It may happen that you want to propose private files for download. This means that such | |
| 355 | - a file could be downloaded only by the authorized person, and should not be seen by any | |
| 356 | - other one. This feature can be used only under HTTPS, not under HTTP. | |
| 357 | - | |
| 358 | - The file may be located anywhere on the server. Hence, the file has a complete absolute | |
| 359 | - path, like for example: | |
| 360 | - | |
| 361 | - /home/georges/my_documents/my_text.pdf | |
| 362 | - | |
| 363 | - which has nothing to do with the directories of the web server. Now, you may also want | |
| 364 | - to show another path or simply just a name to the client, not the actual absolute path | |
| 365 | - above, which may need to remain secret. So for example, the same file may appear to the | |
| 366 | - client as: | |
| 367 | - | |
| 368 | - informations.pdf | |
| 369 | - | |
| 370 | - The page must provide a link with an authorization. The authorization is just a web | |
| 371 | - argument, whose name is "zauth". The value of this web argument is computed by hashing | |
| 372 | - some secret string (known only from the programmer of the web site) with the absolute | |
| 373 | - path of the file. The HTTPS request will have the form: | |
| 374 | - | |
| 375 | - GET /informations.pdf?zauth=d38161f5b4e87e2d46e06ff8b3e233be563794d1 | |
| 376 | - | |
| 377 | - The server will search for a file named | |
| 378 | - | |
| 379 | - zd38161f5b4e87e2d46e06ff8b3e233be563794d1 | |
| 380 | - | |
| 381 | - (i.e. "z" concatenated with the value of the authorization) in the subdirectory | |
| 382 | - 'private_download' of the site directory. This file contains the absolute path of the | |
| 383 | - file, i.e: | |
| 384 | - | |
| 385 | - /home/georges/my_documents/my_text.pdf | |
| 386 | - | |
| 387 | - At that point, the server may hash the secret string and the absolute path together, to | |
| 388 | - check if the client is authorized to download the file. If it is the case, it sends the | |
| 389 | - file (the MIME type is declared as 'application/octet-stream' if it is not recognized). | |
| 390 | - The file is sent under the visible name. | |
| 391 | - | |
| 392 | - The server creates automatically the subdirectory 'private_download/' within the 'site | |
| 393 | - directory' (for each web site) if it does not already exist. Files in this directory | |
| 394 | - are deleted when they become too old (for example, after 3 days of life). | |
| 395 | - | |
| 396 | - Here is the function for computing the value of the authorization, and for making the | |
| 397 | - authorization file in 'private_download'. | |
| 398 | - | |
| 399 | -public define String | |
| 400 | - make_authorization | |
| 401 | - ( | |
| 402 | - String site_directory, | |
| 403 | - String authorization_secret, // known only by the programmer of the web site | |
| 404 | - String absolute_path // on server | |
| 405 | - ). | |
| 406 | - | |
| 407 | - See 'web/making_a_web_site.anubis' for the construction of the link for downloading. | |
| 408 | - | |
| 409 | - | |
| 410 | - | |
| 411 | - | |
| 412 | - | |
| 413 | - | |
| 414 | - | |
| 415 | - | |
| 416 | - *** (8) About web argument names. | |
| 417 | - | |
| 418 | - The server reserves the name "zauth" for the authorization in the private download | |
| 419 | - mecanism. Also, if the name of a web arguments begins by "p" (like 'password'), it does | |
| 420 | - not print the value of the web argument neither on the console or in the journal. A | |
| 421 | - good politics is to prefix all web arguments by letters distinct from 'p' and 'z'. This | |
| 422 | - method is used in 'web/making_a_web_site.anubis'. This will avoid clashes of names. | |
| 423 | - | |
| 424 | - | |
| 425 | - | |
| 426 | - | |
| 427 | - | |
| 428 | - | |
| 429 | - *** (9) A web dispatcher. | |
| 430 | - | |
| 431 | - For hosting several sites you may prefer another method which we now describe. We start | |
| 432 | - a HTTP server on port 80 (or on another port). This server is called the | |
| 433 | - ``dispatcher''. When a requests arrives, the dispatcher examines the ``host'' HTTP | |
| 434 | - header, so that it gets the name of the requested host. Then it sends to the client a | |
| 435 | - page like this one: | |
| 436 | - | |
| 437 | - <html> | |
| 438 | - <head> | |
| 439 | - <meta http-equiv="Refresh" content="0;URL=..."> | |
| 440 | - </head> | |
| 441 | - <body> | |
| 442 | - </body> | |
| 443 | - </html> | |
| 444 | - | |
| 445 | - where the URL represented by '...' is the URL of the requested site. This URL may have | |
| 446 | - the same IP address as the dispatcher, except that the port number is different. It may | |
| 447 | - also have a different IP address. | |
| 448 | - | |
| 449 | - The dispatcher uses the file 'my_anubis/web_sites/dispatcher.info'. This file contains | |
| 450 | - a serialized datum of type 'List(DispatcherInfo)'. | |
| 451 | - | |
| 452 | -public type DispatcherInfo: | |
| 453 | - site(String common_name, | |
| 454 | - Int32 http_port). | |
| 455 | - | |
| 456 | - The dispatcher does not write into this file. It reads it when it starts, and rereads | |
| 457 | - it each time the date of last modification of the file changes, so that the dispatcher | |
| 458 | - always has up to date data. The file may be managed (written and updated) by another | |
| 459 | - program. | |
| 460 | - | |
| 461 | - So, for each site, the dispatcher knows the common name (needed to recognize the 'host' | |
| 462 | - HTTP header), and the pair (ip_address,port) used by the actual site for HTTP. The | |
| 463 | - dispatcher does not worry about HTTPS. HTTPS must be managed by the actual site. | |
| 464 | - | |
| 465 | - The dispatcher is started by: | |
| 466 | - | |
| 467 | -public define One | |
| 468 | - start_web_dispatcher | |
| 469 | - ( | |
| 470 | - Int32 ip_address, // address for listening (typically 0) | |
| 471 | - Int32 port, // typically 80 | |
| 472 | - DenialOfService dos | |
| 473 | - ). | |
| 474 | - | |
| 475 | - A command line tool for managing the file 'my_anubis/web_sites/dispatcher.info' is also | |
| 476 | - provided: | |
| 477 | - | |
| 478 | - global define One | |
| 479 | - manage_web_dispatcher | |
| 480 | - ( | |
| 481 | - List(String) args | |
| 482 | - ). | |
| 483 | - | |
| 484 | - | |
| 485 | - | |
| 486 | - | |
| 487 | - | |
| 488 | - | |
| 489 | - | |
| 490 | - --- That's all for the public part ! -------------------------------------------------- | |
| 491 | - | |
| 492 | - | |
| 493 | - | |
| 494 | - | |
| 495 | - | |
| 496 | - | |
| 497 | - | |
| 498 | - ----------------------------------- Table of Contents --------------------------------- | |
| 499 | - | |
| 500 | - *** [1] Types which are private to this file. | |
| 501 | - | |
| 502 | - *** [2] Tools. | |
| 503 | - *** [2.1] Formating an error message. | |
| 504 | - *** [2.2] Converting IP addresses. | |
| 505 | - *** [2.3] Reading and unputting characters. | |
| 506 | - *** [2.4] Reading and discarding characters. | |
| 507 | - *** [2.5] Reading a character string. | |
| 508 | - *** [2.6] Padding integers with zeros. | |
| 509 | - *** [2.7] Converting web arguments to ASCII. | |
| 510 | - *** [2.8] Server description. | |
| 511 | - | |
| 512 | - *** [3] Managing the journal. | |
| 513 | - *** [3.1] Naming journal files. | |
| 514 | - *** [3.2] Formating HTTP headers. | |
| 515 | - *** [3.3] Formating web arguments. | |
| 516 | - *** [3.4] Formating the whole request. | |
| 517 | - *** [3.5] Putting it in the journal file (and on the console). | |
| 518 | - | |
| 519 | - *** [4] Reading the HTTP request. | |
| 520 | - *** [4.1] Skipping leading blanks. | |
| 521 | - *** [4.2] Reading a new line. | |
| 522 | - *** [4.3] Reading a 'word'. | |
| 523 | - *** [4.4] Separating the URI from the query string. | |
| 524 | - *** [4.5] Reading the web arguments. | |
| 525 | - *** [4.7] Reading the request line. | |
| 526 | - *** [4.8] Reading the HTTP headers. | |
| 527 | - *** [4.9] Getting the size of the request's body. | |
| 528 | - *** [4.10] Reading the body of the request. | |
| 529 | - | |
| 530 | - *** [5] Making the HTTP answer. | |
| 531 | - *** [5.1] Avoiding illegal URIs. | |
| 532 | - *** [5.2] Managing authorizations for downloading private files. | |
| 533 | - *** [5.3] Recognizing MIME types. | |
| 534 | - *** [5.4] Formating HTTP headers. | |
| 535 | - *** [5.5] Sending a file. | |
| 536 | - *** [5.6] Answering a www-url encoded request. | |
| 537 | - *** [5.7] Answering a multipart/form-data encoded request. | |
| 538 | - *** [5.7.1] Finding the boundary. | |
| 539 | - *** [5.7.2] Reading attributes from a multipart entity. | |
| 540 | - *** [5.7.3] Creating a temporary filename for an uploaded file. | |
| 541 | - *** [5.7.4] Saving an uploaded file under a temporary filename. | |
| 542 | - *** [5.7.5] Removing the path from a file name. | |
| 543 | - *** [5.7.6] Reading a multipart entity. | |
| 544 | - *** [5.8] Handling redirections. | |
| 545 | - *** [5.9] Answering both sorts of requests. | |
| 546 | - | |
| 547 | - *** [6] The HTTP/HTTPS servers. | |
| 548 | - *** [6.1] The HTTP request handler. | |
| 549 | - *** [6.2] Server's tasks. | |
| 550 | - *** [6.3] Starting the HTTP/HTTPS servers. | |
| 551 | - | |
| 552 | - *** [7] The web dispatcher. | |
| 553 | - *** [7.1] The dispatcher server. | |
| 554 | - *** [7.2] The dispatcher web site. | |
| 555 | - *** [7.3] Managing the info file. | |
| 556 | - | |
| 557 | - --------------------------------------------------------------------------------------- | |
| 558 | - | |
| 559 | - | |
| 560 | - | |
| 561 | - | |
| 562 | -read tools/basis.anubis | |
| 563 | -read tools/findstring.anubis | |
| 564 | -read tools/connections.anubis | |
| 565 | - | |
| 566 | - | |
| 567 | - | |
| 568 | - | |
| 569 | - | |
| 570 | - *** [1] Types which are private to this file. | |
| 571 | - | |
| 572 | - We use the following self-explanatory types. | |
| 573 | - | |
| 574 | -type Error: | |
| 575 | - cannot_read_from_connection, | |
| 576 | - not_get_or_post_request(String), | |
| 577 | - end_of_line_expected, | |
| 578 | - incorrect_content_length_value, | |
| 579 | - colon_expected, | |
| 580 | - timeout(Int32). | |
| 581 | - | |
| 582 | -type HTTP_RequestType: | |
| 583 | - get, | |
| 584 | - post. | |
| 585 | - | |
| 586 | -type HTTP_RequestLine: | |
| 587 | - request_line (HTTP_RequestType type, | |
| 588 | - String uri, | |
| 589 | - List(Web_arg) query_string). | |
| 590 | - | |
| 591 | -type EncodingType: | |
| 592 | - www_url, | |
| 593 | - multipart_form_data. | |
| 594 | - | |
| 595 | - | |
| 596 | - | |
| 597 | - | |
| 598 | - | |
| 599 | - *** [2] Tools. | |
| 600 | - | |
| 601 | - *** [2.1] Formating an error message. | |
| 602 | - | |
| 603 | - The next function formats an error message. | |
| 604 | - | |
| 605 | -define String | |
| 606 | - format | |
| 607 | - ( | |
| 608 | - Error msg | |
| 609 | - ) = | |
| 610 | - if msg is | |
| 611 | - { | |
| 612 | - cannot_read_from_connection then | |
| 613 | - "Cannot read from connection.\n", | |
| 614 | - not_get_or_post_request(s) then | |
| 615 | - "The request did not begin by 'GET' or 'POST': "+s+".\n", | |
| 616 | - end_of_line_expected then | |
| 617 | - "End of line expected.\n", | |
| 618 | - incorrect_content_length_value then | |
| 619 | - "Incorrect value for HTTP header 'Content-Length'.\n", | |
| 620 | - colon_expected then | |
| 621 | - "':' was expected.\n", | |
| 622 | - timeout(n) then | |
| 623 | - //"time out: "+n+"\n" | |
| 624 | - //"time out.\n" | |
| 625 | - "" | |
| 626 | - }. | |
| 627 | - | |
| 628 | - | |
| 629 | - | |
| 630 | - | |
| 631 | - | |
| 632 | - | |
| 633 | - *** [2.2] Converting IP addresses. | |
| 634 | - | |
| 635 | - We need two conversion functions for IP addresses: | |
| 636 | - | |
| 637 | - (Word8,Word8,Word8,Word8) --> Int32 ip_address | |
| 638 | - Int32 --> String ip_addr_to_string | |
| 639 | - | |
| 640 | - These conversions are defined in 'tools/basis.anubis'. | |
| 641 | - | |
| 642 | - | |
| 643 | - | |
| 644 | - | |
| 645 | - | |
| 646 | - | |
| 647 | - | |
| 648 | - | |
| 649 | - *** [2.3] Reading and unputting characters. | |
| 650 | - | |
| 651 | - We need a mecanism for unputting several characters (actually at least 3). This is | |
| 652 | - because when reading the client connection, we must sometimes go ahead several | |
| 653 | - characters, and virtually put them back into the connection, so that they can be | |
| 654 | - reread. Of course, we do not send them back to the client. We store them in a list | |
| 655 | - (hold by the variable 'unput_chars'), and we manage this list, so that characters may | |
| 656 | - be virtually put back in the connection (this is called 'unputting'). | |
| 657 | - | |
| 658 | -variable List(Word8) unput_chars = []. | |
| 659 | - | |
| 660 | - The most recently read one is the head of list. Fortunately, this variable is private | |
| 661 | - to this virtual machine (hence to this client). | |
| 662 | - | |
| 663 | - | |
| 664 | -define One | |
| 665 | - unput // unputting a character (add it in front of the list) | |
| 666 | - ( | |
| 667 | - Word8 character | |
| 668 | - ) = | |
| 669 | - unput_chars <- (List(Word8))[character . *unput_chars]. | |
| 670 | - | |
| 671 | - | |
| 672 | - | |
| 673 | -define One record_dubious_IP(Int32 addr,DenialOfService dos). | |
| 674 | - | |
| 675 | -variable Int32 sttm = 0. // contains the start time for this connection. | |
| 676 | - | |
| 677 | -define Result(Error,Word8) | |
| 678 | - record_dubious_connection | |
| 679 | - ( | |
| 680 | - Connection conn, | |
| 681 | - Int32 dead_line, | |
| 682 | - DenialOfService dos, | |
| 683 | - ) = | |
| 684 | - if remote_IP_address_and_port(conn) is (addr,port) then | |
| 685 | - record_dubious_IP(addr,dos); | |
| 686 | - print("Recording IP address "+ip_addr_to_string(addr)+ | |
| 687 | - " as dubious after "+(dead_line-*sttm)+" seconds. Total: "+ | |
| 688 | - length(*list_of_dubious(dos))+"\n"); | |
| 689 | - error(timeout(dead_line)). | |
| 690 | - | |
| 691 | - | |
| 692 | -define Result(Error,Word8) | |
| 693 | - read_one_byte | |
| 694 | - ( | |
| 695 | - Connection connection, | |
| 696 | - Int32 dead_line, | |
| 697 | - DenialOfService dos | |
| 698 | - ) = | |
| 699 | - //if now > dead_line then record_dubious_connection(connection,dead_line,dos) else | |
| 700 | - if read(connection,1,600) is // the connection is closed after 10 minutes of inactivity | |
| 701 | - { | |
| 702 | - error then error(cannot_read_from_connection), | |
| 703 | - timeout then error(timeout(600)), | |
| 704 | - //record_dubious_connection(connection,dead_line,dos), | |
| 705 | - ok(ba) then if nth(0,ba) is | |
| 706 | - { | |
| 707 | - failure then error(cannot_read_from_connection), | |
| 708 | - success(c) then ok(c) | |
| 709 | - } | |
| 710 | - }. | |
| 711 | - | |
| 712 | - | |
| 713 | -define Result(Error,Word8) | |
| 714 | - next_char // reading a character (check the list first, and read on the connection | |
| 715 | - // only when the list is empty). | |
| 716 | - ( | |
| 717 | - Connection connection, | |
| 718 | - Int32 dead_line, | |
| 719 | - DenialOfService dos | |
| 720 | - ) = | |
| 721 | - if *unput_chars is | |
| 722 | - { | |
| 723 | - [ ] then read_one_byte(connection,dead_line,dos), | |
| 724 | - | |
| 725 | - [h . t] then | |
| 726 | - unput_chars <- t; | |
| 727 | - ok(h) | |
| 728 | - }. | |
| 729 | - | |
| 730 | - | |
| 731 | - | |
| 732 | - | |
| 733 | - | |
| 734 | - | |
| 735 | - | |
| 736 | - *** [2.4] Reading and discarding characters. | |
| 737 | - | |
| 738 | - The next function reads the specified number of bytes (this is the same as | |
| 739 | - 'characters') from the connection and discards them. This is used for discarding CR LF | |
| 740 | - just before the body of a request. | |
| 741 | - | |
| 742 | -define Result(Error,One) | |
| 743 | - read_and_ignore | |
| 744 | - ( | |
| 745 | - Connection connection, // to client | |
| 746 | - Int32 dead_line, | |
| 747 | - Int32 number_of_characters, // number of characters to read and ignore | |
| 748 | - DenialOfService dos | |
| 749 | - ) = | |
| 750 | - if number_of_characters =< 0 then ok(unique) else | |
| 751 | - if next_char(connection,dead_line,dos) is | |
| 752 | - { | |
| 753 | - error(msg) then error(msg), | |
| 754 | - ok(c) then read_and_ignore(connection,dead_line,number_of_characters-1,dos) | |
| 755 | - }. | |
| 756 | - | |
| 757 | - | |
| 758 | - | |
| 759 | - | |
| 760 | - | |
| 761 | - | |
| 762 | - | |
| 763 | - *** [2.5] Reading a character string. | |
| 764 | - | |
| 765 | - Sometimes values of HTTP attributes or web args are presented in the form of double | |
| 766 | - quoted strings. The next function handles the reading of such things. The leading | |
| 767 | - double quote is already read in. We must read subsequent characters until the next non | |
| 768 | - backslashed double quote. | |
| 769 | - | |
| 770 | -define Result(Error,String) | |
| 771 | - read_string | |
| 772 | - ( | |
| 773 | - Connection connection, // connection with the client | |
| 774 | - Int32 dead_line, | |
| 775 | - List(Word8) so_far, // characters read so far (in reverse order) | |
| 776 | - DenialOfService dos | |
| 777 | - ) = | |
| 778 | - if next_char(connection,dead_line,dos) is | |
| 779 | - { | |
| 780 | - error(msg) then error(msg), | |
| 781 | - ok(c) then | |
| 782 | - if c = '\\' | |
| 783 | - then if next_char(connection,dead_line,dos) is | |
| 784 | - { | |
| 785 | - error(msg) then error(msg), | |
| 786 | - ok(d) then | |
| 787 | - if d = '\"' | |
| 788 | - then read_string(connection,dead_line,['\"' . so_far],dos) | |
| 789 | - else read_string(connection,dead_line,[d, c . so_far],dos) | |
| 790 | - } | |
| 791 | - else if c = '\"' | |
| 792 | - then ok(implode(reverse(so_far))) | |
| 793 | - else read_string(connection,dead_line,[c . so_far],dos) | |
| 794 | - }. | |
| 795 | - | |
| 796 | - | |
| 797 | - | |
| 798 | - | |
| 799 | - | |
| 800 | - | |
| 801 | - | |
| 802 | - *** [2.6] Padding integers with zeros. | |
| 803 | - | |
| 804 | - 'zero_pad_2' transforms an integer (which is assumed to be between 0 and 99) into a | |
| 805 | - string with exactly two digits. This is used for formating days, hours, minutes and | |
| 806 | - seconds. | |
| 807 | - | |
| 808 | -define String | |
| 809 | - zero_pad_2 | |
| 810 | - ( | |
| 811 | - Int32 n | |
| 812 | - ) = | |
| 813 | - with s = integer_to_string(n), | |
| 814 | - if length(s) < 2 | |
| 815 | - then "0"+s | |
| 816 | - else s. | |
| 817 | - | |
| 818 | - | |
| 819 | - | |
| 820 | - | |
| 821 | - | |
| 822 | - | |
| 823 | - | |
| 824 | - *** [2.7] Converting web arguments to ASCII. | |
| 825 | - | |
| 826 | - The function 'web_to_ascii' gets a character string and replaces web encoding by normal | |
| 827 | - ASCII encoding. This amounts to replacing: | |
| 828 | - | |
| 829 | - + by blank | |
| 830 | - %xx by the character whose ASCII code is xx in hexadecimal | |
| 831 | - | |
| 832 | - Note: We assume that '9' < 'A' (which is the case for ASCII code). | |
| 833 | - | |
| 834 | - | |
| 835 | - | |
| 836 | -define Word8 | |
| 837 | - web_decode | |
| 838 | - ( | |
| 839 | - Word8 x1, | |
| 840 | - Word8 x2 | |
| 841 | - ) = | |
| 842 | - with z1 = word8_to_int32(x1), | |
| 843 | - n1 = if z1 =< '9' then (z1 - '0') else (z1 - 'A' + 10), | |
| 844 | - z2 = word8_to_int32(x2), | |
| 845 | - n2 = if z2 =< '9' then (z2 - '0') else (z2 - 'A' + 10), | |
| 846 | - n = (n1 << 4) + n2, | |
| 847 | - truncate_to_word8(n). | |
| 848 | - | |
| 849 | - | |
| 850 | - | |
| 851 | -define String | |
| 852 | - web_to_ascii | |
| 853 | - ( | |
| 854 | - String web_string, | |
| 855 | - Int32 n, // current position in web_string | |
| 856 | - List(Word8) so_far | |
| 857 | - ) = | |
| 858 | - if nth(n,web_string) is | |
| 859 | - { | |
| 860 | - failure then implode(reverse(so_far)), | |
| 861 | - success(c) then | |
| 862 | - if c = '+' | |
| 863 | - then web_to_ascii(web_string,n+1,[' ' . so_far]) | |
| 864 | - else if c = '%' | |
| 865 | - then if nth(n+1,web_string) is | |
| 866 | - { | |
| 867 | - failure then implode(reverse(so_far)), | |
| 868 | - success(x1) then if nth(n+2,web_string) is | |
| 869 | - { | |
| 870 | - failure then implode(reverse(so_far)), | |
| 871 | - success(x2) then web_to_ascii(web_string,n+3,[web_decode(x1,x2) . so_far]) | |
| 872 | - } | |
| 873 | - } | |
| 874 | - else web_to_ascii(web_string,n+1,[c . so_far]) | |
| 875 | - }. | |
| 876 | - | |
| 877 | - | |
| 878 | - | |
| 879 | - | |
| 880 | - | |
| 881 | - | |
| 882 | - | |
| 883 | - | |
| 884 | - *** [3] Managing the journal. | |
| 885 | - | |
| 886 | - Concurrently working machines should not try to access the same file at the same | |
| 887 | - time. This problem may be solved by using the 'protect' mecanism. | |
| 888 | - | |
| 889 | - | |
| 890 | - | |
| 891 | - *** [3.1] Naming journal files. | |
| 892 | - | |
| 893 | - Since journal messages are rather prolific, we should have at least one file per | |
| 894 | - hour. Hence, the name of a journal file must be constructed from the current year, | |
| 895 | - month, day and hour. For example, it may be: | |
| 896 | - | |
| 897 | - 2003_03_12_19 | |
| 898 | - | |
| 899 | - (this is for the journal of 7 PM to 8 PM, 2003/mar/12). | |
| 900 | - | |
| 901 | -define String | |
| 902 | - make_current_journal_file_name | |
| 903 | - = | |
| 904 | - if convert_time(now) is date_and_time(y,m,d,h,_,_,_,_,_) then | |
| 905 | - integer_to_string(y)+"_"+ | |
| 906 | - zero_pad_2(m)+"_"+ | |
| 907 | - zero_pad_2(d)+"_"+ | |
| 908 | - zero_pad_2(h). | |
| 909 | - | |
| 910 | - | |
| 911 | - | |
| 912 | - | |
| 913 | - | |
| 914 | - | |
| 915 | - | |
| 916 | - *** [3.2] Formating HTTP headers. | |
| 917 | - | |
| 918 | - HTTP headers may be shown on the console or written in the journal. The function below | |
| 919 | - formats a list of HTTP headers. | |
| 920 | - | |
| 921 | -define String | |
| 922 | - show_format | |
| 923 | - ( | |
| 924 | - Web_Site_Description desc, | |
| 925 | - List(HTTP_header) headers, | |
| 926 | - ) = | |
| 927 | - if headers is | |
| 928 | - { | |
| 929 | - [ ] then "", | |
| 930 | - [h . t] then if h is http_header(name,value) then | |
| 931 | - if member(journal_headers(desc),name) | |
| 932 | - then " | "+name+": "+value+"\n"+show_format(desc,t) | |
| 933 | - else show_format(desc,t) | |
| 934 | - }. | |
| 935 | - | |
| 936 | - | |
| 937 | - | |
| 938 | - | |
| 939 | - | |
| 940 | - | |
| 941 | - *** [3.3] Formating web arguments. | |
| 942 | - | |
| 943 | - The same thing for web arguments. | |
| 944 | - | |
| 945 | -define String | |
| 946 | - show_format | |
| 947 | - ( | |
| 948 | - List(Web_arg) lwa | |
| 949 | - ) = | |
| 950 | - if lwa is | |
| 951 | - { | |
| 952 | - [ ] then "", | |
| 953 | - [h . t] then if h is | |
| 954 | - { | |
| 955 | - web_arg(n,v) then | |
| 956 | - " | "+n+"="+(if nth(0,n) = success('p') then "<not shown>" else v)+"\n"+show_format(t), | |
| 957 | - upload(n,fn,tfn) then | |
| 958 | - " | "+n+"="+fn+" (uploaded as '"+tfn+"')\n"+show_format(t) | |
| 959 | - } | |
| 960 | - }. | |
| 961 | - | |
| 962 | - | |
| 963 | - | |
| 964 | - | |
| 965 | - | |
| 966 | - | |
| 967 | - *** [3.4] Formating the whole request. | |
| 968 | - | |
| 969 | - It is cheap to transform month numbers into abbreviated month names. This enhances the | |
| 970 | - readability of the journal. | |
| 971 | - | |
| 972 | -define String | |
| 973 | - format_month | |
| 974 | - ( | |
| 975 | - Int32 m | |
| 976 | - ) = | |
| 977 | - if m = 1 then "jan" else | |
| 978 | - if m = 2 then "feb" else | |
| 979 | - if m = 3 then "mar" else | |
| 980 | - if m = 4 then "apr" else | |
| 981 | - if m = 5 then "may" else | |
| 982 | - if m = 6 then "jun" else | |
| 983 | - if m = 7 then "jul" else | |
| 984 | - if m = 8 then "aug" else | |
| 985 | - if m = 9 then "sep" else | |
| 986 | - if m = 10 then "oct" else | |
| 987 | - if m = 11 then "nov" else | |
| 988 | - if m = 12 then "dec" else | |
| 989 | - "???". | |
| 990 | - | |
| 991 | - | |
| 992 | - Below we format a whole HTTP request. This may give this (actually, it depends on how | |
| 993 | - you defined the values of 'journal_headers' and 'journal_extensions'): | |
| 994 | - | |
| 995 | - [3] 2003/mar/10 10:06:57 from 123.456.123.456: /homepage.awp | |
| 996 | - | host: www.the-best-one.com | |
| 997 | - | user-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0 | |
| 998 | - | |
| 999 | - The leading number between brackets is the number of the virtual machine which served | |
| 1000 | - the URI. | |
| 1001 | - | |
| 1002 | -define String | |
| 1003 | - format_request | |
| 1004 | - ( | |
| 1005 | - Web_Site_Description desc, | |
| 1006 | - Connection client_connection, | |
| 1007 | - HTTP_RequestLine request_line, | |
| 1008 | - List(HTTP_header) headers, | |
| 1009 | - List(Web_arg) web_args | |
| 1010 | - ) = | |
| 1011 | - with dt = convert_time(now), | |
| 1012 | - if remote_IP_address_and_port(client_connection) is (addr,port) then | |
| 1013 | - integer_to_string(year(dt))+"/"+format_month(month(dt))+"/"+zero_pad_2(day(dt))+" "+ | |
| 1014 | - zero_pad_2(hour(dt))+":"+zero_pad_2(minute(dt))+":"+zero_pad_2(second(dt))+ | |
| 1015 | - " from "+ip_addr_to_string(addr)+ | |
| 1016 | - ": "+uri(request_line)+"\n"+ | |
| 1017 | - show_format(desc,headers)+ | |
| 1018 | - show_format(web_args). | |
| 1019 | - | |
| 1020 | - | |
| 1021 | - | |
| 1022 | - | |
| 1023 | - | |
| 1024 | - | |
| 1025 | - | |
| 1026 | - *** [3.5] Putting it in the journal file (and on the console). | |
| 1027 | - | |
| 1028 | - We must not forget to 'protect' this operation, so that the messages of two machines | |
| 1029 | - (working for the same site) will not be mixed together. | |
| 1030 | - | |
| 1031 | -define One | |
| 1032 | - log_journal_msg | |
| 1033 | - ( | |
| 1034 | - Web_Site_Description desc, | |
| 1035 | - String msg, | |
| 1036 | - ) = | |
| 1037 | - with msg = to_byte_array("["+virtual_machine_id+"] "+msg+"\n"), | |
| 1038 | - protect | |
| 1039 | - ( | |
| 1040 | - if file(site_directory(desc)+"/journal/"+make_current_journal_file_name,append) is | |
| 1041 | - { | |
| 1042 | - failure then unique, | |
| 1043 | - success(journal_file) then | |
| 1044 | - forget(reliable_write(file(journal_file),msg)) | |
| 1045 | - }; | |
| 1046 | - forget(reliable_write(file(stdout),msg)) | |
| 1047 | - ). | |
| 1048 | - | |
| 1049 | - | |
| 1050 | - | |
| 1051 | - | |
| 1052 | - | |
| 1053 | - | |
| 1054 | - | |
| 1055 | - *** [4] Reading the HTTP request. | |
| 1056 | - | |
| 1057 | - | |
| 1058 | - *** [4.1] Skipping leading blanks. | |
| 1059 | - | |
| 1060 | - One of the peculiarities of HTTP is that the characters 13 (carriage return) and 10 | |
| 1061 | - (line feed) followed by either a space (32) or a tab (9), is considered as a blank not | |
| 1062 | - containing any new line. 'skip_http_blanks' must skip all blanks characters until the | |
| 1063 | - first non blank character, which should not be read in. Obviously, because of the above | |
| 1064 | - peculiarity, we need at least 3 characters of lookahead to do this. In other words, we | |
| 1065 | - must be able to unput at least 3 characters (hopefully we are). | |
| 1066 | - | |
| 1067 | - Strictly blanks characters are 'space' and 'tab'. | |
| 1068 | - | |
| 1069 | -define Bool | |
| 1070 | - is_strict_blank | |
| 1071 | - ( | |
| 1072 | - Word8 c | |
| 1073 | - ) = | |
| 1074 | - if c = ' ' then true else c = '\t'. | |
| 1075 | - | |
| 1076 | - | |
| 1077 | - On the contrary, blanks include 13 and 10. | |
| 1078 | - | |
| 1079 | -define Bool | |
| 1080 | - is_blank | |
| 1081 | - ( | |
| 1082 | - Word8 c | |
| 1083 | - ) = | |
| 1084 | - if c = ' ' then true else | |
| 1085 | - if c = '\t' then true else | |
| 1086 | - if c = 13 then true else | |
| 1087 | - c = 10. | |
| 1088 | - | |
| 1089 | - | |
| 1090 | - Skipping HTTP blanks. | |
| 1091 | - | |
| 1092 | -define Result(Error,One) | |
| 1093 | - skip_http_blanks | |
| 1094 | - ( | |
| 1095 | - Connection connection, | |
| 1096 | - Int32 dead_line, | |
| 1097 | - DenialOfService dos | |
| 1098 | - ) = | |
| 1099 | - if next_char(connection,dead_line,dos) is | |
| 1100 | - { | |
| 1101 | - error(msg) then error(msg), | |
| 1102 | - ok(c) then | |
| 1103 | - if is_strict_blank(c) | |
| 1104 | - then skip_http_blanks(connection,dead_line,dos) | |
| 1105 | - else if c = 13 | |
| 1106 | - then if next_char(connection,dead_line,dos) is | |
| 1107 | - { | |
| 1108 | - error(msg) then error(msg), // (unput(c); ok(unique)), | |
| 1109 | - ok(d) then | |
| 1110 | - if d = 10 | |
| 1111 | - then if next_char(connection,dead_line,dos) is | |
| 1112 | - { | |
| 1113 | - error(msg) then error(msg), // (unput(d); unput(c); ok(unique)), | |
| 1114 | - ok(e) then | |
| 1115 | - if is_strict_blank(e) | |
| 1116 | - then skip_http_blanks(connection,dead_line,dos) | |
| 1117 | - else (unput(e); unput(d); unput(c); ok(unique)) | |
| 1118 | - } | |
| 1119 | - else (unput(d); unput(c); ok(unique)) | |
| 1120 | - } | |
| 1121 | - else (unput(c); ok(unique)) | |
| 1122 | - }. | |
| 1123 | - | |
| 1124 | - | |
| 1125 | - | |
| 1126 | - | |
| 1127 | - | |
| 1128 | - | |
| 1129 | - | |
| 1130 | - | |
| 1131 | - *** [4.2] Reading a new line. | |
| 1132 | - | |
| 1133 | - Normally in HTTP a new line is the sequence 13 10 (carriage return line feed), not | |
| 1134 | - followed by a space or tabulator. If it is followed by a space or tabulator, the three | |
| 1135 | - characters are considered blanks, and no new line has been read. Before trying to read | |
| 1136 | - a new line, we first skip leading spaces and tabs. Then we try to read 13 and 10, and | |
| 1137 | - we read another character. if this character is space or tab, we consider we have read | |
| 1138 | - only blanks and we continue reading in order to find our new line. Otherwise, we unput | |
| 1139 | - this character (which may be for example the first character of the name of the next | |
| 1140 | - header), and answer that we have seen a new line. | |
| 1141 | - | |
| 1142 | - Warning: we must not use this function for reading the last pair (13,10) before the | |
| 1143 | - beginning of the body, because if the body is empty, there is no character to read | |
| 1144 | - after this pair, so that the server could wait for a character which will never | |
| 1145 | - come. This is the reason for 'read_and_ignore' above, which is used precisely for | |
| 1146 | - reading that last (13,10) pair. | |
| 1147 | - | |
| 1148 | -define Result(Error,One) | |
| 1149 | - read_new_line | |
| 1150 | - ( | |
| 1151 | - Connection connection, | |
| 1152 | - Int32 dead_line, | |
| 1153 | - DenialOfService dos | |
| 1154 | - ) = | |
| 1155 | - if skip_http_blanks(connection,dead_line,dos) is | |
| 1156 | - { | |
| 1157 | - error(msg) then error(msg), | |
| 1158 | - ok(_) then | |
| 1159 | - if next_char(connection,dead_line,dos) is | |
| 1160 | - { | |
| 1161 | - error(msg) then error(msg), | |
| 1162 | - ok(c) then | |
| 1163 | - if c = 13 | |
| 1164 | - then if next_char(connection,dead_line,dos) is | |
| 1165 | - { | |
| 1166 | - error(msg) then error(msg), | |
| 1167 | - ok(d) then | |
| 1168 | - if d = 10 | |
| 1169 | - then ok(unique) | |
| 1170 | - else (unput(d); | |
| 1171 | - unput(c); | |
| 1172 | - error(end_of_line_expected)) | |
| 1173 | - } | |
| 1174 | - else (unput(c); | |
| 1175 | - error(end_of_line_expected)) | |
| 1176 | - }}. | |
| 1177 | - | |
| 1178 | - | |
| 1179 | - | |
| 1180 | - | |
| 1181 | - | |
| 1182 | - | |
| 1183 | - | |
| 1184 | - | |
| 1185 | - *** [4.3] Reading a 'word'. | |
| 1186 | - | |
| 1187 | - A 'word' is a sequence of characters which begins either by a double quote or not by a | |
| 1188 | - double quote. (However, any leading blanks are read in and ignored. This is | |
| 1189 | - accomplished by 'skip_http_blanks'.) If it begins by a double quote, it is read like a | |
| 1190 | - string, i.e. it ends at the next (non backslashed) double quote. Otherwise, it is | |
| 1191 | - right delimited by any character which may be considered as 'blank'. If the word is | |
| 1192 | - double quoted, the closing double quote is read in. On the contrary, if the word is not | |
| 1193 | - double quoted, the right delimiting blank character is not read in (it is 'unput' back | |
| 1194 | - into the connection), and may be read in again. This is needed because carriage return | |
| 1195 | - or line feed which are 'blank', also have a meaning in HTTP. | |
| 1196 | - | |
| 1197 | -define Result(Error,String) | |
| 1198 | - read_word_aux | |
| 1199 | - ( | |
| 1200 | - Connection connection, | |
| 1201 | - Int32 dead_line, | |
| 1202 | - List(Word8) so_far, | |
| 1203 | - DenialOfService dos | |
| 1204 | - ) = | |
| 1205 | - if next_char(connection,dead_line,dos) is | |
| 1206 | - { | |
| 1207 | - error(msg) then error(msg), | |
| 1208 | - ok(c) then | |
| 1209 | - if is_blank(c) | |
| 1210 | - then (unput(c); | |
| 1211 | - ok(implode(reverse(so_far)))) | |
| 1212 | - else read_word_aux(connection,dead_line,[c . so_far],dos) | |
| 1213 | - }. | |
| 1214 | - | |
| 1215 | -define Result(Error,String) | |
| 1216 | - read_word | |
| 1217 | - ( | |
| 1218 | - Connection connection, | |
| 1219 | - Int32 dead_line, | |
| 1220 | - DenialOfService dos | |
| 1221 | - ) = | |
| 1222 | - if skip_http_blanks(connection,dead_line,dos) is | |
| 1223 | - { | |
| 1224 | - error(msg) then error(msg), | |
| 1225 | - ok(_) then | |
| 1226 | - if next_char(connection,dead_line,dos) is | |
| 1227 | - { | |
| 1228 | - error(msg) then error(msg), | |
| 1229 | - ok(c) then | |
| 1230 | - if c = '\"' | |
| 1231 | - then read_string(connection,dead_line,[],dos) | |
| 1232 | - else read_word_aux(connection,dead_line,[c],dos) | |
| 1233 | - } | |
| 1234 | - }. | |
| 1235 | - | |
| 1236 | - | |
| 1237 | - | |
| 1238 | - | |
| 1239 | - | |
| 1240 | - | |
| 1241 | - | |
| 1242 | - | |
| 1243 | - *** [4.4] Separating the URI from the query string. | |
| 1244 | - | |
| 1245 | - A 'query string' may be postfixed to the URI, just after a question mark. For example, | |
| 1246 | - the client may send the following request: | |
| 1247 | - | |
| 1248 | - GET /catalog.awp?item=3&color=blue | |
| 1249 | - | |
| 1250 | - We separate this into an URI: "/catalog.awp" and the string: "item=3&color=blue" which | |
| 1251 | - will be later transformed into the list: | |
| 1252 | - | |
| 1253 | - [web_arg("item","3"),web_arg("color","blue")] | |
| 1254 | - | |
| 1255 | - | |
| 1256 | -define (String,String) | |
| 1257 | - separate_uri_from_query_string | |
| 1258 | - ( | |
| 1259 | - String uri_and_query_string, | |
| 1260 | - Int32 n | |
| 1261 | - ) = | |
| 1262 | - if nth(n,uri_and_query_string) is | |
| 1263 | - { | |
| 1264 | - failure then (uri_and_query_string,""), | |
| 1265 | - success(c) then | |
| 1266 | - if c = '?' | |
| 1267 | - then (substr(uri_and_query_string,0,n), | |
| 1268 | - substr(uri_and_query_string,n+1,length(uri_and_query_string)-(n+1))) | |
| 1269 | - else separate_uri_from_query_string(uri_and_query_string,n+1) | |
| 1270 | - }. | |
| 1271 | - | |
| 1272 | - | |
| 1273 | - | |
| 1274 | - | |
| 1275 | - | |
| 1276 | - | |
| 1277 | - | |
| 1278 | - | |
| 1279 | - | |
| 1280 | - *** [4.5] Reading the web arguments. | |
| 1281 | - | |
| 1282 | - HTTP/HTTPS requests are sent in one of two formats: | |
| 1283 | - | |
| 1284 | - (1) www-url encoded | |
| 1285 | - (2) multipart/form-data encoded | |
| 1286 | - | |
| 1287 | - The first one is the normal (historical) way of encoding. The second one is required | |
| 1288 | - for uploading files. A server which is supposed to accept upload of files must handle | |
| 1289 | - both formats. The first thing to do is to decide the format of the request. This is | |
| 1290 | - easily done by examining the HTTP headers. If we find the header: | |
| 1291 | - | |
| 1292 | - Content-Type: multipart/form-data | |
| 1293 | - | |
| 1294 | - the request is multipart/form-data encoded. Otherwise, it is 'www-url' encoded. We | |
| 1295 | - first consider 'www-url' encoded requests. | |
| 1296 | - | |
| 1297 | - For a 'www-url' encoded request, the web argument are either in the query string or in | |
| 1298 | - the body of the request, or both. The format is the same for both: | |
| 1299 | - | |
| 1300 | - name=value&name=value&... | |
| 1301 | - | |
| 1302 | - However, we may also have | |
| 1303 | - | |
| 1304 | - name | |
| 1305 | - name= | |
| 1306 | - name=&... | |
| 1307 | - name&... | |
| 1308 | - | |
| 1309 | - i.e. some parts may be missing. Hence, we must be careful. | |
| 1310 | - | |
| 1311 | - Furthermore, web arguments must be translated from web to ASCII when www-url encoded. | |
| 1312 | - | |
| 1313 | -define Bool | |
| 1314 | - is_ampersand_or_equal | |
| 1315 | - ( | |
| 1316 | - Word8 c | |
| 1317 | - ) = | |
| 1318 | - if c = '&' then true else c = '='. | |
| 1319 | - | |
| 1320 | - | |
| 1321 | - | |
| 1322 | - The function 'read_name_or_value' reads the string 's' starting at position 'n' until | |
| 1323 | - either the end of the string or the first '&' or '='. | |
| 1324 | - | |
| 1325 | -define String | |
| 1326 | - read_name_or_value | |
| 1327 | - ( | |
| 1328 | - String s, | |
| 1329 | - Int32 start, | |
| 1330 | - Int32 i | |
| 1331 | - ) = | |
| 1332 | - if nth(i,s) is | |
| 1333 | - { | |
| 1334 | - failure then substr(s,start,i - start), | |
| 1335 | - success(c) then | |
| 1336 | - if is_ampersand_or_equal(c) | |
| 1337 | - then substr(s,start,i-start) // the separator is not included | |
| 1338 | - else read_name_or_value(s,start,i+1) | |
| 1339 | - }. | |
| 1340 | - | |
| 1341 | - | |
| 1342 | -define List(Web_arg) | |
| 1343 | - read_www_url_encoded_web_args | |
| 1344 | - ( | |
| 1345 | - String s, | |
| 1346 | - Int32 start, | |
| 1347 | - ) = | |
| 1348 | - with first = read_name_or_value(s,start,start), | |
| 1349 | - if first = "" | |
| 1350 | - then [] | |
| 1351 | - else with i = start+length(first), | |
| 1352 | - if nth(i,s) is | |
| 1353 | - { | |
| 1354 | - failure then [web_arg(first,"")], | |
| 1355 | - success(c) then | |
| 1356 | - if c = '&' | |
| 1357 | - then [web_arg(first,"") . read_www_url_encoded_web_args(s,i+1)] | |
| 1358 | - else if c = '=' | |
| 1359 | - then with second1 = read_name_or_value(s,i+1,i+1), | |
| 1360 | - // print("\""+second1+"\"\n"); | |
| 1361 | - with second = web_to_ascii(second1,0,[]), | |
| 1362 | - [web_arg(first,second) . read_www_url_encoded_web_args(s,i+length(second1)+2)] | |
| 1363 | - else alert | |
| 1364 | - }. | |
| 1365 | - | |
| 1366 | - | |
| 1367 | - | |
| 1368 | - | |
| 1369 | - | |
| 1370 | - *** [4.7] Reading the request line. | |
| 1371 | - | |
| 1372 | - 'read_request_line' reads three words and a new line from the connection. It tries to | |
| 1373 | - recognize "GET" or "POST" in the first word, separates the URI from the query string in | |
| 1374 | - the second word, transforms the query string into a list of 'Web_arg', and finally | |
| 1375 | - returns a datum of type 'HTTP_RequestLine' if no error arose. | |
| 1376 | - | |
| 1377 | - | |
| 1378 | -define Result(Error,HTTP_RequestType) | |
| 1379 | - identify_get_or_post | |
| 1380 | - ( | |
| 1381 | - String s | |
| 1382 | - ) = | |
| 1383 | - with s = to_lower(s), | |
| 1384 | - if s = "get" then ok(get) else | |
| 1385 | - if s = "post" then ok(post) else | |
| 1386 | - error(not_get_or_post_request(s)). | |
| 1387 | - | |
| 1388 | -define Result(Error,HTTP_RequestLine) | |
| 1389 | - read_request_line | |
| 1390 | - ( | |
| 1391 | - Connection connection, | |
| 1392 | - Int32 dead_line, | |
| 1393 | - DenialOfService dos | |
| 1394 | - ) = | |
| 1395 | - if read_word(connection,dead_line,dos) is | |
| 1396 | - { | |
| 1397 | - error(msg) then error(msg), | |
| 1398 | - ok(get_or_post) then if read_word(connection,dead_line,dos) is | |
| 1399 | - { | |
| 1400 | - error(msg) then error(msg), | |
| 1401 | - ok(uri_and_query_string) then if read_word(connection,dead_line,dos) is | |
| 1402 | - { | |
| 1403 | - error(msg) then error(msg), | |
| 1404 | - ok(http_version) then if read_new_line(connection,dead_line,dos) is | |
| 1405 | - { | |
| 1406 | - error(msg) then error(msg), | |
| 1407 | - ok(_) then if separate_uri_from_query_string(uri_and_query_string,0) is | |
| 1408 | - (uri,query_string) then if identify_get_or_post(get_or_post) is | |
| 1409 | - { | |
| 1410 | - error(msg) then error(msg), | |
| 1411 | - ok(request_type) then | |
| 1412 | - ok(request_line(request_type,uri,read_www_url_encoded_web_args(query_string,0))) | |
| 1413 | - } | |
| 1414 | - } | |
| 1415 | - } | |
| 1416 | - } | |
| 1417 | - }. | |
| 1418 | - | |
| 1419 | - | |
| 1420 | - | |
| 1421 | - | |
| 1422 | - | |
| 1423 | - | |
| 1424 | - | |
| 1425 | - *** [4.8] Reading the HTTP headers. | |
| 1426 | - | |
| 1427 | - Each header is made of a name (containing only letters, the underscore, digits and the | |
| 1428 | - minus sign), a colon, a value, and a new line. The first empty line ends the headers. | |
| 1429 | - | |
| 1430 | - | |
| 1431 | - The next function tests characters acceptable in a header name. | |
| 1432 | - | |
| 1433 | -define Bool | |
| 1434 | - is_header_name_char | |
| 1435 | - ( | |
| 1436 | - Word8 c | |
| 1437 | - ) = | |
| 1438 | - with n = word8_to_int32(c), | |
| 1439 | - if ('a' =< n & n =< 'z') then true else | |
| 1440 | - if ('A' =< n & n =< 'Z') then true else | |
| 1441 | - if ('0' =< n & n =< '9') then true else | |
| 1442 | - if c = '-' then true else | |
| 1443 | - c = '_'. | |
| 1444 | - | |
| 1445 | -define Result(Error,String) | |
| 1446 | - read_header_name | |
| 1447 | - ( | |
| 1448 | - Connection connection, | |
| 1449 | - Int32 dead_line, | |
| 1450 | - List(Word8) so_far, | |
| 1451 | - DenialOfService dos | |
| 1452 | - ) = | |
| 1453 | - if next_char(connection,dead_line,dos) is | |
| 1454 | - { | |
| 1455 | - error(msg) then error(msg), | |
| 1456 | - ok(c) then | |
| 1457 | - if is_header_name_char(c) | |
| 1458 | - then read_header_name(connection,dead_line,[to_lower(c) . so_far],dos) | |
| 1459 | - else unput(c); ok(implode(reverse(so_far))) | |
| 1460 | - }. | |
| 1461 | - | |
| 1462 | -define Result(Error,One) | |
| 1463 | - skip_colon | |
| 1464 | - ( | |
| 1465 | - Connection connection, | |
| 1466 | - Int32 dead_line, | |
| 1467 | - DenialOfService dos | |
| 1468 | - ) = | |
| 1469 | - if skip_http_blanks(connection,dead_line,dos) is | |
| 1470 | - { | |
| 1471 | - error(msg) then error(msg), | |
| 1472 | - ok(_) then | |
| 1473 | - if next_char(connection,dead_line,dos) is | |
| 1474 | - { | |
| 1475 | - error(msg) then error(msg), | |
| 1476 | - ok(c) then | |
| 1477 | - if c = ':' | |
| 1478 | - then ok(unique) | |
| 1479 | - else error(colon_expected) | |
| 1480 | - }}. | |
| 1481 | - | |
| 1482 | - | |
| 1483 | -define Result(Error,String) | |
| 1484 | - read_header_value | |
| 1485 | - ( | |
| 1486 | - Connection connection, | |
| 1487 | - Int32 dead_line, | |
| 1488 | - List(Word8) so_far, | |
| 1489 | - DenialOfService dos | |
| 1490 | - ) = | |
| 1491 | - if next_char(connection,dead_line,dos) is | |
| 1492 | - { | |
| 1493 | - error(msg) then error(msg), | |
| 1494 | - ok(c) then | |
| 1495 | - if c = 13 | |
| 1496 | - then if next_char(connection,dead_line,dos) is | |
| 1497 | - { | |
| 1498 | - error(msg) then error(msg), | |
| 1499 | - ok(d) then | |
| 1500 | - if d = 10 | |
| 1501 | - then if next_char(connection,dead_line,dos) is | |
| 1502 | - { | |
| 1503 | - error(msg) then error(msg), | |
| 1504 | - ok(e) then | |
| 1505 | - if is_strict_blank(e) | |
| 1506 | - then read_header_value(connection,dead_line,[e . so_far],dos) | |
| 1507 | - else (unput(e); ok(implode(reverse(so_far)))) | |
| 1508 | - } | |
| 1509 | - else read_header_value(connection,dead_line,[d, c . so_far],dos) | |
| 1510 | - } | |
| 1511 | - else read_header_value(connection,dead_line,[c . so_far],dos) | |
| 1512 | - }. | |
| 1513 | - | |
| 1514 | - | |
| 1515 | - Reading a single header. | |
| 1516 | - | |
| 1517 | -define Result(Error,Maybe(HTTP_header)) | |
| 1518 | - read_header | |
| 1519 | - ( | |
| 1520 | - Connection connection, | |
| 1521 | - Int32 dead_line, | |
| 1522 | - DenialOfService dos | |
| 1523 | - ) = | |
| 1524 | - if read_header_name(connection,dead_line,[],dos) is | |
| 1525 | - { | |
| 1526 | - error(msg) then error(msg), | |
| 1527 | - ok(name) then | |
| 1528 | - if name = "" then | |
| 1529 | - if read_and_ignore(connection,dead_line,2,dos) /* 13 and 10 */ is | |
| 1530 | - { | |
| 1531 | - error(msg) then error(msg), | |
| 1532 | - ok(_) then // this is the blank line | |
| 1533 | - ok(failure) // end of headers | |
| 1534 | - } | |
| 1535 | - else if skip_colon(connection,dead_line,dos) is | |
| 1536 | - { | |
| 1537 | - error(msg) then error(msg), | |
| 1538 | - ok(_) then if skip_http_blanks(connection,dead_line,dos) is | |
| 1539 | - { | |
| 1540 | - error(msg) then error(msg), | |
| 1541 | - ok(_) then if read_header_value(connection,dead_line,[],dos) is | |
| 1542 | - { | |
| 1543 | - error(msg) then error(msg), | |
| 1544 | - ok(value) then | |
| 1545 | - ok(success(http_header(name,value))) | |
| 1546 | - } | |
| 1547 | - } | |
| 1548 | - } | |
| 1549 | - }. | |
| 1550 | - | |
| 1551 | - | |
| 1552 | - | |
| 1553 | - Reading all the headers. | |
| 1554 | - | |
| 1555 | -define Result(Error,List(HTTP_header)) | |
| 1556 | - read_http_headers | |
| 1557 | - ( | |
| 1558 | - Connection connection, | |
| 1559 | - Int32 dead_line, | |
| 1560 | - DenialOfService dos | |
| 1561 | - ) = | |
| 1562 | - if read_header(connection,dead_line,dos) is | |
| 1563 | - { | |
| 1564 | - error(msg) then error(msg), | |
| 1565 | - ok(mbh) then if mbh is | |
| 1566 | - { | |
| 1567 | - failure then ok([ ]), | |
| 1568 | - success(header) then | |
| 1569 | - if read_http_headers(connection,dead_line,dos) is | |
| 1570 | - { | |
| 1571 | - error(msg) then error(msg), | |
| 1572 | - ok(others) then ok([header . others]) | |
| 1573 | - } | |
| 1574 | - } | |
| 1575 | - }. | |
| 1576 | - | |
| 1577 | - | |
| 1578 | - | |
| 1579 | - | |
| 1580 | - | |
| 1581 | - | |
| 1582 | - | |
| 1583 | - *** [4.9] Getting the size of the request's body. | |
| 1584 | - | |
| 1585 | - The size of the body of the request is given under the 'Content-Length' header. If this | |
| 1586 | - header is not present, the size is assumed to be zero. | |
| 1587 | - | |
| 1588 | -define Result(Error,Int32) | |
| 1589 | - get_body_size | |
| 1590 | - ( | |
| 1591 | - List(HTTP_header) headers | |
| 1592 | - ) = | |
| 1593 | - if headers is | |
| 1594 | - { | |
| 1595 | - [ ] then ok(0), | |
| 1596 | - [h . t] then if h is http_header(name,value) then | |
| 1597 | - if name = "content-length" | |
| 1598 | - then if string_to_integer(value) is | |
| 1599 | - { | |
| 1600 | - failure then error(incorrect_content_length_value), | |
| 1601 | - success(n) then ok(n) | |
| 1602 | - } | |
| 1603 | - else get_body_size(t) | |
| 1604 | - }. | |
| 1605 | - | |
| 1606 | - | |
| 1607 | - | |
| 1608 | - | |
| 1609 | - | |
| 1610 | - | |
| 1611 | - | |
| 1612 | - | |
| 1613 | - | |
| 1614 | - | |
| 1615 | - *** [4.10] Reading the body of the request. | |
| 1616 | - | |
| 1617 | - The body of the request may be very big (it contains uploaded files, if any). We read | |
| 1618 | - it using the primitive 'read', which returns the number of bytes read, which may be | |
| 1619 | - less than the number of bytes we wanted to read. This is not an error, but simply due | |
| 1620 | - to the fact the buffer associated with the connection in the Linux (or MS-Windows) | |
| 1621 | - kernel has a limited size. Hence, we must read bytes again until we have read the | |
| 1622 | - required number of bytes. However, if the number of bytes read is zero, the connection | |
| 1623 | - may be broken. In that case, we must not try to read indefinitely. On the contrary, we | |
| 1624 | - make at most 10 retries, with a small sleeping time between any two of them. | |
| 1625 | - | |
| 1626 | -define Result(Error,ByteArray) | |
| 1627 | - read_http_body | |
| 1628 | - ( | |
| 1629 | - Connection connection, | |
| 1630 | - Int32 body_size, | |
| 1631 | - ByteArray so_far, // when calling this function, 'so_far' is the empty byte array | |
| 1632 | - Int32 retries // this function is called with retries = 10 | |
| 1633 | - ) = | |
| 1634 | - if body_size = 0 then ok(constant_byte_array(0,0)) else | |
| 1635 | - if retries =< 0 then error(cannot_read_from_connection) else | |
| 1636 | - if read(connection,body_size,60) is | |
| 1637 | - { | |
| 1638 | - error then error(cannot_read_from_connection), | |
| 1639 | - timeout then error(timeout(60)), | |
| 1640 | - ok(new_bytes) then with | |
| 1641 | - ba = so_far + new_bytes, // contains all the bytes read so far | |
| 1642 | - nr = length(ba), // total read since the beginning | |
| 1643 | - nn = length(new_bytes), // number of bytes just read | |
| 1644 | - if nr < body_size // must read more bytes | |
| 1645 | - then if nn > 0 // if connection seems to work | |
| 1646 | - then read_http_body(connection,body_size,ba,1000) // continue reading | |
| 1647 | - else sleep(100); // otherwise, sleep 1/10 of second | |
| 1648 | - read_http_body(connection,body_size,ba, // and retry reading | |
| 1649 | - retries-1) // but no more than 10 times | |
| 1650 | - else ok(ba) // required number of bytes has been read | |
| 1651 | - }. | |
| 1652 | - | |
| 1653 | - | |
| 1654 | - Note: During sleeping, 'anbexec' runs other machines. Actually, calling 'sleep', even | |
| 1655 | - for one millisecond, is some way of giving up explicitly, so that other virtual | |
| 1656 | - machines may work. | |
| 1657 | - | |
| 1658 | - | |
| 1659 | - | |
| 1660 | - | |
| 1661 | - | |
| 1662 | - | |
| 1663 | - | |
| 1664 | - | |
| 1665 | - | |
| 1666 | - | |
| 1667 | - | |
| 1668 | - | |
| 1669 | - *** [5] Making the HTTP answer. | |
| 1670 | - | |
| 1671 | - At that point we have read the request line, the headers and the body of the | |
| 1672 | - request, and we must decide what to do. | |
| 1673 | - | |
| 1674 | - Actually, we can do one of the following: | |
| 1675 | - | |
| 1676 | - - send a file, | |
| 1677 | - - execute 'tickets_and_web_page' in case of an ".awp" URI. | |
| 1678 | - | |
| 1679 | - The uploaded file (which are in the body of the request) are saved into temporary files | |
| 1680 | - below. | |
| 1681 | - | |
| 1682 | - | |
| 1683 | - | |
| 1684 | - | |
| 1685 | - | |
| 1686 | - *** [5.1] Avoiding illegal URIs. | |
| 1687 | - | |
| 1688 | - For security reasons, we must avoid illegal URIs, for example those which may climb up | |
| 1689 | - in the file hierarchy. First we accept only few characters in URIs. | |
| 1690 | - | |
| 1691 | -define Bool | |
| 1692 | - is_legal_uri_char | |
| 1693 | - ( | |
| 1694 | - Word8 c | |
| 1695 | - ) = | |
| 1696 | - with n = word8_to_int32(c), | |
| 1697 | - if ('a' =< n & n =< 'z') then true else // accept 'a' to 'z' | |
| 1698 | - if ('A' =< n & n =< 'Z') then true else // accept 'A' to 'Z' | |
| 1699 | - if ('0' =< n & n =< '9') then true else // accept '0' to '9' | |
| 1700 | - if c = '.' then true else // accept '.' '-' '/' and '_' | |
| 1701 | - if c = '-' then true else | |
| 1702 | - if c = '/' then true else | |
| 1703 | - c = '_'. | |
| 1704 | - | |
| 1705 | - We do not accept ~ which is some way of climbing. Of course, we cannot disallow single | |
| 1706 | - dots, which are most often present in legal URIs, but we must avoid double dots .. | |
| 1707 | - which mean 'climb up'. | |
| 1708 | - | |
| 1709 | -define Bool | |
| 1710 | - is_illegal_uri | |
| 1711 | - ( | |
| 1712 | - String uri, | |
| 1713 | - Int32 n | |
| 1714 | - ) = | |
| 1715 | - if nth(n,uri) is | |
| 1716 | - { | |
| 1717 | - failure then false, | |
| 1718 | - success(c) then | |
| 1719 | - if c = '.' // first dot | |
| 1720 | - then if nth(n+1,uri) is | |
| 1721 | - { | |
| 1722 | - failure then false, | |
| 1723 | - success(d) then | |
| 1724 | - if d = '.' // second dot | |
| 1725 | - then true | |
| 1726 | - else is_illegal_uri(uri,n+1) | |
| 1727 | - } | |
| 1728 | - else is_illegal_uri(uri,n+1) | |
| 1729 | - }. | |
| 1730 | - | |
| 1731 | - | |
| 1732 | - | |
| 1733 | - | |
| 1734 | - | |
| 1735 | - | |
| 1736 | - *** [5.2] Managing authorizations for downloading private files. | |
| 1737 | - | |
| 1738 | - Computing the authorization and making the authorization file (containing the absolute | |
| 1739 | - path of the file on the server). | |
| 1740 | - | |
| 1741 | - | |
| 1742 | -define String | |
| 1743 | - compute_authorization | |
| 1744 | - ( | |
| 1745 | - String authorization_secret, | |
| 1746 | - String absolute_path | |
| 1747 | - ) = | |
| 1748 | - to_ascii(sha1((authorization_secret, | |
| 1749 | - absolute_path))). | |
| 1750 | - | |
| 1751 | - | |
| 1752 | -public define String | |
| 1753 | - make_authorization | |
| 1754 | - ( | |
| 1755 | - String site_directory, | |
| 1756 | - String authorization_secret, | |
| 1757 | - String absolute_path | |
| 1758 | - ) = | |
| 1759 | - with private_download_dir = site_directory+"/private_download", | |
| 1760 | - auth = compute_authorization(authorization_secret, | |
| 1761 | - absolute_path), | |
| 1762 | - forget(save(absolute_path, | |
| 1763 | - private_download_dir+"/z"+auth)); | |
| 1764 | - auth. | |
| 1765 | - | |
| 1766 | - | |
| 1767 | - The function 'send_file' defined below handles the recognition of authorizations. | |
| 1768 | - | |
| 1769 | - | |
| 1770 | - | |
| 1771 | - | |
| 1772 | - | |
| 1773 | - *** [5.3] Recognizing MIME types. | |
| 1774 | - | |
| 1775 | - The extension of the (redirected) URI must be either ".awp" or recognized as associated | |
| 1776 | - to a MIME type. Otherwise, the server will not send the file. This is for security, but | |
| 1777 | - also because, we must generate a 'Content-Type' header in the answer, with the right | |
| 1778 | - MIME type. | |
| 1779 | - | |
| 1780 | -define String | |
| 1781 | - get_uri_extension_aux | |
| 1782 | - ( | |
| 1783 | - String uri, | |
| 1784 | - Int32 n // used for searching backwards | |
| 1785 | - ) = | |
| 1786 | - if nth(n,uri) is | |
| 1787 | - { | |
| 1788 | - failure then "", | |
| 1789 | - success(c) then | |
| 1790 | - if c = '.' then substr(uri,n,length(uri)-n) | |
| 1791 | - else if c = '/' then "" | |
| 1792 | - else get_uri_extension_aux(uri,n-1) | |
| 1793 | - }. | |
| 1794 | - | |
| 1795 | -public define String | |
| 1796 | - get_uri_extension | |
| 1797 | - ( | |
| 1798 | - String uri | |
| 1799 | - ) = | |
| 1800 | - get_uri_extension_aux(uri, | |
| 1801 | - length(uri)-1). // search starts at the right end | |
| 1802 | - | |
| 1803 | - | |
| 1804 | - | |
| 1805 | -define Maybe(String) | |
| 1806 | - recognize_mime_type_from_ext | |
| 1807 | - ( | |
| 1808 | - String ext, | |
| 1809 | - List(MIME) l | |
| 1810 | - ) = | |
| 1811 | - if l is | |
| 1812 | - { | |
| 1813 | - [ ] then success("application/octet-stream"), // failure, | |
| 1814 | - [h . t] then if h is mime(mime_type,extension) then | |
| 1815 | - if ext = extension | |
| 1816 | - then success(mime_type) | |
| 1817 | - else recognize_mime_type_from_ext(ext,t) | |
| 1818 | - }. | |
| 1819 | - | |
| 1820 | -define Maybe(String) | |
| 1821 | - recognize_mime_type_from_uri | |
| 1822 | - ( | |
| 1823 | - Web_Site_Description desc, | |
| 1824 | - String uri | |
| 1825 | - ) = | |
| 1826 | - recognize_mime_type_from_ext(get_uri_extension(uri),known_mime_types(desc)). | |
| 1827 | - | |
| 1828 | - | |
| 1829 | - | |
| 1830 | - | |
| 1831 | - | |
| 1832 | - | |
| 1833 | - | |
| 1834 | - | |
| 1835 | - *** [5.4] Formating HTTP headers. | |
| 1836 | - | |
| 1837 | - This is the formating for sending to the client (hence, it has nothing to do with the | |
| 1838 | - component 'journal_headers' in the web site description). | |
| 1839 | - | |
| 1840 | -define Printable_tree | |
| 1841 | - format_headers | |
| 1842 | - ( | |
| 1843 | - List(HTTP_header) headers | |
| 1844 | - ) = | |
| 1845 | - if headers is | |
| 1846 | - { | |
| 1847 | - [ ] then [ ], | |
| 1848 | - [h . t] then if h is http_header(name,value) then | |
| 1849 | - [name,": ",value,crlf . format_headers(t)] | |
| 1850 | - }. | |
| 1851 | - | |
| 1852 | - | |
| 1853 | - | |
| 1854 | - | |
| 1855 | - | |
| 1856 | - | |
| 1857 | - *** [5.5] Sending a file. | |
| 1858 | - | |
| 1859 | - We send 2 headers 'Content-Type' and 'Content-Length'. | |
| 1860 | - | |
| 1861 | -define List(HTTP_header) | |
| 1862 | - headers_for_send_file | |
| 1863 | - ( | |
| 1864 | - String mime_type, | |
| 1865 | - Int32 size, | |
| 1866 | - ) = | |
| 1867 | - [ | |
| 1868 | - http_header("Content-Type",mime_type), | |
| 1869 | - http_header("Content-Length",integer_to_string(size)), | |
| 1870 | - ]. | |
| 1871 | - | |
| 1872 | - | |
| 1873 | - | |
| 1874 | - Sending the body of the answer (i.e. the file itself). | |
| 1875 | - | |
| 1876 | -define One | |
| 1877 | - send_file_body | |
| 1878 | - ( | |
| 1879 | - Web_Site_Description desc, | |
| 1880 | - Connection connection, // connection with the client | |
| 1881 | - Connection file, // file to be sent already opened | |
| 1882 | - Int32 size, // size of file | |
| 1883 | - Int32 sent, // bytes already sent | |
| 1884 | - String filename // name of file | |
| 1885 | - ) = | |
| 1886 | - if sent >= size then unique else | |
| 1887 | - if read(file,min(10000,size-sent),60) is | |
| 1888 | - { | |
| 1889 | - error then log_journal_msg(desc,"Cannot read from file '"+filename+"'.\n"), | |
| 1890 | - timeout then log_journal_msg(desc,"Cannot read from file timeoput'"+filename+"'.\n"), | |
| 1891 | - ok(ba) then | |
| 1892 | - with nr = length(ba), // get the number of bytes read | |
| 1893 | - if reliable_write(connection,ba) is | |
| 1894 | - { | |
| 1895 | - failure then log_journal_msg(desc,"Cannot write into connection.\n"), | |
| 1896 | - success(nw) then | |
| 1897 | - send_file_body(desc,connection,file,size,sent+nw,filename) | |
| 1898 | - } | |
| 1899 | - }. | |
| 1900 | - | |
| 1901 | - | |
| 1902 | - | |
| 1903 | - | |
| 1904 | - Sending the answer line, the headers and the body. | |
| 1905 | - | |
| 1906 | -define One | |
| 1907 | - send_file | |
| 1908 | - ( | |
| 1909 | - Web_Site_Description desc, | |
| 1910 | - Connection connection, | |
| 1911 | - List(HTTP_header) headers, | |
| 1912 | - Int32 size, | |
| 1913 | - Connection file, | |
| 1914 | - String filename, | |
| 1915 | - One -> One action_before_send_file | |
| 1916 | - ) = | |
| 1917 | - action_before_send_file(unique); | |
| 1918 | - forget(reliable_write(connection,to_byte_array("HTTP/1.1 200 OK"+crlf))); | |
| 1919 | - forget(reliable_write(connection,[format_headers(headers) , crlf])); | |
| 1920 | - send_file_body(desc,connection,file,size,0,filename). | |
| 1921 | - | |
| 1922 | - | |
| 1923 | - | |
| 1924 | - Checking if a connection is under SSL. | |
| 1925 | - | |
| 1926 | -define Bool | |
| 1927 | - is_SSL | |
| 1928 | - ( | |
| 1929 | - Connection c | |
| 1930 | - ) = | |
| 1931 | - if c is | |
| 1932 | - { | |
| 1933 | - file_r(_) then false, | |
| 1934 | - file_w(_) then false, | |
| 1935 | - file_rw(_) then false, | |
| 1936 | - tcp(_) then false, | |
| 1937 | - ssl(_) then true | |
| 1938 | - }. | |
| 1939 | - | |
| 1940 | - | |
| 1941 | - | |
| 1942 | - Before opening and sending a file, we check the MIME type. It must be recognized, | |
| 1943 | - except if there is a valid authorization for private download. | |
| 1944 | - | |
| 1945 | -define One | |
| 1946 | - send_file | |
| 1947 | - ( | |
| 1948 | - Web_Site_Description desc, | |
| 1949 | - Connection connection, | |
| 1950 | - String uri, | |
| 1951 | - Maybe(String) mbauthorization, | |
| 1952 | - One -> One action_before_send_file | |
| 1953 | - ) = | |
| 1954 | - if mbauthorization is | |
| 1955 | - { | |
| 1956 | - //--- file without authorization: take it from public --- | |
| 1957 | - failure then if recognize_mime_type_from_uri(desc,uri) is | |
| 1958 | - { | |
| 1959 | - failure then log_journal_msg(desc,"No MIME type found for '"+uri+"'.\n"), | |
| 1960 | - success(mime_type) then | |
| 1961 | - with path = site_directory(desc)+"/public"+uri, | |
| 1962 | - if (Maybe(RStream))connect to file path is | |
| 1963 | - { | |
| 1964 | - failure then log_journal_msg(desc,"Cannot find file '"+path+"'.\n"), | |
| 1965 | - success(f) then with size = file_size(f), | |
| 1966 | - send_file(desc, | |
| 1967 | - connection, | |
| 1968 | - headers_for_send_file(mime_type,size), | |
| 1969 | - size, | |
| 1970 | - file(f), | |
| 1971 | - uri, | |
| 1972 | - action_before_send_file) | |
| 1973 | - } | |
| 1974 | - }, | |
| 1975 | - | |
| 1976 | - //--- file with authorization: apply 'private download' mecanism --- | |
| 1977 | - success(authorization) then | |
| 1978 | - with private_download_dir = site_directory(desc)+"/private_download", | |
| 1979 | - if (RetrieveResult(String))retrieve(private_download_dir+"/z"+authorization) | |
| 1980 | - is ok(absolute_path) | |
| 1981 | - then ( | |
| 1982 | - with new_hash = compute_authorization(authorization_secret(desc), | |
| 1983 | - absolute_path), | |
| 1984 | - if (Maybe(RStream))connect to file absolute_path is | |
| 1985 | - { | |
| 1986 | - failure then log_journal_msg(desc,"Cannot find file '"+absolute_path+"'.\n"), | |
| 1987 | - success(f) then with size = file_size(f), | |
| 1988 | - send_file(desc, | |
| 1989 | - connection, | |
| 1990 | - headers_for_send_file(if recognize_mime_type_from_uri(desc,uri) is | |
| 1991 | - { | |
| 1992 | - failure then "application/octet-stream" | |
| 1993 | - success(mime_type) then mime_type | |
| 1994 | - }, | |
| 1995 | - size), | |
| 1996 | - size, | |
| 1997 | - file(f), | |
| 1998 | - uri, | |
| 1999 | - action_before_send_file) | |
| 2000 | - } | |
| 2001 | - ) | |
| 2002 | - else log_journal_msg(desc,"Cannot find or read authorization file.\n") | |
| 2003 | - }. | |
| 2004 | - | |
| 2005 | - | |
| 2006 | - define One | |
| 2007 | - send_file | |
| 2008 | - ( | |
| 2009 | - Web_Site_Description desc, | |
| 2010 | - Connection connection, | |
| 2011 | - String uri, | |
| 2012 | - Maybe(String) mbauthorization, | |
| 2013 | - One -> One action_before_send_file | |
| 2014 | - ) = | |
| 2015 | - if mbauthorization is | |
| 2016 | - { | |
| 2017 | - //--- file without authorization: take it from public --- | |
| 2018 | - failure then if recognize_mime_type_from_uri(desc,uri) is | |
| 2019 | - { | |
| 2020 | - failure then log_journal_msg(desc,"No MIME type found for '"+uri+"'.\n"), | |
| 2021 | - success(mime_type) then | |
| 2022 | - with path = site_directory(desc)+"/public"+uri, | |
| 2023 | - if (Maybe(RStream))connect to file path is | |
| 2024 | - { | |
| 2025 | - failure then log_journal_msg(desc,"Cannot find file '"+path+"'.\n"), | |
| 2026 | - success(f) then with size = file_size(f), | |
| 2027 | - send_file(desc, | |
| 2028 | - connection, | |
| 2029 | - headers_for_send_file(mime_type,size), | |
| 2030 | - size, | |
| 2031 | - file(f), | |
| 2032 | - uri, | |
| 2033 | - action_before_send_file) | |
| 2034 | - } | |
| 2035 | - }, | |
| 2036 | - | |
| 2037 | - //--- file with authorization: apply 'private download' mecanism --- | |
| 2038 | - success(authorization) then | |
| 2039 | - if is_SSL(connection) | |
| 2040 | - then ( | |
| 2041 | - with private_download_dir = site_directory(desc)+"/private_download", | |
| 2042 | - if (RetrieveResult(String))retrieve(private_download_dir+"/z"+authorization) | |
| 2043 | - is ok(absolute_path) | |
| 2044 | - then ( | |
| 2045 | - with new_hash = compute_authorization(authorization_secret(desc), | |
| 2046 | - absolute_path), | |
| 2047 | - if (Maybe(RStream))connect to file absolute_path is | |
| 2048 | - { | |
| 2049 | - failure then log_journal_msg(desc,"Cannot find file '"+absolute_path+"'.\n"), | |
| 2050 | - success(f) then with size = file_size(f), | |
| 2051 | - send_file(desc, | |
| 2052 | - connection, | |
| 2053 | - headers_for_send_file(if recognize_mime_type_from_uri(desc,uri) is | |
| 2054 | - { | |
| 2055 | - failure then "application/octet-stream" | |
| 2056 | - success(mime_type) then mime_type | |
| 2057 | - }, | |
| 2058 | - size), | |
| 2059 | - size, | |
| 2060 | - file(f), | |
| 2061 | - uri, | |
| 2062 | - action_before_send_file) | |
| 2063 | - } | |
| 2064 | - ) | |
| 2065 | - else log_journal_msg(desc,"Cannot find or read authorization file.\n") | |
| 2066 | - ) | |
| 2067 | - else //-- file with authorization, but under HTTP: take it from private_download | |
| 2068 | - if recognize_mime_type_from_uri(desc,uri) is | |
| 2069 | - { | |
| 2070 | - failure then log_journal_msg(desc,"No MIME type found for '"+uri+"'.\n"), | |
| 2071 | - success(mime_type) then | |
| 2072 | - with path = site_directory(desc)+"/private_download"+uri, | |
| 2073 | - if (Maybe(RStream))connect to file path is | |
| 2074 | - { | |
| 2075 | - failure then log_journal_msg(desc,"Cannot find file '"+path+"'.\n"), | |
| 2076 | - success(f) then with size = file_size(f), | |
| 2077 | - send_file(desc, | |
| 2078 | - connection, | |
| 2079 | - headers_for_send_file(mime_type,size), | |
| 2080 | - size, | |
| 2081 | - file(f), | |
| 2082 | - uri, | |
| 2083 | - action_before_send_file) | |
| 2084 | - } | |
| 2085 | - } | |
| 2086 | - }. | |
| 2087 | - | |
| 2088 | - | |
| 2089 | - | |
| 2090 | - | |
| 2091 | - | |
| 2092 | - | |
| 2093 | - | |
| 2094 | - *** [5.6] Answering a www-url encoded request. | |
| 2095 | - | |
| 2096 | - Standard headers are for answering ".awp" requests. | |
| 2097 | - | |
| 2098 | -define List(HTTP_header) | |
| 2099 | - standard_headers | |
| 2100 | - ( | |
| 2101 | - Int32 answer_body_size, | |
| 2102 | - String charset | |
| 2103 | - ) = | |
| 2104 | - [ | |
| 2105 | - //http_header("Content-Type","text/html"), | |
| 2106 | - http_header("Content-Type","text/html; charset="+charset), | |
| 2107 | - http_header("Content-length",integer_to_string(answer_body_size)) | |
| 2108 | - ]. | |
| 2109 | - | |
| 2110 | - | |
| 2111 | -define One | |
| 2112 | - www_url_answer | |
| 2113 | - ( | |
| 2114 | - String host_name, | |
| 2115 | - Web_Site_Description desc, | |
| 2116 | - Connection connection, // with the client | |
| 2117 | - Int32 ip_addr, // of the client | |
| 2118 | - HTTP_RequestLine request_line, | |
| 2119 | - List(HTTP_header) headers, | |
| 2120 | - ByteArray body, | |
| 2121 | - One -> String generate_tt // trust ticket generation | |
| 2122 | - ) = | |
| 2123 | - with all_web_args = query_string(request_line) + | |
| 2124 | - read_www_url_encoded_web_args(to_string(body),0), | |
| 2125 | - uri = uri(request_line), | |
| 2126 | - ext = get_uri_extension(uri), | |
| 2127 | - (if member(journal_extensions(desc),ext) | |
| 2128 | - then log_journal_msg(desc, | |
| 2129 | - format_request(desc,connection,request_line,headers,all_web_args)) | |
| 2130 | - else unique); | |
| 2131 | - if is_illegal_uri(uri,0) | |
| 2132 | - then log_journal_msg(desc,"Received illegal URI: "+uri+"\n") | |
| 2133 | - else (if (ext = ".awp" | ext = "") | |
| 2134 | - then (with answer_headers_body = awp_handler(desc)(host_name, | |
| 2135 | - http_info(ip_addr,uri,headers,generate_tt), | |
| 2136 | - all_web_args, | |
| 2137 | - is_SSL(connection)), | |
| 2138 | - if answer_headers_body is (additional_headers,answer_body) then | |
| 2139 | - forget(reliable_write(connection, | |
| 2140 | - [ "HTTP/1.1 200 OK", crlf, | |
| 2141 | - format_headers(standard_headers(length(answer_body),charset(desc))), | |
| 2142 | - format_headers(additional_headers), | |
| 2143 | - crlf . | |
| 2144 | - answer_body]))) | |
| 2145 | - else (send_file(desc, | |
| 2146 | - connection, | |
| 2147 | - uri, | |
| 2148 | - if web_arg_value(all_web_args,"zauth") is | |
| 2149 | - { | |
| 2150 | - not_found then failure, | |
| 2151 | - found(v) then success(v) | |
| 2152 | - }, | |
| 2153 | - (One u) |-> before_send_file(desc)(all_web_args)))). | |
| 2154 | - | |
| 2155 | - | |
| 2156 | - | |
| 2157 | - | |
| 2158 | - | |
| 2159 | - | |
| 2160 | - | |
| 2161 | - *** [5.7] Answering a multipart/form-data encoded request. | |
| 2162 | - | |
| 2163 | - In order to support upload of files, we must be able to read web arguments which are | |
| 2164 | - encoded in a multipart/form-data body. The first thing to do is to find the | |
| 2165 | - boundary. The boundary is a special string which delimits the various parts of the | |
| 2166 | - 'multipart' body. It is found within the value of the 'Content-Type' HTTP header, as | |
| 2167 | - the value of the 'boundary' attribute. | |
| 2168 | - | |
| 2169 | - | |
| 2170 | - | |
| 2171 | - | |
| 2172 | - | |
| 2173 | - *** [5.7.1] Finding the boundary. | |
| 2174 | - | |
| 2175 | - Hence, we just have to find the string 'boundary=' within the value of the | |
| 2176 | - 'Content-Type' header, and read the value of the boundary from there. | |
| 2177 | - | |
| 2178 | -define Bool | |
| 2179 | - delimits_boundary | |
| 2180 | - ( | |
| 2181 | - Word8 c | |
| 2182 | - ) = | |
| 2183 | - if c = ' ' then true else | |
| 2184 | - if c = 13 then true else | |
| 2185 | - if c = 10 then true else | |
| 2186 | - if c = 0 then true else | |
| 2187 | - if c = ',' then true else | |
| 2188 | - c = ';'. | |
| 2189 | - | |
| 2190 | - | |
| 2191 | -define Maybe(String) | |
| 2192 | - get_boundary_value_3 | |
| 2193 | - ( | |
| 2194 | - String s, | |
| 2195 | - Int32 i, | |
| 2196 | - List(Word8) so_far | |
| 2197 | - ) = | |
| 2198 | - if nth(i,s) is | |
| 2199 | - { | |
| 2200 | - failure then success(implode(reverse(so_far))), | |
| 2201 | - success(c) then | |
| 2202 | - if delimits_boundary(c) | |
| 2203 | - then success(implode(reverse(so_far))) | |
| 2204 | - else get_boundary_value_3(s,i+1,[c . so_far]) | |
| 2205 | - }. | |
| 2206 | - | |
| 2207 | - | |
| 2208 | - | |
| 2209 | -define Maybe(String) | |
| 2210 | - get_boundary_value_2 | |
| 2211 | - ( | |
| 2212 | - String s, | |
| 2213 | - Int32 i, | |
| 2214 | - ) = | |
| 2215 | - if nth(i,s) is | |
| 2216 | - { | |
| 2217 | - failure then failure, | |
| 2218 | - success(c) then | |
| 2219 | - if is_blank(c) | |
| 2220 | - then get_boundary_value_2(s,i+1) | |
| 2221 | - else get_boundary_value_3(s,i+1,[c]) | |
| 2222 | - }. | |
| 2223 | - | |
| 2224 | -define Maybe(String) | |
| 2225 | - get_boundary_value_1 | |
| 2226 | - ( | |
| 2227 | - String s, // string into which we must find '= ...' | |
| 2228 | - Int32 i // position of start of search | |
| 2229 | - ) = | |
| 2230 | - if nth(i,s) is | |
| 2231 | - { | |
| 2232 | - failure then failure, | |
| 2233 | - success(c) then | |
| 2234 | - if is_blank(c) | |
| 2235 | - then get_boundary_value_1(s,i+1) | |
| 2236 | - else if c = '=' | |
| 2237 | - then get_boundary_value_2(s,i+1) | |
| 2238 | - else failure | |
| 2239 | - }. | |
| 2240 | - | |
| 2241 | - | |
| 2242 | -define Maybe(String) | |
| 2243 | - get_boundary | |
| 2244 | - ( | |
| 2245 | - String content_type_header_value | |
| 2246 | - ) = | |
| 2247 | - if find("boundary",content_type_header_value,0) is | |
| 2248 | - { | |
| 2249 | - failure then failure, | |
| 2250 | - success(n) then // 'boundary' has been found at position n | |
| 2251 | - get_boundary_value_1(content_type_header_value,n+8) | |
| 2252 | - }. | |
| 2253 | - | |
| 2254 | -define Maybe(String) | |
| 2255 | - get_boundary | |
| 2256 | - ( | |
| 2257 | - List(HTTP_header) headers | |
| 2258 | - ) = | |
| 2259 | - if headers is | |
| 2260 | - { | |
| 2261 | - [ ] then failure, | |
| 2262 | - [h . t] then if h is http_header(name,value) then | |
| 2263 | - if name = "content-type" | |
| 2264 | - then get_boundary(value) | |
| 2265 | - else get_boundary(t) | |
| 2266 | - }. | |
| 2267 | - | |
| 2268 | - | |
| 2269 | - | |
| 2270 | - | |
| 2271 | - | |
| 2272 | - | |
| 2273 | - | |
| 2274 | - | |
| 2275 | - *** [5.7.2] Reading attributes from a multipart entity. | |
| 2276 | - | |
| 2277 | - Entities in a multipart/form-data body are separated by instances of the string: | |
| 2278 | - | |
| 2279 | - --bbbbb | |
| 2280 | - | |
| 2281 | - where bbbbb is the boundary computed above. Actually, the body has the form: | |
| 2282 | - | |
| 2283 | - --bbbbb | |
| 2284 | - <entity 1> | |
| 2285 | - --bbbbb | |
| 2286 | - <entity 2> | |
| 2287 | - --bbbbb | |
| 2288 | - ... | |
| 2289 | - --bbbbb | |
| 2290 | - <last entity> | |
| 2291 | - --bbbbb | |
| 2292 | - | |
| 2293 | - | |
| 2294 | - We have to extract an entity which is in the body between offsets 'start' and 'end' | |
| 2295 | - (computed when boundaries have been localized). The entity itself is made of two parts: | |
| 2296 | - headers and body. The body is separated from the headers by a blank line. This blank | |
| 2297 | - line (a double crlf) marks the beginning of the body of the entity. Within the headers | |
| 2298 | - of the entity, we look for a 'Content-Disposition' header, which should look like this: | |
| 2299 | - | |
| 2300 | - Content-Disposition: form-data; name="..."; filename="..." crlf | |
| 2301 | - | |
| 2302 | - We are just interested in the name and the file name. Hence we first search | |
| 2303 | - 'Content-Disposition', then we search 'name' and read the value, and we do the same for | |
| 2304 | - 'filename'. | |
| 2305 | - | |
| 2306 | - If the 'filename' attribute is not present, the web arg is an ordinary one, otherwise, | |
| 2307 | - it is an uploaded file. | |
| 2308 | - | |
| 2309 | - | |
| 2310 | - Below is a variant of 'find' (see 'tools/findstring.anubis'), with an extra 'end' | |
| 2311 | - argument. | |
| 2312 | - | |
| 2313 | -define Maybe(Int32) | |
| 2314 | - find | |
| 2315 | - ( | |
| 2316 | - String what, | |
| 2317 | - ByteArray where, | |
| 2318 | - Int32 start, | |
| 2319 | - Int32 end | |
| 2320 | - ) = | |
| 2321 | - if find(to_byte_array(what),where,start) is | |
| 2322 | - { | |
| 2323 | - failure then failure, | |
| 2324 | - success(n) then | |
| 2325 | - if n+length(what) >= end | |
| 2326 | - then failure | |
| 2327 | - else success(n) | |
| 2328 | - }. | |
| 2329 | - | |
| 2330 | - | |
| 2331 | -define String | |
| 2332 | - read_attribute_value | |
| 2333 | - ( | |
| 2334 | - ByteArray where, | |
| 2335 | - Int32 start, | |
| 2336 | - Int32 end, | |
| 2337 | - List(Word8) so_far | |
| 2338 | - ) = | |
| 2339 | - if start >= end then implode(reverse(so_far)) else | |
| 2340 | - if nth(start,where) is | |
| 2341 | - { | |
| 2342 | - failure then implode(reverse(so_far)), | |
| 2343 | - success(c) then | |
| 2344 | - if c = '\"' | |
| 2345 | - then implode(reverse(so_far)) | |
| 2346 | - else read_attribute_value(where,start+1,end,[c . so_far]) | |
| 2347 | - }. | |
| 2348 | - | |
| 2349 | -define Maybe(String) | |
| 2350 | - find_attribute | |
| 2351 | - ( | |
| 2352 | - String name, | |
| 2353 | - ByteArray where, | |
| 2354 | - Int32 start, | |
| 2355 | - Int32 end | |
| 2356 | - ) = | |
| 2357 | - with name = name+"=\"", | |
| 2358 | - if find(to_byte_array(name),where,start) is | |
| 2359 | - { | |
| 2360 | - failure then failure, | |
| 2361 | - success(n) then | |
| 2362 | - if n+length(name) >= end | |
| 2363 | - then failure | |
| 2364 | - else success(read_attribute_value(where,n+length(name),end,[])) | |
| 2365 | - }. | |
| 2366 | - | |
| 2367 | - | |
| 2368 | - | |
| 2369 | -define Maybe((String,Maybe(String))) | |
| 2370 | - find_name_and_filename | |
| 2371 | - ( | |
| 2372 | - ByteArray body, | |
| 2373 | - Int32 start, | |
| 2374 | - Int32 end | |
| 2375 | - ) = | |
| 2376 | - if find(to_byte_array("Content-Disposition"),body,start) is | |
| 2377 | - { | |
| 2378 | - failure then failure, | |
| 2379 | - success(n) then | |
| 2380 | - if find_attribute("name",body,n+19,end) is | |
| 2381 | - { | |
| 2382 | - failure then failure, | |
| 2383 | - success(name_value) then if find_attribute("filename",body,n+19,end) is | |
| 2384 | - { | |
| 2385 | - failure then success((name_value,failure)), | |
| 2386 | - success(filename_value) then success((name_value,success(filename_value))) | |
| 2387 | - } | |
| 2388 | - } | |
| 2389 | - }. | |
| 2390 | - | |
| 2391 | - | |
| 2392 | - | |
| 2393 | - | |
| 2394 | - | |
| 2395 | - | |
| 2396 | - | |
| 2397 | - | |
| 2398 | - | |
| 2399 | - | |
| 2400 | - *** [5.7.3] Creating a temporary filename for an uploaded file. | |
| 2401 | - | |
| 2402 | -variable Int32 uploaded_file_count = 0. | |
| 2403 | - | |
| 2404 | - This variable is local to the virtual machine. Hence, its value is 0 each time a new | |
| 2405 | - requests arrives. Temporary uploaded files are stored in the directory represented by | |
| 2406 | - 'upload_temporary_directory'. The filenames have the form: | |
| 2407 | - | |
| 2408 | - _m_n | |
| 2409 | - | |
| 2410 | - where 'm' is the number of the virtual machine, and 'n' a number obtained by | |
| 2411 | - incrementing 'uploaded_file_count'. Notice that the program must do something with this | |
| 2412 | - file (move it to some directory/name), otherwise, it will probably be overwritten the | |
| 2413 | - next time the same machine works. | |
| 2414 | - | |
| 2415 | - | |
| 2416 | - | |
| 2417 | - | |
| 2418 | - | |
| 2419 | - | |
| 2420 | - *** [5.7.4] Saving an uploaded file under a temporary filename. | |
| 2421 | - | |
| 2422 | -define Maybe(String) // returns the temporary file name | |
| 2423 | - save_uploaded_file | |
| 2424 | - ( | |
| 2425 | - Web_Site_Description desc, | |
| 2426 | - ByteArray body, | |
| 2427 | - Int32 start, | |
| 2428 | - Int32 end | |
| 2429 | - ) = | |
| 2430 | - uploaded_file_count <- 1 + *uploaded_file_count; | |
| 2431 | - with tfn = "_"+integer_to_string(virtual_machine_id)+"_"+integer_to_string(*uploaded_file_count), | |
| 2432 | - if (Maybe(WStream))connect to file site_directory(desc)+"/upload_temporary/"+tfn is | |
| 2433 | - { | |
| 2434 | - failure then failure, | |
| 2435 | - success(f) then | |
| 2436 | - if reliable_write(file(f),extract(body,start,end)) is | |
| 2437 | - { | |
| 2438 | - failure then failure, | |
| 2439 | - success(nw) then | |
| 2440 | - if nw = end - start | |
| 2441 | - then success(tfn) | |
| 2442 | - else failure | |
| 2443 | - } | |
| 2444 | - }. | |
| 2445 | - | |
| 2446 | - | |
| 2447 | - | |
| 2448 | - | |
| 2449 | - | |
| 2450 | - | |
| 2451 | - | |
| 2452 | - | |
| 2453 | - *** [5.7.5] Removing the path from a file name. | |
| 2454 | - | |
| 2455 | - When a file is uploaded, the browser sends the complete path of the file on the client | |
| 2456 | - machine as the file name. Actually, this is not quite normal. Nevertheless, we need to | |
| 2457 | - remove the path, and keep only the file name. This is achieved by 'remove_path' below. | |
| 2458 | - | |
| 2459 | -define Int32 | |
| 2460 | - file_name_begin | |
| 2461 | - ( | |
| 2462 | - String full_name, | |
| 2463 | - Int32 i | |
| 2464 | - ) = | |
| 2465 | - if nth(i,full_name) is | |
| 2466 | - { | |
| 2467 | - failure then 0, | |
| 2468 | - success(c) then | |
| 2469 | - if c = '/' then i+1 else | |
| 2470 | - if c = '\\' then i+1 else | |
| 2471 | - file_name_begin(full_name,i-1) | |
| 2472 | - }. | |
| 2473 | - | |
| 2474 | -define String | |
| 2475 | - remove_path | |
| 2476 | - ( | |
| 2477 | - String full_name | |
| 2478 | - ) = | |
| 2479 | - with l = length(full_name), | |
| 2480 | - b = file_name_begin(full_name,l-1), | |
| 2481 | - substr(full_name,b,l-b). | |
| 2482 | - | |
| 2483 | - | |
| 2484 | - | |
| 2485 | - | |
| 2486 | - | |
| 2487 | - *** [5.7.6] Reading a multipart entity. | |
| 2488 | - | |
| 2489 | -define Maybe(Web_arg) | |
| 2490 | - get_multipart_entity | |
| 2491 | - ( | |
| 2492 | - Web_Site_Description desc, | |
| 2493 | - ByteArray body, | |
| 2494 | - Int32 start, | |
| 2495 | - Int32 end | |
| 2496 | - ) = | |
| 2497 | - if find(to_byte_array(crlf+crlf),body,start) is | |
| 2498 | - { | |
| 2499 | - failure then failure, | |
| 2500 | - success(k) then | |
| 2501 | - if k >= end // must be within this entity, not the next one | |
| 2502 | - then failure | |
| 2503 | - else if find_name_and_filename(body,start,k) is | |
| 2504 | - { | |
| 2505 | - failure then failure, | |
| 2506 | - success(n_mbfn) then if n_mbfn is (name,mbfn) then | |
| 2507 | - if mbfn is | |
| 2508 | - { | |
| 2509 | - failure then | |
| 2510 | - success(web_arg(name,to_string(extract(body,k+4,end-2)))), | |
| 2511 | - // we must substract 2 to end because of crlf just before the boundary | |
| 2512 | - | |
| 2513 | - success(fn) then | |
| 2514 | - if save_uploaded_file(desc,body,k+4,end-2) is | |
| 2515 | - { | |
| 2516 | - failure then failure, | |
| 2517 | - success(tfn) then | |
| 2518 | - success(upload(name,remove_path(fn), | |
| 2519 | - site_directory(desc)+"/upload_temporary/"+tfn)) | |
| 2520 | - | |
| 2521 | - } | |
| 2522 | - } | |
| 2523 | - } | |
| 2524 | - }. | |
| 2525 | - | |
| 2526 | - | |
| 2527 | - | |
| 2528 | -define List(Web_arg) | |
| 2529 | - read_multipart_form_data_encoded_web_args | |
| 2530 | - ( | |
| 2531 | - Web_Site_Description desc, | |
| 2532 | - ByteArray body, | |
| 2533 | - ByteArray __boundary, | |
| 2534 | - Int32 i, | |
| 2535 | - ) = | |
| 2536 | - if find(__boundary,body,i) is | |
| 2537 | - { | |
| 2538 | - failure then [ ], | |
| 2539 | - success(n) then | |
| 2540 | - if find(__boundary,body,n+length(__boundary)) is | |
| 2541 | - { | |
| 2542 | - failure then [ ], | |
| 2543 | - success(m) then | |
| 2544 | - if get_multipart_entity(desc,body,n+length(__boundary),m) is | |
| 2545 | - { | |
| 2546 | - failure then [ ], | |
| 2547 | - success(wa) then | |
| 2548 | - [wa . read_multipart_form_data_encoded_web_args(desc,body,__boundary,m)] | |
| 2549 | - } | |
| 2550 | - } | |
| 2551 | - }. | |
| 2552 | - | |
| 2553 | - | |
| 2554 | - | |
| 2555 | -define One | |
| 2556 | - multipart_form_data_answer | |
| 2557 | - ( | |
| 2558 | - String host_name, | |
| 2559 | - Web_Site_Description desc, | |
| 2560 | - Connection connection, | |
| 2561 | - Int32 ip_addr, | |
| 2562 | - HTTP_RequestLine request_line, | |
| 2563 | - List(HTTP_header) headers, | |
| 2564 | - ByteArray body, | |
| 2565 | - One -> String generate_tt | |
| 2566 | - ) = | |
| 2567 | - if get_boundary(headers) is | |
| 2568 | - { | |
| 2569 | - failure then unique, | |
| 2570 | - success(boundary) then | |
| 2571 | - with all_web_args = query_string(request_line) + | |
| 2572 | - read_multipart_form_data_encoded_web_args(desc, | |
| 2573 | - body, | |
| 2574 | - to_byte_array("--"+boundary), | |
| 2575 | - 0), | |
| 2576 | - uri = uri(request_line), | |
| 2577 | - ext = get_uri_extension(uri), | |
| 2578 | - log_journal_msg(desc, | |
| 2579 | - format_request(desc,connection,request_line,headers,all_web_args)); | |
| 2580 | - if is_illegal_uri(uri,0) | |
| 2581 | - then log_journal_msg(desc,"Received illegal URI: "+uri+"\n") | |
| 2582 | - else | |
| 2583 | - if (ext = ".awp" | ext = "") then | |
| 2584 | - (with answer_headers_body = awp_handler(desc)(host_name, | |
| 2585 | - http_info(ip_addr,uri,headers,generate_tt), | |
| 2586 | - all_web_args, | |
| 2587 | - is_SSL(connection)), | |
| 2588 | - if answer_headers_body is (additional_headers,answer_body) then | |
| 2589 | - forget(reliable_write(connection, | |
| 2590 | - [ "HTTP/1.1 200 OK",crlf, | |
| 2591 | - format_headers(standard_headers(length(answer_body),charset(desc))), | |
| 2592 | - format_headers(additional_headers), | |
| 2593 | - crlf . | |
| 2594 | - answer_body]))) | |
| 2595 | - else unique | |
| 2596 | - }. | |
| 2597 | - | |
| 2598 | - | |
| 2599 | - | |
| 2600 | - | |
| 2601 | - | |
| 2602 | - | |
| 2603 | - | |
| 2604 | - | |
| 2605 | - *** [5.8] Handling redirections. | |
| 2606 | - | |
| 2607 | - 'redirections' (of type 'List(Redirection)') contains redirection directives. Each one | |
| 2608 | - has the form: | |
| 2609 | - | |
| 2610 | - redirect(required_uri,required_host,corresponding_uri). | |
| 2611 | - | |
| 2612 | - The host required by the client may be found in the 'Host' HTTP header. The URI | |
| 2613 | - required by the client is given below as 'uri'. We just have to find the required host | |
| 2614 | - in the headers, and to find the corresponding redirection directive. | |
| 2615 | - | |
| 2616 | - | |
| 2617 | - In the next fonction, the required host and URI are known. We just have to search in | |
| 2618 | - the 'redirections' list. | |
| 2619 | - | |
| 2620 | -define String | |
| 2621 | - handle_redirection | |
| 2622 | - ( | |
| 2623 | - String required_uri, | |
| 2624 | - String required_host, | |
| 2625 | - List(Redirection) redirections | |
| 2626 | - ) = | |
| 2627 | - if redirections is | |
| 2628 | - { | |
| 2629 | - [ ] then required_uri, | |
| 2630 | - [h . t] then if h is redirect(uri,host,target) then | |
| 2631 | - if host = required_host | |
| 2632 | - then if uri = required_uri | |
| 2633 | - then target | |
| 2634 | - else handle_redirection(required_uri,required_host,t) | |
| 2635 | - else handle_redirection(required_uri,required_host,t) | |
| 2636 | - }. | |
| 2637 | - | |
| 2638 | - | |
| 2639 | - | |
| 2640 | - The host name may be encumbered by a port number, like | |
| 2641 | - | |
| 2642 | - www.our-business.com:1607 | |
| 2643 | - | |
| 2644 | - We must remove this port number, otherwise the host name may not be recognized. | |
| 2645 | - | |
| 2646 | -define String | |
| 2647 | - strip_port | |
| 2648 | - ( | |
| 2649 | - String name, | |
| 2650 | - Int32 i | |
| 2651 | - ) = | |
| 2652 | - if nth(i,name) is | |
| 2653 | - { | |
| 2654 | - failure then name, | |
| 2655 | - success(c) then | |
| 2656 | - if c = ':' | |
| 2657 | - then substr(name,0,i) | |
| 2658 | - else strip_port(name,i+1) | |
| 2659 | - }. | |
| 2660 | - | |
| 2661 | - | |
| 2662 | - | |
| 2663 | - | |
| 2664 | - | |
| 2665 | - Finding the 'Host' header. No redirection is performed if this header is not found. | |
| 2666 | - | |
| 2667 | -define String | |
| 2668 | - handle_redirection // returns the redirected URI | |
| 2669 | - ( | |
| 2670 | - List(Redirection) redirections, | |
| 2671 | - String uri, // original URI | |
| 2672 | - List(HTTP_header) headers | |
| 2673 | - ) = | |
| 2674 | - if headers is | |
| 2675 | - { | |
| 2676 | - [ ] then uri, | |
| 2677 | - [h . t] then if h is http_header(name,value) then | |
| 2678 | - if name = "host" | |
| 2679 | - then handle_redirection(uri,strip_port(value,0),redirections) | |
| 2680 | - else handle_redirection(redirections,uri,t) | |
| 2681 | - }. | |
| 2682 | - | |
| 2683 | - | |
| 2684 | - | |
| 2685 | - | |
| 2686 | - | |
| 2687 | - | |
| 2688 | - | |
| 2689 | - | |
| 2690 | - *** [5.9] Answering both sorts of requests. | |
| 2691 | - | |
| 2692 | - We must decide if the request is www-url encoded or multipart/form-data encoded. This | |
| 2693 | - is achieved through the header 'Content-Type'. | |
| 2694 | - | |
| 2695 | -define EncodingType | |
| 2696 | - get_encoding_type | |
| 2697 | - ( | |
| 2698 | - List(HTTP_header) headers | |
| 2699 | - ) = | |
| 2700 | - if headers is | |
| 2701 | - { | |
| 2702 | - [ ] then www_url, // this is the default | |
| 2703 | - [h . t] then if h is http_header(name,value) then | |
| 2704 | - if name = "content-type" | |
| 2705 | - then if find("multipart/form-data",value,0) is | |
| 2706 | - { | |
| 2707 | - failure then www_url, | |
| 2708 | - success(_) then multipart_form_data | |
| 2709 | - } | |
| 2710 | - else get_encoding_type(t) | |
| 2711 | - }. | |
| 2712 | - | |
| 2713 | - | |
| 2714 | - | |
| 2715 | -define One | |
| 2716 | - send_answer | |
| 2717 | - ( | |
| 2718 | - String host_name, | |
| 2719 | - Web_Site_Description desc, | |
| 2720 | - Connection connection, | |
| 2721 | - HTTP_RequestLine rqline, | |
| 2722 | - List(HTTP_header) headers, | |
| 2723 | - ByteArray body, | |
| 2724 | - One -> String generate_tt | |
| 2725 | - ) = | |
| 2726 | - if rqline is request_line(type,uri,qstring) then | |
| 2727 | - with rqline = request_line(type,handle_redirection(redirections(desc),uri,headers),qstring), | |
| 2728 | - if remote_IP_address_and_port(connection) is (ip_addr,_) then | |
| 2729 | - if get_encoding_type(headers) is | |
| 2730 | - { | |
| 2731 | - www_url then | |
| 2732 | - www_url_answer(host_name,desc,connection,ip_addr,rqline,headers,body,generate_tt), | |
| 2733 | - multipart_form_data then | |
| 2734 | - multipart_form_data_answer(host_name,desc,connection,ip_addr,rqline,headers,body,generate_tt) | |
| 2735 | - }. | |
| 2736 | - | |
| 2737 | - | |
| 2738 | - | |
| 2739 | - | |
| 2740 | - | |
| 2741 | - | |
| 2742 | - | |
| 2743 | - *** [6] The HTTP/HTTPS server. | |
| 2744 | - | |
| 2745 | - The command 'start_server' (declared in 'predefined.anubis') starts a virtual machine | |
| 2746 | - which opens a server TCP/IP connection, and which continuously listens to this | |
| 2747 | - connection. When a request arrives, this machine delegates the work of deciphering and | |
| 2748 | - answering the request to another virtual machine, and continues to listen. The job of | |
| 2749 | - the delegated machine is defined by the HTTP request handler below. | |
| 2750 | - | |
| 2751 | - | |
| 2752 | - | |
| 2753 | - | |
| 2754 | - | |
| 2755 | - *** [6.1] Determining the requested host. | |
| 2756 | - | |
| 2757 | - When a request arrives to one of our two servers, we must decide which site (host) is | |
| 2758 | - requested. | |
| 2759 | - | |
| 2760 | -define Maybe(String) | |
| 2761 | - get_host_header_value | |
| 2762 | - ( | |
| 2763 | - List(HTTP_header) headers | |
| 2764 | - ) = | |
| 2765 | - if headers is | |
| 2766 | - { | |
| 2767 | - [ ] then failure, | |
| 2768 | - [h . t] then if h is http_header(name,value) then | |
| 2769 | - if name = "host" | |
| 2770 | - then success(strip_port(value,0)) | |
| 2771 | - else get_host_header_value(t) | |
| 2772 | - }. | |
| 2773 | - | |
| 2774 | -define Maybe((String,Web_Site_Description)) | |
| 2775 | - get_site | |
| 2776 | - ( | |
| 2777 | - String requested_host, | |
| 2778 | - List(Web_Site_Description) sites | |
| 2779 | - ) = | |
| 2780 | - if sites is | |
| 2781 | - { | |
| 2782 | - [ ] then print("Requested host '"+requested_host+"' does not exist.\n"); failure, | |
| 2783 | - [site1 . others] then | |
| 2784 | - if site1 is web_site_description(common_names,_,_,_,_,_,_,_,_,_) then | |
| 2785 | - if member(common_names,requested_host) | |
| 2786 | - then success((requested_host,site1)) | |
| 2787 | - else get_site(requested_host,others) | |
| 2788 | - }. | |
| 2789 | - | |
| 2790 | - | |
| 2791 | -define Maybe((String,Web_Site_Description)) | |
| 2792 | - get_site | |
| 2793 | - ( | |
| 2794 | - List(HTTP_header) headers, | |
| 2795 | - List(Web_Site_Description) sites | |
| 2796 | - ) = | |
| 2797 | - if get_host_header_value(headers) is | |
| 2798 | - { | |
| 2799 | - failure then print("No 'Host' HTTP header.\n"); failure, | |
| 2800 | - success(requested_host) then | |
| 2801 | - //here we treat the case with only one site. hence we accept any host request | |
| 2802 | - //print("*** there is " +length(sites) + " sites \n"); | |
| 2803 | - if length(sites) = 1 then | |
| 2804 | - with site = force_nth(0, sites), | |
| 2805 | - //print("ONE server OK\n"); | |
| 2806 | - success((requested_host, site)) | |
| 2807 | - else | |
| 2808 | - get_site(requested_host,sites) | |
| 2809 | - }. | |
| 2810 | - | |
| 2811 | - | |
| 2812 | - | |
| 2813 | - | |
| 2814 | - | |
| 2815 | - *** [6.2] The HTTP request handler. | |
| 2816 | - | |
| 2817 | - Here is the HTTP/HTTPS handler. It is called at each new request in a separate virtual | |
| 2818 | - machine. It reads the headers of the HTTP request, determines the host, determines body | |
| 2819 | - size, reads the body of the HTTP request, and answers the request. | |
| 2820 | - | |
| 2821 | - | |
| 2822 | - | |
| 2823 | -define One -> String make_generate_trust_ticket(DenialOfService dos). | |
| 2824 | - | |
| 2825 | - | |
| 2826 | -define One | |
| 2827 | - http_https_handler | |
| 2828 | - ( | |
| 2829 | - List(Web_Site_Description) sites, | |
| 2830 | - Connection connection, | |
| 2831 | - Bool is_https, | |
| 2832 | - DenialOfService dos | |
| 2833 | - ) = | |
| 2834 | - with start_time = (Int32)now, | |
| 2835 | - sttm <- start_time; | |
| 2836 | - if dos is denial_of_service(mc_v,rld_v,hd_v,ad_v,ld_v,ra_v) then | |
| 2837 | - if remote_IP_address_and_port(connection) is (ip_addr,port) then | |
| 2838 | - if read_request_line(connection,start_time+*rld_v,dos) is | |
| 2839 | - { | |
| 2840 | - error(msg) then print(format(msg)), | |
| 2841 | - ok(request_line) then | |
| 2842 | - if read_http_headers(connection,start_time+*hd_v,dos) is | |
| 2843 | - { | |
| 2844 | - error(msg) then print(format(msg)), | |
| 2845 | - ok(headers) then if get_site(headers,sites) is | |
| 2846 | - { | |
| 2847 | - failure then unique, | |
| 2848 | - success(p) then if p is (host_name,desc) then | |
| 2849 | - if get_body_size(headers) is | |
| 2850 | - { | |
| 2851 | - error(msg) then log_journal_msg(desc,format(msg)), | |
| 2852 | - ok(body_size) then | |
| 2853 | - if read_http_body(connection,body_size,constant_byte_array(0,0),1000) is | |
| 2854 | - { | |
| 2855 | - error(msg) then log_journal_msg(desc,format(msg)), | |
| 2856 | - ok(body) then | |
| 2857 | - send_answer(host_name,desc,connection,request_line,headers,body, | |
| 2858 | - make_generate_trust_ticket(dos)) | |
| 2859 | - } | |
| 2860 | - } | |
| 2861 | - } | |
| 2862 | - } | |
| 2863 | - }. | |
| 2864 | - | |
| 2865 | - | |
| 2866 | - Below are the two tools for constructing the handlers required by 'start_server' and | |
| 2867 | - 'start_ssl_server' (see 'predefined.anubis'). | |
| 2868 | - | |
| 2869 | -define Bool is_dubious_IP(Int32 ip, DenialOfService dos). | |
| 2870 | - | |
| 2871 | -define Server -> ((RWStream) -> One) | |
| 2872 | - make_http_handler | |
| 2873 | - ( | |
| 2874 | - List(Web_Site_Description) sites, | |
| 2875 | - DenialOfService dos | |
| 2876 | - ) = | |
| 2877 | - (Server server) |-> (RWStream connection) |-> | |
| 2878 | - if remote_IP_address_and_port(connection) is (addr,_) then | |
| 2879 | - if is_dubious_IP(addr,dos) | |
| 2880 | - then print("Rejecting dubious IP address "+ip_addr_to_string(addr)+"\n") | |
| 2881 | - else http_https_handler(sites,tcp(connection),false,dos). | |
| 2882 | - | |
| 2883 | -define Server -> (SSL_Connection -> One) | |
| 2884 | - make_https_handler | |
| 2885 | - ( | |
| 2886 | - List(Web_Site_Description) sites, | |
| 2887 | - DenialOfService dos | |
| 2888 | - ) = | |
| 2889 | - (Server server) |-> (SSL_Connection connection) |-> | |
| 2890 | - http_https_handler(sites,ssl(connection),true,dos). | |
| 2891 | - | |
| 2892 | - | |
| 2893 | - | |
| 2894 | - | |
| 2895 | - *** [6.3] Server's tasks. | |
| 2896 | - | |
| 2897 | - Some tasks must be executed periodically, for example for cleaning up directories from | |
| 2898 | - short life time files. | |
| 2899 | - | |
| 2900 | - The next function removes from the given directory (and recursively from its | |
| 2901 | - subdirectories) all the files which are more than 10 minutes old. | |
| 2902 | - | |
| 2903 | -define One | |
| 2904 | - cleanup_directory_10mn | |
| 2905 | - ( | |
| 2906 | - String dir // path of private download directory (or subdirectory) with trailing slash | |
| 2907 | - ) = | |
| 2908 | - forget(map((FileDescription fd) |-> if fd is | |
| 2909 | - { | |
| 2910 | - no_info(name) then forget(remove(dir+name)), | |
| 2911 | - file(name,_,_,d) then if d+600 < now then forget(remove(dir+name)) else unique, | |
| 2912 | - link(name,_,_,d) then if d+600 < now then forget(remove(dir+name)) else unique, | |
| 2913 | - directory(name,_,_) then cleanup_directory_10mn(dir+name+"/"), | |
| 2914 | - }, | |
| 2915 | - directory_full_list(dir,"*","*","*"))). | |
| 2916 | - | |
| 2917 | - | |
| 2918 | -define One | |
| 2919 | - http_servers_tasks | |
| 2920 | - ( | |
| 2921 | - List(Web_Site_Description) sites, | |
| 2922 | - List(Server) servers, | |
| 2923 | - Int32 period, | |
| 2924 | - Int32 next_time, | |
| 2925 | - ) = | |
| 2926 | - if mapand(is_down,servers) | |
| 2927 | - then unique | |
| 2928 | - else if now > next_time | |
| 2929 | - then | |
| 2930 | - ( | |
| 2931 | - /* | |
| 2932 | - forget(map((Web_Site_Description wsd) |-> | |
| 2933 | - cleanup_directory_10mn(site_directory(wsd)+"/private_download/"), | |
| 2934 | - sites)); | |
| 2935 | - */ | |
| 2936 | - http_servers_tasks(sites,servers,period,next_time+period) | |
| 2937 | - ) | |
| 2938 | - else | |
| 2939 | - ( | |
| 2940 | - sleep(1000); | |
| 2941 | - http_servers_tasks(sites,servers,period,next_time) | |
| 2942 | - ). | |
| 2943 | - | |
| 2944 | - | |
| 2945 | -public define One | |
| 2946 | - start_http_servers_tasks | |
| 2947 | - ( | |
| 2948 | - List(Web_Site_Description) sites, | |
| 2949 | - List(Server) servers, | |
| 2950 | - Int32 period | |
| 2951 | - ) = | |
| 2952 | - delegate http_servers_tasks(sites,servers,period,now), | |
| 2953 | - unique. | |
| 2954 | - | |
| 2955 | - | |
| 2956 | - | |
| 2957 | - | |
| 2958 | - *** [6.4] Protection against 'denial of service' attacks. | |
| 2959 | - | |
| 2960 | - | |
| 2961 | - *** [6.4.1] Counting connections. | |
| 2962 | - | |
| 2963 | -define Bool // returns false if the counter cannot be incremented (too many connections) | |
| 2964 | - increment_connections_counter | |
| 2965 | - ( | |
| 2966 | - Var(Int32) counter | |
| 2967 | - ) = | |
| 2968 | - protect with n = *counter, | |
| 2969 | - if n >= 100 | |
| 2970 | - then false | |
| 2971 | - else (counter <- (*counter)+1); true. | |
| 2972 | - | |
| 2973 | -define One | |
| 2974 | - decrement_connections_counter | |
| 2975 | - ( | |
| 2976 | - Var(Int32) counter | |
| 2977 | - ) = | |
| 2978 | - protect counter <- (*counter)-1. | |
| 2979 | - | |
| 2980 | - | |
| 2981 | - | |
| 2982 | - | |
| 2983 | - | |
| 2984 | - *** [6.4.2] Recording dubious IP addresses. | |
| 2985 | - | |
| 2986 | - | |
| 2987 | -define List(DubiousIP) | |
| 2988 | - record_dubious_IP | |
| 2989 | - ( | |
| 2990 | - Int32 ip, | |
| 2991 | - List(DubiousIP) l | |
| 2992 | - ) = | |
| 2993 | - if l is | |
| 2994 | - { | |
| 2995 | - [ ] then [dubious_ip(ip,now)], | |
| 2996 | - [h . t] then if h is dubious_ip(addr,time) then | |
| 2997 | - if addr = ip | |
| 2998 | - then [dubious_ip(addr,now) . t] | |
| 2999 | - else [h . record_dubious_IP(ip,t)] | |
| 3000 | - }. | |
| 3001 | - | |
| 3002 | - | |
| 3003 | -define One | |
| 3004 | - record_dubious_IP | |
| 3005 | - ( | |
| 3006 | - Int32 dubious_IP, | |
| 3007 | - Var(List(DubiousIP)) v | |
| 3008 | - ) = | |
| 3009 | - protect v <- record_dubious_IP(dubious_IP,*v). | |
| 3010 | - | |
| 3011 | - | |
| 3012 | -define One | |
| 3013 | - record_dubious_IP | |
| 3014 | - ( | |
| 3015 | - Int32 addr, | |
| 3016 | - DenialOfService dos | |
| 3017 | - ) = | |
| 3018 | - record_dubious_IP(addr,list_of_dubious(dos)). | |
| 3019 | - | |
| 3020 | - | |
| 3021 | -public define DenialOfService | |
| 3022 | - load_denial_of_service_info | |
| 3023 | - = | |
| 3024 | - if (RetrieveResult(DenialOfService))retrieve(my_anubis_directory+"/web_sites/dos_info") is | |
| 3025 | - ok(dos) then dos else denial_of_service( | |
| 3026 | - var(100), | |
| 3027 | - var(1000), | |
| 3028 | - var(1500), | |
| 3029 | - var(2000), | |
| 3030 | - var([]), | |
| 3031 | - var([])). | |
| 3032 | - | |
| 3033 | - | |
| 3034 | - | |
| 3035 | - | |
| 3036 | - *** [6.4.3] Testing if an address is dubious. | |
| 3037 | - | |
| 3038 | -define Bool | |
| 3039 | - is_dubious_IP | |
| 3040 | - ( | |
| 3041 | - Int32 ip, | |
| 3042 | - List(DubiousIP) l | |
| 3043 | - ) = | |
| 3044 | - if l is | |
| 3045 | - { | |
| 3046 | - [ ] then false, | |
| 3047 | - [h . t] then if h is dubious_ip(addr,time) then | |
| 3048 | - if ip = addr | |
| 3049 | - then true | |
| 3050 | - else is_dubious_IP(ip,t) | |
| 3051 | - }. | |
| 3052 | - | |
| 3053 | - | |
| 3054 | -define Bool | |
| 3055 | - is_dubious_IP | |
| 3056 | - ( | |
| 3057 | - Int32 ip, | |
| 3058 | - DenialOfService dos | |
| 3059 | - ) = | |
| 3060 | - if dos is | |
| 3061 | - { | |
| 3062 | - denial_of_service(mc_v,rld_v,hd_v,ad_v,ld_v,ra_v) then | |
| 3063 | - if member(*ra_v,ip) then false else | |
| 3064 | - is_dubious_IP(ip,*ld_v) | |
| 3065 | - }. | |
| 3066 | - | |
| 3067 | - | |
| 3068 | - | |
| 3069 | - | |
| 3070 | - *** [6.4.4] Removing inactive dubious IP addresses. | |
| 3071 | - | |
| 3072 | -define List(DubiousIP) | |
| 3073 | - remove_inactive_dubious_IP | |
| 3074 | - ( | |
| 3075 | - List(DubiousIP) l, | |
| 3076 | - Int32 ref_time, | |
| 3077 | - ) = | |
| 3078 | - if l is | |
| 3079 | - { | |
| 3080 | - [ ] then [ ], | |
| 3081 | - [h . t] then if h is dubious_ip(addr,time) then | |
| 3082 | - if time < ref_time | |
| 3083 | - then (print(ip_addr_to_string(addr)+" removed from dubious addresses list.\n"); | |
| 3084 | - remove_inactive_dubious_IP(t,ref_time)) | |
| 3085 | - else [h . remove_inactive_dubious_IP(t,ref_time)] | |
| 3086 | - }. | |
| 3087 | - | |
| 3088 | -define One | |
| 3089 | - remove_inactive_dubious_IP | |
| 3090 | - ( | |
| 3091 | - Var(List(DubiousIP)) v | |
| 3092 | - ) = | |
| 3093 | - protect | |
| 3094 | - with ref_time = now - 600, // 10 minutes | |
| 3095 | - v <- remove_inactive_dubious_IP(*v,ref_time). | |
| 3096 | - | |
| 3097 | - | |
| 3098 | - The above function will be executed periodically by the servers's tasks machine. | |
| 3099 | - | |
| 3100 | - | |
| 3101 | - | |
| 3102 | - *** [6.4.5] Making the function for generating trust tickets. | |
| 3103 | - | |
| 3104 | -define One -> String | |
| 3105 | - make_generate_trust_ticket | |
| 3106 | - ( | |
| 3107 | - DenialOfService dos | |
| 3108 | - ) = | |
| 3109 | - (One _) |-> "". | |
| 3110 | - | |
| 3111 | - | |
| 3112 | - | |
| 3113 | - | |
| 3114 | - | |
| 3115 | - | |
| 3116 | - | |
| 3117 | - *** [6.5] Starting the HTTP/HTTPS server. | |
| 3118 | - | |
| 3119 | - | |
| 3120 | - The next function creates the directories for all sites (if they don't already exist). | |
| 3121 | - | |
| 3122 | -define One | |
| 3123 | - create_directories | |
| 3124 | - ( | |
| 3125 | - List(Web_Site_Description) sites | |
| 3126 | - ) = | |
| 3127 | - if sites is | |
| 3128 | - { | |
| 3129 | - [ ] then unique, | |
| 3130 | - [s1 . others] then | |
| 3131 | - with site_dir = site_directory(s1), | |
| 3132 | - forget(make_directory(site_dir+"/public",default_directory_mode)); | |
| 3133 | - forget(make_directory(site_dir+"/upload_temporary",default_directory_mode)); | |
| 3134 | - forget(make_directory(site_dir+"/private_download",default_directory_mode)); | |
| 3135 | - forget(make_directory(site_dir+"/journal",default_directory_mode)); | |
| 3136 | - create_directories(others) | |
| 3137 | - }. | |
| 3138 | - | |
| 3139 | - | |
| 3140 | - | |
| 3141 | - | |
| 3142 | - | |
| 3143 | - Below are the commands for starting an HTTP server and an HTTPS server. | |
| 3144 | - | |
| 3145 | - | |
| 3146 | -define StartServerResult | |
| 3147 | - start_http_server | |
| 3148 | - ( | |
| 3149 | - Int32 ip_address, | |
| 3150 | - Int32 port, | |
| 3151 | - Server -> ((RWStream) -> One) handler, | |
| 3152 | - Int32 retries, | |
| 3153 | - DenialOfService dos | |
| 3154 | - ) = | |
| 3155 | - if start_server(ip_address, | |
| 3156 | - port, | |
| 3157 | - handler, | |
| 3158 | - identity) is ok(server) | |
| 3159 | - then print(" \r"); | |
| 3160 | - ok(server) | |
| 3161 | - else print("Port "+port+": retry number "+retries+"\r"); | |
| 3162 | - sleep(1000); | |
| 3163 | - start_http_server(ip_address,port,handler,retries+1,dos). | |
| 3164 | - | |
| 3165 | -public define StartServerResult | |
| 3166 | - start_http_server | |
| 3167 | - ( | |
| 3168 | - Int32 ip_address, | |
| 3169 | - Int32 port, | |
| 3170 | - List(Web_Site_Description) sites, | |
| 3171 | - DenialOfService dos | |
| 3172 | - ) = | |
| 3173 | - create_directories(sites); | |
| 3174 | - start_http_server(ip_address,port, | |
| 3175 | - make_http_handler(sites,dos), | |
| 3176 | - 0, | |
| 3177 | - dos). | |
| 3178 | - | |
| 3179 | - | |
| 3180 | - For the HTTPS server, we have a problem which is due to the fact that 'anbexec' is not | |
| 3181 | - yet able to manipulate several SSL server certificates. 'anbexec' and | |
| 3182 | - 'predefined.anubis' must be changed. Sorry ! This will be done as soon as possible. The | |
| 3183 | - 'solution' for the time being is to provide the common name of the unique SSL server | |
| 3184 | - certificate. | |
| 3185 | - | |
| 3186 | - | |
| 3187 | -define StartServerResult | |
| 3188 | - start_https_server | |
| 3189 | - ( | |
| 3190 | - Int32 ip_address, | |
| 3191 | - Int32 port, | |
| 3192 | - String certificate_common_name, | |
| 3193 | - Server -> (SSL_Connection -> One) handler, | |
| 3194 | - Int32 retries, | |
| 3195 | - DenialOfService dos | |
| 3196 | - ) = | |
| 3197 | - if start_ssl_server(ip_address, | |
| 3198 | - port, | |
| 3199 | - certificate_common_name, | |
| 3200 | - handler, | |
| 3201 | - identity) is ok(server) | |
| 3202 | - then print(" \r"); | |
| 3203 | - ok(server) | |
| 3204 | - else print("Port "+port+": retry number "+retries+"\r"); | |
| 3205 | - sleep(1000); | |
| 3206 | - start_https_server(ip_address,port, | |
| 3207 | - certificate_common_name, | |
| 3208 | - handler,retries+1, | |
| 3209 | - dos). | |
| 3210 | - | |
| 3211 | - | |
| 3212 | -public define StartServerResult | |
| 3213 | - start_https_server | |
| 3214 | - ( | |
| 3215 | - Int32 ip_address, | |
| 3216 | - Int32 port, | |
| 3217 | - String certificate_common_name, // of SSL server certificate | |
| 3218 | - List(Web_Site_Description) sites, | |
| 3219 | - DenialOfService dos | |
| 3220 | - ) = | |
| 3221 | - create_directories(sites); | |
| 3222 | - start_https_server(ip_address,port,certificate_common_name, | |
| 3223 | - make_https_handler(sites,dos), | |
| 3224 | - 0,dos). | |
| 3225 | - | |
| 3226 | - | |
| 3227 | - | |
| 3228 | - | |
| 3229 | - | |
| 3230 | - | |
| 3231 | - | |
| 3232 | - | |
| 3233 | - | |
| 3234 | - *** [7] The web dispatcher. | |
| 3235 | - | |
| 3236 | - | |
| 3237 | - *** [7.1] The dispatcher server. | |
| 3238 | - | |
| 3239 | -define One | |
| 3240 | - send_dispatching_page | |
| 3241 | - ( | |
| 3242 | - RWStream conn, | |
| 3243 | - String common_name, | |
| 3244 | - Int32 port | |
| 3245 | - ) = | |
| 3246 | - print("Dispatching '"+common_name+"' to port "+port+"\n"); | |
| 3247 | - forget(reliable_write(conn,to_byte_array( | |
| 3248 | - "<html><head><meta http-equiv=\"Refresh\" content=\"0;URL="+ | |
| 3249 | - "http://"+common_name+":"+port+"/"+ | |
| 3250 | - "\"></head><body></body></html>" | |
| 3251 | - ))). | |
| 3252 | - | |
| 3253 | - | |
| 3254 | - | |
| 3255 | -define Maybe(DispatcherInfo) | |
| 3256 | - find_host | |
| 3257 | - ( | |
| 3258 | - List(DispatcherInfo) l, | |
| 3259 | - String host | |
| 3260 | - ) = | |
| 3261 | - if l is | |
| 3262 | - { | |
| 3263 | - [ ] then failure, | |
| 3264 | - [h . t] then if h is site(name,port) then | |
| 3265 | - if name = host | |
| 3266 | - then success(h) | |
| 3267 | - else find_host(t,host) | |
| 3268 | - }. | |
| 3269 | - | |
| 3270 | - | |
| 3271 | - | |
| 3272 | -define Server -> ((RWStream) -> One) | |
| 3273 | - make_dispatcher_handler | |
| 3274 | - ( | |
| 3275 | - Var(List(DispatcherInfo)) info_v, | |
| 3276 | - DenialOfService dos | |
| 3277 | - ) = | |
| 3278 | - (Server server) |-> (RWStream conn) |-> | |
| 3279 | - with start_time = (Int32)now, | |
| 3280 | - if read_request_line(tcp(conn),start_time+*request_line_delay(dos),dos) is | |
| 3281 | - { | |
| 3282 | - error(msg) then print(format(msg)), | |
| 3283 | - ok(request_line) then | |
| 3284 | - if read_http_headers(tcp(conn),start_time+*headers_delay(dos),dos) is | |
| 3285 | - { | |
| 3286 | - error(msg) then print(format(msg)), | |
| 3287 | - ok(headers) then if get_host_header_value(headers) is | |
| 3288 | - { | |
| 3289 | - failure then print("No 'HOST' HTTP header.\n"), | |
| 3290 | - success(host) then | |
| 3291 | - if find_host(*info_v,host) is | |
| 3292 | - { | |
| 3293 | - failure then print("Host: '"+host+"' not registered.\n"), | |
| 3294 | - success(s) then if s is site(common_name,ip_port) then | |
| 3295 | - send_dispatching_page(conn,common_name,ip_port) | |
| 3296 | - } | |
| 3297 | - } | |
| 3298 | - } | |
| 3299 | - }. | |
| 3300 | - | |
| 3301 | - | |
| 3302 | -define One | |
| 3303 | - dispatcher_update_error | |
| 3304 | - ( | |
| 3305 | - String file_path | |
| 3306 | - ) = | |
| 3307 | - print("web_dispatcher: unable to reread file: '"+file_path+"'.\n"). | |
| 3308 | - | |
| 3309 | - | |
| 3310 | -define Bool | |
| 3311 | - dispatcher_update_data | |
| 3312 | - ( | |
| 3313 | - String info_file_path, | |
| 3314 | - Var(List(DispatcherInfo)) info_v, | |
| 3315 | - Var(Int32) info_date_v | |
| 3316 | - ) = | |
| 3317 | - if directory_full_list(my_anubis_directory+"/web_sites","dispatcher.info","","") is | |
| 3318 | - { | |
| 3319 | - [ ] then false, | |
| 3320 | - [h . t] then if h is | |
| 3321 | - { | |
| 3322 | - no_info(n) then false, | |
| 3323 | - file(n,_,_,d) then if n = "dispatcher.info" | |
| 3324 | - then (info_date_v <- d; | |
| 3325 | - if (RetrieveResult(List(DispatcherInfo)))retrieve(info_file_path) is | |
| 3326 | - { | |
| 3327 | - cannot_find_file then false, | |
| 3328 | - read_error then false, | |
| 3329 | - type_error then false, | |
| 3330 | - ok(info) then info_v <- info; true | |
| 3331 | - }) | |
| 3332 | - else false, | |
| 3333 | - link(_,_,_,_) then false, | |
| 3334 | - directory(_,_,_) then false | |
| 3335 | - } | |
| 3336 | - }. | |
| 3337 | - | |
| 3338 | - | |
| 3339 | - | |
| 3340 | - The loop within which the dispatcher updates its data every 3 seconds: | |
| 3341 | - | |
| 3342 | -define One | |
| 3343 | - dispatcher_update_task | |
| 3344 | - ( | |
| 3345 | - String info_file_path, | |
| 3346 | - Var(List(DispatcherInfo)) info_v, | |
| 3347 | - Var(Int32) info_date_v | |
| 3348 | - ) = | |
| 3349 | - sleep(3000); | |
| 3350 | - (if dispatcher_update_data(info_file_path,info_v,info_date_v) | |
| 3351 | - then unique | |
| 3352 | - else dispatcher_update_error(info_file_path)); | |
| 3353 | - dispatcher_update_task(info_file_path,info_v,info_date_v). | |
| 3354 | - | |
| 3355 | - | |
| 3356 | -public define One | |
| 3357 | - start_web_dispatcher | |
| 3358 | - ( | |
| 3359 | - Int32 ip_address, // address for listening (typically 0: listen on all interfaces) | |
| 3360 | - Int32 http_port, // typically 80 | |
| 3361 | - DenialOfService dos | |
| 3362 | - ) = | |
| 3363 | - with info_file_path = my_anubis_directory+"/web_sites/dispatcher.info", | |
| 3364 | - info_v = var((List(DispatcherInfo))[]), | |
| 3365 | - info_date_v = var((Int32)0), | |
| 3366 | - if dispatcher_update_data(info_file_path,info_v,info_date_v) | |
| 3367 | - then if start_server(ip_address, | |
| 3368 | - http_port, | |
| 3369 | - make_dispatcher_handler(info_v,dos), | |
| 3370 | - (One u)|->u) is | |
| 3371 | - { | |
| 3372 | - cannot_create_the_socket then | |
| 3373 | - print("Cannot create the socket for HTTP server.\n"), | |
| 3374 | - cannot_bind_to_port then | |
| 3375 | - print("Cannot bind HTTP server to port "+http_port+".\n"), | |
| 3376 | - cannot_listen_on_port then | |
| 3377 | - print("HTTP server cannot listen on port "+http_port+".\n"), | |
| 3378 | - ok(http_server) then | |
| 3379 | - dispatcher_update_task(info_file_path,info_v,info_date_v) | |
| 3380 | - } | |
| 3381 | - else dispatcher_update_error(info_file_path). | |
| 3382 | - | |
| 3383 | - | |
| 3384 | - | |
| 3385 | - *** [7.2] The dispatcher web site. | |
| 3386 | - | |
| 3387 | - global define One | |
| 3388 | - web_dispatcher | |
| 3389 | - ( | |
| 3390 | - List(String) args | |
| 3391 | - ) = | |
| 3392 | - start_web_dispatcher(0,80,load_denial_of_service_info). | |
| 3393 | - | |
| 3394 | - | |
| 3395 | - | |
| 3396 | - | |
| 3397 | - | |
| 3398 | - | |
| 3399 | - *** [7.3] Managing the info file. | |
| 3400 | - | |
| 3401 | -define Int32 | |
| 3402 | - register_ip_address | |
| 3403 | - = | |
| 3404 | - if ip_address(prompt(" numerical IP address (for HTTP): ")) is | |
| 3405 | - { | |
| 3406 | - failure then print(" *** Error: incorrect IP address.\n"); | |
| 3407 | - register_ip_address, | |
| 3408 | - success(n) then n | |
| 3409 | - }. | |
| 3410 | - | |
| 3411 | - | |
| 3412 | -define Int32 | |
| 3413 | - register_ip_port | |
| 3414 | - = | |
| 3415 | - if string_to_integer(prompt(" IP port (for HTTP): ")) is | |
| 3416 | - { | |
| 3417 | - failure then print(" *** Error: incorrect IP port.\n"); | |
| 3418 | - register_ip_port, | |
| 3419 | - success(p) then if (0 =< p & p =< 65535) | |
| 3420 | - then p | |
| 3421 | - else print(" *** Error: IP port out of bounds.\n"); | |
| 3422 | - register_ip_port | |
| 3423 | - }. | |
| 3424 | - | |
| 3425 | - | |
| 3426 | -define One | |
| 3427 | - register_new_site | |
| 3428 | - ( | |
| 3429 | - Var(List(DispatcherInfo)) info_v | |
| 3430 | - ) = | |
| 3431 | - print("\n"); | |
| 3432 | - print(" Registering a new site:\n"); | |
| 3433 | - with name = prompt(" Site name: "), | |
| 3434 | - with addr = register_ip_address, | |
| 3435 | - with port = register_ip_port, | |
| 3436 | - (protect info_v <- [site(name,port) . *info_v]); | |
| 3437 | - print(" Site "+name+" at "+ip_addr_to_string(addr)+":"+port+" added\n (but not saved to disk).\n"). | |
| 3438 | - | |
| 3439 | - | |
| 3440 | -define List(DispatcherInfo) | |
| 3441 | - find_sites | |
| 3442 | - ( | |
| 3443 | - List(DispatcherInfo) l, | |
| 3444 | - String name | |
| 3445 | - ) = | |
| 3446 | - if l is | |
| 3447 | - { | |
| 3448 | - [ ] then [ ], | |
| 3449 | - [h . t] then if h is site(n,_) then | |
| 3450 | - if find(name,n,0) is | |
| 3451 | - { | |
| 3452 | - failure then find_sites(t,name), | |
| 3453 | - success(_) then [h . find_sites(t,name)] | |
| 3454 | - } | |
| 3455 | - }. | |
| 3456 | - | |
| 3457 | - | |
| 3458 | -define String | |
| 3459 | - pad | |
| 3460 | - ( | |
| 3461 | - String s, | |
| 3462 | - Int32 l | |
| 3463 | - ) = | |
| 3464 | - if length(s) >= l | |
| 3465 | - then s | |
| 3466 | - else s+constant_string(l-length(s),' '). | |
| 3467 | - | |
| 3468 | - | |
| 3469 | - | |
| 3470 | -define One | |
| 3471 | - show_sites_1 | |
| 3472 | - ( | |
| 3473 | - List(DispatcherInfo) l, | |
| 3474 | - Int32 i | |
| 3475 | - ) = | |
| 3476 | - if l is | |
| 3477 | - { | |
| 3478 | - [ ] then unique, | |
| 3479 | - [h . t] then if h is site(name,port) then | |
| 3480 | - print(" ["+i+"] "+pad(name,40)+" "+" "+port+"\n"); | |
| 3481 | - show_sites_1(t,i+1) | |
| 3482 | - }. | |
| 3483 | - | |
| 3484 | - | |
| 3485 | -define One | |
| 3486 | - show_sites | |
| 3487 | - ( | |
| 3488 | - List(DispatcherInfo) l, | |
| 3489 | - Int32 i | |
| 3490 | - ) = | |
| 3491 | - print(" Name Port\n"); | |
| 3492 | - print(" --------------------------------------------------------\n"); | |
| 3493 | - show_sites_1(l,i). | |
| 3494 | - | |
| 3495 | -define List(DispatcherInfo) | |
| 3496 | - replace_info | |
| 3497 | - ( | |
| 3498 | - List(DispatcherInfo) l, | |
| 3499 | - String site_name, | |
| 3500 | - Int32 new_port | |
| 3501 | - ) = | |
| 3502 | - if l is | |
| 3503 | - { | |
| 3504 | - [ ] then alert, | |
| 3505 | - [h . t] then if h is site(n,_) then | |
| 3506 | - if n = site_name | |
| 3507 | - then [site(n,new_port) . t] | |
| 3508 | - else [h . replace_info(t,site_name,new_port)] | |
| 3509 | - }. | |
| 3510 | - | |
| 3511 | -define List(DispatcherInfo) | |
| 3512 | - delete_info | |
| 3513 | - ( | |
| 3514 | - List(DispatcherInfo) l, | |
| 3515 | - String site_name, | |
| 3516 | - ) = | |
| 3517 | - if l is | |
| 3518 | - { | |
| 3519 | - [ ] then alert, | |
| 3520 | - [h . t] then if h is site(n,_) then | |
| 3521 | - if n = site_name | |
| 3522 | - then t | |
| 3523 | - else [h . delete_info(t,site_name)] | |
| 3524 | - }. | |
| 3525 | - | |
| 3526 | - | |
| 3527 | -define One | |
| 3528 | - update_site | |
| 3529 | - ( | |
| 3530 | - Var(List(DispatcherInfo)) info_v, | |
| 3531 | - String site_name, | |
| 3532 | - Int32 old_port | |
| 3533 | - ) = | |
| 3534 | - print("\n"); | |
| 3535 | - print(" Updating site '"+site_name+"': (currently: "+old_port+")\n"); | |
| 3536 | - with new_port = register_ip_port, | |
| 3537 | - answer = prompt(" Update '"+site_name+"' as: "+new_port+" [Y/N] ? "), | |
| 3538 | - if (answer = "Y" | answer = "y") | |
| 3539 | - then info_v <- replace_info(*info_v,site_name,new_port) | |
| 3540 | - else unique. | |
| 3541 | - | |
| 3542 | - | |
| 3543 | - | |
| 3544 | -define Bool | |
| 3545 | - compare | |
| 3546 | - ( | |
| 3547 | - DispatcherInfo d1, | |
| 3548 | - DispatcherInfo d2 | |
| 3549 | - ) = | |
| 3550 | - if d1 is site(n1,_) then | |
| 3551 | - if d2 is site(n2,_) then | |
| 3552 | - string_less(n1,n2). | |
| 3553 | - | |
| 3554 | - | |
| 3555 | - | |
| 3556 | -define One | |
| 3557 | - update_site | |
| 3558 | - ( | |
| 3559 | - Var(List(DispatcherInfo)) info_v | |
| 3560 | - ) = | |
| 3561 | - print("\n"); | |
| 3562 | - with prefix = prompt(" Search for site to update: "), | |
| 3563 | - if find_sites(*info_v,prefix) is | |
| 3564 | - { | |
| 3565 | - [ ] then print(" No site found.\n"); | |
| 3566 | - update_site(info_v), | |
| 3567 | - [h . t] then | |
| 3568 | - show_sites(qsort([h . t],compare),1); | |
| 3569 | - with i1 = prompt(" Choose a site to update [1/.../"+(length(t)+1)+"]: "), | |
| 3570 | - if string_to_integer(i1) is | |
| 3571 | - { | |
| 3572 | - failure then print(" *** Error: site number not recognized.\n"); | |
| 3573 | - update_site(info_v), | |
| 3574 | - success(ii1) then if nth(ii1-1,*info_v) is | |
| 3575 | - { | |
| 3576 | - failure then print(" *** Error: site number "+i1+" does not exist.\n"); | |
| 3577 | - update_site(info_v), | |
| 3578 | - success(site_info) then if site_info is site(name,old_port) then | |
| 3579 | - update_site(info_v,name,old_port) | |
| 3580 | - } | |
| 3581 | - } | |
| 3582 | - }. | |
| 3583 | - | |
| 3584 | - | |
| 3585 | -define One | |
| 3586 | - delete_site | |
| 3587 | - ( | |
| 3588 | - Var(List(DispatcherInfo)) info_v, | |
| 3589 | - String site_name, | |
| 3590 | - Int32 old_port | |
| 3591 | - ) = | |
| 3592 | - print("\n"); | |
| 3593 | - print(" Deleting site '"+site_name+"': (currently: "+old_port+")\n"); | |
| 3594 | - with answer = prompt(" Are you sure you want to delete site: '"+site_name+"' [Y/N] ? "), | |
| 3595 | - if (answer = "Y" | answer = "y") | |
| 3596 | - then info_v <- delete_info(*info_v,site_name) | |
| 3597 | - else print(" Site '"+site_name+"' not deleted.\n"). | |
| 3598 | - | |
| 3599 | - | |
| 3600 | -define One | |
| 3601 | - delete_site | |
| 3602 | - ( | |
| 3603 | - Var(List(DispatcherInfo)) info_v | |
| 3604 | - ) = | |
| 3605 | - print("\n"); | |
| 3606 | - with prefix = prompt(" Search for site to delete: "), | |
| 3607 | - if find_sites(*info_v,prefix) is | |
| 3608 | - { | |
| 3609 | - [ ] then print(" No site found.\n"); | |
| 3610 | - delete_site(info_v), | |
| 3611 | - [h . t] then | |
| 3612 | - show_sites(qsort([h . t],compare),1); | |
| 3613 | - with i1 = prompt(" Choose a site to delete [1/.../"+(length(t)+1)+"]: "), | |
| 3614 | - if string_to_integer(i1) is | |
| 3615 | - { | |
| 3616 | - failure then print(" *** Error: site number not recognized.\n"); | |
| 3617 | - delete_site(info_v), | |
| 3618 | - success(ii1) then if nth(ii1-1,*info_v) is | |
| 3619 | - { | |
| 3620 | - failure then print(" *** Error: site number "+i1+" does not exist.\n"); | |
| 3621 | - delete_site(info_v), | |
| 3622 | - success(site_info) then if site_info is site(name,old_port) then | |
| 3623 | - delete_site(info_v,name,old_port) | |
| 3624 | - } | |
| 3625 | - } | |
| 3626 | - }. | |
| 3627 | - | |
| 3628 | - | |
| 3629 | -define One | |
| 3630 | - manager | |
| 3631 | - ( | |
| 3632 | - Var(List(DispatcherInfo)) info_v, | |
| 3633 | - String file_path | |
| 3634 | - ) = | |
| 3635 | - print("\n"); | |
| 3636 | - print(" --- Welcome to the Web Dispatcher Manager ---\n"); | |
| 3637 | - with l = length(*info_v), | |
| 3638 | - print(" "+l+" site"+(if l>1 then "s" else "")+" currently registred.\n"); | |
| 3639 | - print(" [L] List registered sites.\n"); | |
| 3640 | - print(" [R] Register a new site.\n"); | |
| 3641 | - print(" [U] Update a registred site.\n"); | |
| 3642 | - print(" [D] Delete a registred site.\n"); | |
| 3643 | - with propose_write_v = var((Bool)true), | |
| 3644 | - action = prompt(" Choose an action [L/R/U/D]: "), | |
| 3645 | - (if (action = "L" | action = "l") then (show_sites(*info_v,1); propose_write_v <- false) else | |
| 3646 | - if (action = "R" | action = "r") then register_new_site(info_v) else | |
| 3647 | - if (action = "U" | action = "u") then update_site(info_v) else | |
| 3648 | - if (action = "D" | action = "d") then delete_site(info_v) else | |
| 3649 | - print("Action not recognized.\n")); | |
| 3650 | - print("\n"); | |
| 3651 | - if *propose_write_v then | |
| 3652 | - with result = prompt(" Write modifications to data base [Y/N] ?"), | |
| 3653 | - if (result = "Y" | result = "y") | |
| 3654 | - then if save(*info_v,file_path) is | |
| 3655 | - { | |
| 3656 | - cannot_open_file then print(" File '"+file_path+"' not found.\n"), | |
| 3657 | - write_error then print(" Error while writing file '"+file_path+"'.\n"), | |
| 3658 | - ok then print(" Data base has been modified.\n") | |
| 3659 | - } | |
| 3660 | - else print(" Data base not modified.\n") | |
| 3661 | - else unique. | |
| 3662 | - | |
| 3663 | - | |
| 3664 | - | |
| 3665 | -global define One | |
| 3666 | - manage_web_dispatcher | |
| 3667 | - ( | |
| 3668 | - List(String) args | |
| 3669 | - ) = | |
| 3670 | - with info_v = var((List(DispatcherInfo))[]), | |
| 3671 | - with file_path = my_anubis_directory+"/web_sites/dispatcher.info", | |
| 3672 | - if (RetrieveResult(List(DispatcherInfo)))retrieve(file_path) is | |
| 3673 | - { | |
| 3674 | - cannot_find_file then print("File '"+file_path+"' does not exist.\n"); | |
| 3675 | - with answer = prompt("Create it [Y/N] ? "), | |
| 3676 | - if (answer = "Y" | answer = "y") | |
| 3677 | - then if save((List(DispatcherInfo))[],file_path) is | |
| 3678 | - { | |
| 3679 | - cannot_open_file then | |
| 3680 | - print("Cannot create file '"+file_path+"'.\n"), | |
| 3681 | - write_error then | |
| 3682 | - print("Error while creating file '"+file_path+"'.\n"), | |
| 3683 | - ok then manager(info_v,file_path) | |
| 3684 | - } | |
| 3685 | - else unique, | |
| 3686 | - read_error then print("Error while reading file '"+file_path+"'.\n"), | |
| 3687 | - type_error then print("File '"+file_path+"' is corrupted.\n"), | |
| 3688 | - ok(info) then info_v <- info; | |
| 3689 | - manager(info_v,file_path) | |
| 3690 | - }. | |
| 3691 | - | |
| 3692 | - | |
| 3693 | - | |
| 3694 | - | |
| 3695 | - | |
| 1 | + | |
| 2 | + *Project* The Anubis Project | |
| 3 | + | |
| 4 | + *Title* A Multi Host HTTP/HTTPS Server | |
| 5 | + | |
| 6 | + *Copyright* Copyright (c) Anubis Team 2003-2007. | |
| 7 | + | |
| 8 | + | |
| 9 | + *Authors* Alain Prouté | |
| 10 | + David René | |
| 11 | + Cédric Ricard | |
| 12 | + | |
| 13 | + | |
| 14 | + *Revised* July 2007. | |
| 15 | + | |
| 16 | + | |
| 17 | + | |
| 18 | + *Overviews* | |
| 19 | + In this file a HTTP/HTTPS server is defined, which is able to handle multiple hosts | |
| 20 | + (virtual hosts). It answers HTTP/HTTPS requests, sends files (images or any other kind | |
| 21 | + of file), constructs HTML pages on the fly using informations received from the client | |
| 22 | + (when the URI ends by '.awp'), handles uploading of files and redirections. It is | |
| 23 | + multitasking by itself, and can handle any number of sites and clients simultaneously. | |
| 24 | + It should better be used in conjunction with 'making_a_web_site.anubis' to be found in | |
| 25 | + the same directory. If you use 'web/making_a_web_site.anubis', you don't need to read | |
| 26 | + this file. | |
| 27 | + | |
| 28 | + | |
| 29 | + ----------------------------------- Table of Contents --------------------------------- | |
| 30 | + | |
| 31 | + *** (1) Multihosting and redirections. | |
| 32 | + *** (2) The incompatibility between SSL and virtual hosts. | |
| 33 | + *** (3) HTTP headers and web arguments. | |
| 34 | + *** (4) Site descriptions. | |
| 35 | + *** (5) Protection against denial of service attacks. | |
| 36 | + *** (6) Starting your HTTP and HTTPS servers. | |
| 37 | + *** (7) Private download. | |
| 38 | + *** (8) About web argument names. | |
| 39 | + *** (9) A web dispatcher. | |
| 40 | + | |
| 41 | + --------------------------------------------------------------------------------------- | |
| 42 | + | |
| 43 | + | |
| 44 | + | |
| 45 | + | |
| 46 | + *** (1) Multihosting and redirections. | |
| 47 | + | |
| 48 | + This HTTP/HTTPS server can handle several host (also called 'virtual hosts'), in other | |
| 49 | + words, you may have several sites on the same server, with the same IP address and same | |
| 50 | + port numbers, but distinct 'host names'. | |
| 51 | + | |
| 52 | + A HTTP request sent by a browser contains the following informations: | |
| 53 | + | |
| 54 | + - a 'host name', | |
| 55 | + - an URI (Uniform Resource Identifier), | |
| 56 | + - HTTP headers, | |
| 57 | + - web arguments (in the form 'name=value'). | |
| 58 | + | |
| 59 | + Actually, the host name is just the value of the HTTP header whose name is 'Host'. The | |
| 60 | + host name indicates which site is requested. Hence, it is the primary information for | |
| 61 | + branching to the right site. If there is no 'Host' HTTP header in the request, the | |
| 62 | + request is denied. | |
| 63 | + | |
| 64 | + From now on, we may assume that the host is determined, and consequently that we are | |
| 65 | + concerned by only one site. Each site has his own directories on the server's | |
| 66 | + disk. | |
| 67 | + | |
| 68 | + Each site also has a list of 'redirections'. A redirection is a triplet, like this one: | |
| 69 | + | |
| 70 | + redirect("/", "www.our-business.com", "/homepage.awp") | |
| 71 | + | |
| 72 | + meaning that if the host is "www.our-business.com", and if the requested URI is "/", | |
| 73 | + then the URI to be served is "/homepage.awp". 'redirect' is a constructor of the type | |
| 74 | + 'Redirection' defined in 'web/common.anubis'. | |
| 75 | + | |
| 76 | + Now, an URI may end by ".awp" (meaning 'Anubis Web Page') or not. If it does, the | |
| 77 | + server understands that an HTML page must be constructed on the fly, and to that end it | |
| 78 | + calls the 'awp handler' of the site. Otherwise, the URI must end by a known extension, | |
| 79 | + like ".jpg", ".png", ".txt", etc... and represents a file path relative to the | |
| 80 | + 'public' directory of the site. If these conditions are satisfied, the file is sent to | |
| 81 | + the client. Known extensions are recorded in 'web/mime.anubis'. | |
| 82 | + | |
| 83 | + | |
| 84 | + | |
| 85 | + | |
| 86 | + *** (2) The incompatibility between SSL and virtual hosts. | |
| 87 | + | |
| 88 | + Handling virtual hosts makes a problem under SSL (i.e. when using HTTPS), which is due | |
| 89 | + to the fact that the guys at Netscape who designed SSL probably did not have the | |
| 90 | + question of virtual hosts in mind. Indeed, the SSL handshake is completed before the | |
| 91 | + server can know about the value of the 'Host' HTTP header, so that it cannot know which | |
| 92 | + server certificate must be sent to the client. This makes a problem, because the | |
| 93 | + browser will not accept a certificate whose common name does not correspond to the name | |
| 94 | + of the requested host. The user will have to accept the certificate manually, which is | |
| 95 | + not good for the security image of the site. This problem has at least two solutions | |
| 96 | + (as far as Anubis is concerned). | |
| 97 | + | |
| 98 | + Solution 1. Arrange so that the network interface on which the server is listening | |
| 99 | + has at least as many different IP addresses as you have virtual hosts. Such | |
| 100 | + supplementary IP addresses are called 'IP Aliases'. In this case, start one HTTPS | |
| 101 | + server for each virtual host, each one listening on a different address. For the time | |
| 102 | + being, this method is applicable under Anubis only if you start as many instances of | |
| 103 | + 'anbexec' as you have virtual hosts, because each instance of 'anbexec' can handle only | |
| 104 | + one server certificate. Of course, getting IP aliases is another problem to be solved | |
| 105 | + with your Internet provider. | |
| 106 | + | |
| 107 | + Solution 2. We propose a simple solution, using only one server certificate (hence | |
| 108 | + only one instance of 'anbexec'). Since, we have only one server certificate, we must | |
| 109 | + introduce a notion of 'main host', i.e. a host containing all other 'virtual | |
| 110 | + hosts'. The unique server certificate belong to the main host, so that only the main | |
| 111 | + host is identified by the client. The client must trust the main host and be confident | |
| 112 | + that the main host redirects him to the right virtual host. Actually, the process will | |
| 113 | + be transparent to the client, except that the client will see the name of the main host | |
| 114 | + instead of the name of the virtual host in the 'location' field of the browser. | |
| 115 | + | |
| 116 | + So, assume that the name of main host is 'www.securedhost.com', and that the names of | |
| 117 | + the virtual hosts are: | |
| 118 | + | |
| 119 | + actual name simplified name | |
| 120 | + ----------------------------------------------------- | |
| 121 | + www.virtual1.com virtual1 | |
| 122 | + www.virtual2.com virtual2 | |
| 123 | + www.virtual3.com virtual3 | |
| 124 | + | |
| 125 | + Then the (confidential) document '/doc/my_document.pdf' on 'www.virtual2.com' will have | |
| 126 | + the URL: | |
| 127 | + | |
| 128 | + https://www.securedhost.com/virtual2/doc/my_document.pdf | |
| 129 | + | |
| 130 | + In order to work transparently, this solution must combine HTTP and HTTPS. Indeed, the | |
| 131 | + vitual host must have a first page reachable under HTTP, through the URL: | |
| 132 | + | |
| 133 | + http://www.virtual2.com/ | |
| 134 | + | |
| 135 | + The HTTP server will redirect this URL to the awp handler of virtual host 'virtual2'. | |
| 136 | + The handler of this virtual host is able to generate a first page containing the | |
| 137 | + following HTML meta: | |
| 138 | + | |
| 139 | + <meta http-equiv="Refresh" content="0;URL=https://www.securedhost.com/virtual2/">, | |
| 140 | + | |
| 141 | + so that the client is immediately redirected to the main host under HTTPS (hence | |
| 142 | + accepting tranparently the server certificate). The awp handler of 'virtual2' then | |
| 143 | + redirects this URL to the home page (maybe a login page) of 'virtual2'. | |
| 144 | + | |
| 145 | + See 'web/making_a_web_site.anubis' for the sequel of this story. | |
| 146 | + | |
| 147 | + | |
| 148 | + | |
| 149 | + | |
| 150 | + | |
| 151 | + *** (3) HTTP headers and web arguments. | |
| 152 | + | |
| 153 | + Each HTTP request which arrives on the server contains a request line followed by a | |
| 154 | + series of HTTP headers. Each HTTP header is a pair '(name,value)' assigning a value to | |
| 155 | + a name. The type 'HTTP_header' is defined in 'web/common.anubis'. | |
| 156 | + | |
| 157 | + The request may also have a 'body'. The body contains either 'web arguments' or | |
| 158 | + uploaded files (or both). The request line itself may also contain web arguments (in a | |
| 159 | + so-called 'query string'). Like HTTP headers, 'web arguments' are pairs | |
| 160 | + '(name,value)', but the difference is that these pairs are generated by the page within | |
| 161 | + which the client clicks, while HTTP headers are generated by the browser itself. The | |
| 162 | + type 'Web_arg' is defined in 'web/common.anubis'. It has two alternatives, one for | |
| 163 | + ordinary web arguments (pairs) and one for uploaded files. | |
| 164 | + | |
| 165 | +read CXM_common.anubis | |
| 166 | +read tools/basis.anubis | |
| 167 | +read system/string.anubis | |
| 168 | +read CXM_mime.anubis | |
| 169 | + | |
| 170 | + | |
| 171 | + | |
| 172 | + *** (4) Site descriptions. | |
| 173 | + | |
| 174 | + The type HTTP_Info gathers informations comming along with the client's request. These | |
| 175 | + informations are rarely used for composing HTML pages. Nevertheless, they are at your | |
| 176 | + disposal. | |
| 177 | + | |
| 178 | +public type HTTP_Info: | |
| 179 | + http_info | |
| 180 | + ( | |
| 181 | + Int32 ip_address, // IP address of the client | |
| 182 | + String uri, // URI requested by the client | |
| 183 | + List(HTTP_header) http_headers, // HTTP headers sent by the client | |
| 184 | + One -> String generate_trust_ticket // may be used against denial of | |
| 185 | + // service attacks | |
| 186 | + ). | |
| 187 | + | |
| 188 | + | |
| 189 | + | |
| 190 | + Each site is described by a 'web site description', which is a datum of type | |
| 191 | + 'Web_Site_Description'. | |
| 192 | + | |
| 193 | +public type Web_Site_Description: | |
| 194 | + web_site_description( | |
| 195 | + List(String) common_names, | |
| 196 | + String site_directory, | |
| 197 | + List(Redirection) redirections, | |
| 198 | + String charset, | |
| 199 | + List(String) journal_extensions, | |
| 200 | + List(String) journal_headers, | |
| 201 | + String authorization_secret, | |
| 202 | + List(MIME) known_mime_types, | |
| 203 | + (String host_name, | |
| 204 | + HTTP_Info http_info, | |
| 205 | + List(Web_arg) lwa, | |
| 206 | + Bool is_https) -> (List(HTTP_header), | |
| 207 | + Printable_tree) awp_handler, | |
| 208 | + (List(Web_arg) lwa) -> One before_send_file). | |
| 209 | + | |
| 210 | + The component 'common_names' is the list of names of the site, like for example | |
| 211 | + "www.our-business.com". The reason why we have a list of common names instead of a | |
| 212 | + single common name, is that it may be useful to have a common name like "192.168.0.1" | |
| 213 | + for testing. | |
| 214 | + | |
| 215 | + 'charset' is a string which will determine the character encoding to be used by the | |
| 216 | + browser. Typically, this string is one of: "UTF-8", "ISO-8859-1", "Windows-1252", | |
| 217 | + etc... | |
| 218 | + | |
| 219 | + 'journal_extensions' is the list of URI extensions for which you want a log in the | |
| 220 | + journal (and on the console). When a request arrives, and if the extension is a member | |
| 221 | + of this list, a message is printed into the journal of the site including the date, the | |
| 222 | + IP address of the client, the complete HTTP request line. The HTTP headers whose name | |
| 223 | + is a member of 'journal_headers' are also printed in the journal. A reasonable minimum | |
| 224 | + for these two components is: | |
| 225 | + | |
| 226 | + [".awp"] for journal_extensions | |
| 227 | + ["user-agent"] for journal_headers | |
| 228 | + | |
| 229 | + 'authorization_secret' is a string which should just be unguessable. You may choose | |
| 230 | + something like (but don't choose this one !): | |
| 231 | + | |
| 232 | + "Hg8kJe42gCML9jNH-74" | |
| 233 | + | |
| 234 | + i.e. a sequence of characters typed at random, long enough to be unguessable. This is | |
| 235 | + used by the 'private download' mecanism, which is discussed later in this file. | |
| 236 | + | |
| 237 | + The component 'awp_handler' is a function of type: | |
| 238 | + | |
| 239 | + (String host_name, | |
| 240 | + HTTP_Info http_info, | |
| 241 | + List(Web_arg) web_args, | |
| 242 | + Bool is_https) -> Printable_tree | |
| 243 | + | |
| 244 | + ('Printable_tree' is a substitute for 'String' and is defined in | |
| 245 | + 'tools/basis.anubis'). This function is the 'awp handler' for the site. When the URI | |
| 246 | + ends by ".awp", this function is called, and the result (an HTML page) is sent to the | |
| 247 | + client over the connection. The last operand to this function is a boolean which is | |
| 248 | + 'true' when the requests arrives through the HTTPS channel, and 'false' when it arrives | |
| 249 | + through the HTTP channel. | |
| 250 | + | |
| 251 | + | |
| 252 | + | |
| 253 | + | |
| 254 | + | |
| 255 | + | |
| 256 | + | |
| 257 | + *** (5) Protection against denial of service attacks. | |
| 258 | + | |
| 259 | + We need to protect our servers against 'denial of service' attacks. The attack may be | |
| 260 | + send automatically from machines which are infested by viruses. In that case, our | |
| 261 | + server is saturated of connections (all virtual machines at work), but nothing is | |
| 262 | + comming on the connections. In order to avoid this problem, we propose the following: | |
| 263 | + | |
| 264 | + (1) Limit the number of simultaneous connections (say to 100). | |
| 265 | + (2) Close a connection if the request is not complete after say 10 seconds. | |
| 266 | + (3) Close the connection if the request is bigger than a given size (normal requests | |
| 267 | + are small except when there are uploaded files. | |
| 268 | + (4) Close the connection during the sending of the answer if the client is waiting | |
| 269 | + too much. | |
| 270 | + (5) Record all IP addresses with which we have encountered one of the problems above. | |
| 271 | + (6) Immediately close the connections if the IP address is in our list. | |
| 272 | + (7) Remove an address from the list only after 5 minutes of inactivity of this | |
| 273 | + address. | |
| 274 | + (8) Maintain a list of reliable IP addresses. | |
| 275 | + | |
| 276 | + Of course, all the above are approximative solutions which may in some circumstances | |
| 277 | + become either cumbersome or also partially block the system. So, it is needed to have a | |
| 278 | + set of dynamically modifiable parameters in order to master the behavior of this | |
| 279 | + mecanism. | |
| 280 | + | |
| 281 | + | |
| 282 | + Each dubious IP address is recorded together with its last activity time. | |
| 283 | + | |
| 284 | +public type DubiousIP: | |
| 285 | + dubious_ip (Int32 address, | |
| 286 | + Int32 last_activity). | |
| 287 | + | |
| 288 | + | |
| 289 | +public type DenialOfService: | |
| 290 | + denial_of_service(Var(Int32) max_connections, | |
| 291 | + Var(Int32) request_line_delay, // seconds | |
| 292 | + Var(Int32) headers_delay, | |
| 293 | + Var(Int32) answer_delay, | |
| 294 | + Var(List(DubiousIP)) list_of_dubious, | |
| 295 | + Var(List(Int32)) reliable_addresses). | |
| 296 | + | |
| 297 | + The informations in this set of variables are stored serialized into the file | |
| 298 | + 'my_anubis/web_sites/dos_info'. If this file does not exist a set if variables with | |
| 299 | + default values is created. The values are saved on the disk each time they are | |
| 300 | + modified. | |
| 301 | + | |
| 302 | +public define DenialOfService load_denial_of_service_info. | |
| 303 | + | |
| 304 | + | |
| 305 | + | |
| 306 | + *** (6) Starting your HTTP and HTTPS servers. | |
| 307 | + | |
| 308 | + When your web site descriptions are ready, you can start a pair of servers (a HTTP | |
| 309 | + server and a HTTPS server) for serving your web sites. Notice that there are always | |
| 310 | + two servers, regardless of the number of web sites, and that each web sites normally | |
| 311 | + uses the two servers. | |
| 312 | + | |
| 313 | + | |
| 314 | +public define StartServerResult | |
| 315 | + start_http_server | |
| 316 | + ( | |
| 317 | + Int32 ip_address, | |
| 318 | + Int32 http_port, | |
| 319 | + List(Web_Site_Description) web_sites, | |
| 320 | + DenialOfService dos | |
| 321 | + ). | |
| 322 | + | |
| 323 | +public define StartServerResult | |
| 324 | + start_https_server | |
| 325 | + ( | |
| 326 | + Int32 ip_address, | |
| 327 | + Int32 https_port, | |
| 328 | + String certificate_common_name, | |
| 329 | + List(Web_Site_Description) web_sites, | |
| 330 | + DenialOfService dos | |
| 331 | + ). | |
| 332 | + | |
| 333 | + The first argument 'ip_address' is the IP address on which the servers listen. If you | |
| 334 | + put 0, the servers listen on all adresses of the machine (which is useful if the | |
| 335 | + machine has several network interfaces). Otherwise, use the function 'ip_address' | |
| 336 | + defined in 'tools/basis.anubis' for composing a particular IP address. | |
| 337 | + | |
| 338 | + The next arguments are the port numbers for HTTP and HTTPS. The usual values are 80 and | |
| 339 | + 443, but you may have reasons to choose other values. | |
| 340 | + | |
| 341 | + The next argument is the list of your web site descriptions. All the sites described in | |
| 342 | + this list will be accessible on the server. | |
| 343 | + | |
| 344 | + The argument 'dos' is a set of dynamic variables containing the informations for | |
| 345 | + protecting the servers against denial of service attacks. | |
| 346 | + | |
| 347 | + | |
| 348 | + | |
| 349 | + | |
| 350 | + | |
| 351 | + | |
| 352 | + *** (7) Private download. | |
| 353 | + | |
| 354 | + It may happen that you want to propose private files for download. This means that such | |
| 355 | + a file could be downloaded only by the authorized person, and should not be seen by any | |
| 356 | + other one. This feature can be used only under HTTPS, not under HTTP. | |
| 357 | + | |
| 358 | + The file may be located anywhere on the server. Hence, the file has a complete absolute | |
| 359 | + path, like for example: | |
| 360 | + | |
| 361 | + /home/georges/my_documents/my_text.pdf | |
| 362 | + | |
| 363 | + which has nothing to do with the directories of the web server. Now, you may also want | |
| 364 | + to show another path or simply just a name to the client, not the actual absolute path | |
| 365 | + above, which may need to remain secret. So for example, the same file may appear to the | |
| 366 | + client as: | |
| 367 | + | |
| 368 | + informations.pdf | |
| 369 | + | |
| 370 | + The page must provide a link with an authorization. The authorization is just a web | |
| 371 | + argument, whose name is "zauth". The value of this web argument is computed by hashing | |
| 372 | + some secret string (known only from the programmer of the web site) with the absolute | |
| 373 | + path of the file. The HTTPS request will have the form: | |
| 374 | + | |
| 375 | + GET /informations.pdf?zauth=d38161f5b4e87e2d46e06ff8b3e233be563794d1 | |
| 376 | + | |
| 377 | + The server will search for a file named | |
| 378 | + | |
| 379 | + zd38161f5b4e87e2d46e06ff8b3e233be563794d1 | |
| 380 | + | |
| 381 | + (i.e. "z" concatenated with the value of the authorization) in the subdirectory | |
| 382 | + 'private_download' of the site directory. This file contains the absolute path of the | |
| 383 | + file, i.e: | |
| 384 | + | |
| 385 | + /home/georges/my_documents/my_text.pdf | |
| 386 | + | |
| 387 | + At that point, the server may hash the secret string and the absolute path together, to | |
| 388 | + check if the client is authorized to download the file. If it is the case, it sends the | |
| 389 | + file (the MIME type is declared as 'application/octet-stream' if it is not recognized). | |
| 390 | + The file is sent under the visible name. | |
| 391 | + | |
| 392 | + The server creates automatically the subdirectory 'private_download/' within the 'site | |
| 393 | + directory' (for each web site) if it does not already exist. Files in this directory | |
| 394 | + are deleted when they become too old (for example, after 3 days of life). | |
| 395 | + | |
| 396 | + Here is the function for computing the value of the authorization, and for making the | |
| 397 | + authorization file in 'private_download'. | |
| 398 | + | |
| 399 | +public define String | |
| 400 | + make_authorization | |
| 401 | + ( | |
| 402 | + String site_directory, | |
| 403 | + String authorization_secret, // known only by the programmer of the web site | |
| 404 | + String absolute_path // on server | |
| 405 | + ). | |
| 406 | + | |
| 407 | + See 'web/making_a_web_site.anubis' for the construction of the link for downloading. | |
| 408 | + | |
| 409 | + | |
| 410 | + | |
| 411 | + | |
| 412 | + | |
| 413 | + | |
| 414 | + | |
| 415 | + | |
| 416 | + *** (8) About web argument names. | |
| 417 | + | |
| 418 | + The server reserves the name "zauth" for the authorization in the private download | |
| 419 | + mecanism. Also, if the name of a web arguments begins by "p" (like 'password'), it does | |
| 420 | + not print the value of the web argument neither on the console or in the journal. A | |
| 421 | + good politics is to prefix all web arguments by letters distinct from 'p' and 'z'. This | |
| 422 | + method is used in 'web/making_a_web_site.anubis'. This will avoid clashes of names. | |
| 423 | + | |
| 424 | + | |
| 425 | + | |
| 426 | + | |
| 427 | + | |
| 428 | + | |
| 429 | + *** (9) A web dispatcher. | |
| 430 | + | |
| 431 | + For hosting several sites you may prefer another method which we now describe. We start | |
| 432 | + a HTTP server on port 80 (or on another port). This server is called the | |
| 433 | + ``dispatcher''. When a requests arrives, the dispatcher examines the ``host'' HTTP | |
| 434 | + header, so that it gets the name of the requested host. Then it sends to the client a | |
| 435 | + page like this one: | |
| 436 | + | |
| 437 | + <html> | |
| 438 | + <head> | |
| 439 | + <meta http-equiv="Refresh" content="0;URL=..."> | |
| 440 | + </head> | |
| 441 | + <body> | |
| 442 | + </body> | |
| 443 | + </html> | |
| 444 | + | |
| 445 | + where the URL represented by '...' is the URL of the requested site. This URL may have | |
| 446 | + the same IP address as the dispatcher, except that the port number is different. It may | |
| 447 | + also have a different IP address. | |
| 448 | + | |
| 449 | + The dispatcher uses the file 'my_anubis/web_sites/dispatcher.info'. This file contains | |
| 450 | + a serialized datum of type 'List(DispatcherInfo)'. | |
| 451 | + | |
| 452 | +public type DispatcherInfo: | |
| 453 | + site(String common_name, | |
| 454 | + Int32 http_port). | |
| 455 | + | |
| 456 | + The dispatcher does not write into this file. It reads it when it starts, and rereads | |
| 457 | + it each time the date of last modification of the file changes, so that the dispatcher | |
| 458 | + always has up to date data. The file may be managed (written and updated) by another | |
| 459 | + program. | |
| 460 | + | |
| 461 | + So, for each site, the dispatcher knows the common name (needed to recognize the 'host' | |
| 462 | + HTTP header), and the pair (ip_address,port) used by the actual site for HTTP. The | |
| 463 | + dispatcher does not worry about HTTPS. HTTPS must be managed by the actual site. | |
| 464 | + | |
| 465 | + The dispatcher is started by: | |
| 466 | + | |
| 467 | +public define One | |
| 468 | + start_web_dispatcher | |
| 469 | + ( | |
| 470 | + Int32 ip_address, // address for listening (typically 0) | |
| 471 | + Int32 port, // typically 80 | |
| 472 | + DenialOfService dos | |
| 473 | + ). | |
| 474 | + | |
| 475 | + A command line tool for managing the file 'my_anubis/web_sites/dispatcher.info' is also | |
| 476 | + provided: | |
| 477 | + | |
| 478 | + global define One | |
| 479 | + manage_web_dispatcher | |
| 480 | + ( | |
| 481 | + List(String) args | |
| 482 | + ). | |
| 483 | + | |
| 484 | + | |
| 485 | + | |
| 486 | + | |
| 487 | + | |
| 488 | + | |
| 489 | + | |
| 490 | + --- That's all for the public part ! -------------------------------------------------- | |
| 491 | + | |
| 492 | + | |
| 493 | + | |
| 494 | + | |
| 495 | + | |
| 496 | + | |
| 497 | + | |
| 498 | + ----------------------------------- Table of Contents --------------------------------- | |
| 499 | + | |
| 500 | + *** [1] Types which are private to this file. | |
| 501 | + | |
| 502 | + *** [2] Tools. | |
| 503 | + *** [2.1] Formating an error message. | |
| 504 | + *** [2.2] Converting IP addresses. | |
| 505 | + *** [2.3] Reading and unputting characters. | |
| 506 | + *** [2.4] Reading and discarding characters. | |
| 507 | + *** [2.5] Reading a character string. | |
| 508 | + *** [2.6] Padding integers with zeros. | |
| 509 | + *** [2.7] Converting web arguments to ASCII. | |
| 510 | + *** [2.8] Server description. | |
| 511 | + | |
| 512 | + *** [3] Managing the journal. | |
| 513 | + *** [3.1] Naming journal files. | |
| 514 | + *** [3.2] Formating HTTP headers. | |
| 515 | + *** [3.3] Formating web arguments. | |
| 516 | + *** [3.4] Formating the whole request. | |
| 517 | + *** [3.5] Putting it in the journal file (and on the console). | |
| 518 | + | |
| 519 | + *** [4] Reading the HTTP request. | |
| 520 | + *** [4.1] Skipping leading blanks. | |
| 521 | + *** [4.2] Reading a new line. | |
| 522 | + *** [4.3] Reading a 'word'. | |
| 523 | + *** [4.4] Separating the URI from the query string. | |
| 524 | + *** [4.5] Reading the web arguments. | |
| 525 | + *** [4.7] Reading the request line. | |
| 526 | + *** [4.8] Reading the HTTP headers. | |
| 527 | + *** [4.9] Getting the size of the request's body. | |
| 528 | + *** [4.10] Reading the body of the request. | |
| 529 | + | |
| 530 | + *** [5] Making the HTTP answer. | |
| 531 | + *** [5.1] Avoiding illegal URIs. | |
| 532 | + *** [5.2] Managing authorizations for downloading private files. | |
| 533 | + *** [5.3] Recognizing MIME types. | |
| 534 | + *** [5.4] Formating HTTP headers. | |
| 535 | + *** [5.5] Sending a file. | |
| 536 | + *** [5.6] Answering a www-url encoded request. | |
| 537 | + *** [5.7] Answering a multipart/form-data encoded request. | |
| 538 | + *** [5.7.1] Finding the boundary. | |
| 539 | + *** [5.7.2] Reading attributes from a multipart entity. | |
| 540 | + *** [5.7.3] Creating a temporary filename for an uploaded file. | |
| 541 | + *** [5.7.4] Saving an uploaded file under a temporary filename. | |
| 542 | + *** [5.7.5] Removing the path from a file name. | |
| 543 | + *** [5.7.6] Reading a multipart entity. | |
| 544 | + *** [5.8] Handling redirections. | |
| 545 | + *** [5.9] Answering both sorts of requests. | |
| 546 | + | |
| 547 | + *** [6] The HTTP/HTTPS servers. | |
| 548 | + *** [6.1] The HTTP request handler. | |
| 549 | + *** [6.2] Server's tasks. | |
| 550 | + *** [6.3] Starting the HTTP/HTTPS servers. | |
| 551 | + | |
| 552 | + *** [7] The web dispatcher. | |
| 553 | + *** [7.1] The dispatcher server. | |
| 554 | + *** [7.2] The dispatcher web site. | |
| 555 | + *** [7.3] Managing the info file. | |
| 556 | + | |
| 557 | + --------------------------------------------------------------------------------------- | |
| 558 | + | |
| 559 | + | |
| 560 | + | |
| 561 | + | |
| 562 | +read tools/basis.anubis | |
| 563 | +read tools/findstring.anubis | |
| 564 | +read tools/connections.anubis | |
| 565 | + | |
| 566 | + | |
| 567 | + | |
| 568 | + | |
| 569 | + | |
| 570 | + *** [1] Types which are private to this file. | |
| 571 | + | |
| 572 | + We use the following self-explanatory types. | |
| 573 | + | |
| 574 | +type Error: | |
| 575 | + cannot_read_from_connection, | |
| 576 | + not_get_or_post_request(String), | |
| 577 | + end_of_line_expected, | |
| 578 | + incorrect_content_length_value, | |
| 579 | + colon_expected, | |
| 580 | + timeout(Int32). | |
| 581 | + | |
| 582 | +type HTTP_RequestType: | |
| 583 | + get, | |
| 584 | + post. | |
| 585 | + | |
| 586 | +type HTTP_RequestLine: | |
| 587 | + request_line (HTTP_RequestType type, | |
| 588 | + String uri, | |
| 589 | + List(Web_arg) query_string). | |
| 590 | + | |
| 591 | +type EncodingType: | |
| 592 | + www_url, | |
| 593 | + multipart_form_data. | |
| 594 | + | |
| 595 | + | |
| 596 | + | |
| 597 | + | |
| 598 | + | |
| 599 | + *** [2] Tools. | |
| 600 | + | |
| 601 | + *** [2.1] Formating an error message. | |
| 602 | + | |
| 603 | + The next function formats an error message. | |
| 604 | + | |
| 605 | +define String | |
| 606 | + format | |
| 607 | + ( | |
| 608 | + Error msg | |
| 609 | + ) = | |
| 610 | + if msg is | |
| 611 | + { | |
| 612 | + cannot_read_from_connection then | |
| 613 | + "Cannot read from connection.\n", | |
| 614 | + not_get_or_post_request(s) then | |
| 615 | + "The request did not begin by 'GET' or 'POST': "+s+".\n", | |
| 616 | + end_of_line_expected then | |
| 617 | + "End of line expected.\n", | |
| 618 | + incorrect_content_length_value then | |
| 619 | + "Incorrect value for HTTP header 'Content-Length'.\n", | |
| 620 | + colon_expected then | |
| 621 | + "':' was expected.\n", | |
| 622 | + timeout(n) then | |
| 623 | + //"time out: "+n+"\n" | |
| 624 | + //"time out.\n" | |
| 625 | + "" | |
| 626 | + }. | |
| 627 | + | |
| 628 | + | |
| 629 | + | |
| 630 | + | |
| 631 | + | |
| 632 | + | |
| 633 | + *** [2.2] Converting IP addresses. | |
| 634 | + | |
| 635 | + We need two conversion functions for IP addresses: | |
| 636 | + | |
| 637 | + (Word8,Word8,Word8,Word8) --> Int32 ip_address | |
| 638 | + Int32 --> String ip_addr_to_string | |
| 639 | + | |
| 640 | + These conversions are defined in 'tools/basis.anubis'. | |
| 641 | + | |
| 642 | + | |
| 643 | + | |
| 644 | + | |
| 645 | + | |
| 646 | + | |
| 647 | + | |
| 648 | + | |
| 649 | + *** [2.3] Reading and unputting characters. | |
| 650 | + | |
| 651 | + We need a mecanism for unputting several characters (actually at least 3). This is | |
| 652 | + because when reading the client connection, we must sometimes go ahead several | |
| 653 | + characters, and virtually put them back into the connection, so that they can be | |
| 654 | + reread. Of course, we do not send them back to the client. We store them in a list | |
| 655 | + (hold by the variable 'unput_chars'), and we manage this list, so that characters may | |
| 656 | + be virtually put back in the connection (this is called 'unputting'). | |
| 657 | + | |
| 658 | +variable List(Word8) unput_chars = []. | |
| 659 | + | |
| 660 | + The most recently read one is the head of list. Fortunately, this variable is private | |
| 661 | + to this virtual machine (hence to this client). | |
| 662 | + | |
| 663 | + | |
| 664 | +define One | |
| 665 | + unput // unputting a character (add it in front of the list) | |
| 666 | + ( | |
| 667 | + Word8 character | |
| 668 | + ) = | |
| 669 | + unput_chars <- (List(Word8))[character . *unput_chars]. | |
| 670 | + | |
| 671 | + | |
| 672 | + | |
| 673 | +define One record_dubious_IP(Int32 addr,DenialOfService dos). | |
| 674 | + | |
| 675 | +variable Int32 sttm = 0. // contains the start time for this connection. | |
| 676 | + | |
| 677 | +define Result(Error,Word8) | |
| 678 | + record_dubious_connection | |
| 679 | + ( | |
| 680 | + Connection conn, | |
| 681 | + Int32 dead_line, | |
| 682 | + DenialOfService dos, | |
| 683 | + ) = | |
| 684 | + if remote_IP_address_and_port(conn) is (addr,port) then | |
| 685 | + record_dubious_IP(addr,dos); | |
| 686 | + print("Recording IP address "+ip_addr_to_string(addr)+ | |
| 687 | + " as dubious after "+(dead_line-*sttm)+" seconds. Total: "+ | |
| 688 | + length(*list_of_dubious(dos))+"\n"); | |
| 689 | + error(timeout(dead_line)). | |
| 690 | + | |
| 691 | + | |
| 692 | +define Result(Error,Word8) | |
| 693 | + read_one_byte | |
| 694 | + ( | |
| 695 | + Connection connection, | |
| 696 | + Int32 dead_line, | |
| 697 | + DenialOfService dos | |
| 698 | + ) = | |
| 699 | + //if now > dead_line then record_dubious_connection(connection,dead_line,dos) else | |
| 700 | + if read(connection,1,600) is // the connection is closed after 10 minutes of inactivity | |
| 701 | + { | |
| 702 | + error then error(cannot_read_from_connection), | |
| 703 | + timeout then error(timeout(600)), | |
| 704 | + //record_dubious_connection(connection,dead_line,dos), | |
| 705 | + ok(ba) then if nth(0,ba) is | |
| 706 | + { | |
| 707 | + failure then error(cannot_read_from_connection), | |
| 708 | + success(c) then ok(c) | |
| 709 | + } | |
| 710 | + }. | |
| 711 | + | |
| 712 | + | |
| 713 | +define Result(Error,Word8) | |
| 714 | + next_char // reading a character (check the list first, and read on the connection | |
| 715 | + // only when the list is empty). | |
| 716 | + ( | |
| 717 | + Connection connection, | |
| 718 | + Int32 dead_line, | |
| 719 | + DenialOfService dos | |
| 720 | + ) = | |
| 721 | + if *unput_chars is | |
| 722 | + { | |
| 723 | + [ ] then read_one_byte(connection,dead_line,dos), | |
| 724 | + | |
| 725 | + [h . t] then | |
| 726 | + unput_chars <- t; | |
| 727 | + ok(h) | |
| 728 | + }. | |
| 729 | + | |
| 730 | + | |
| 731 | + | |
| 732 | + | |
| 733 | + | |
| 734 | + | |
| 735 | + | |
| 736 | + *** [2.4] Reading and discarding characters. | |
| 737 | + | |
| 738 | + The next function reads the specified number of bytes (this is the same as | |
| 739 | + 'characters') from the connection and discards them. This is used for discarding CR LF | |
| 740 | + just before the body of a request. | |
| 741 | + | |
| 742 | +define Result(Error,One) | |
| 743 | + read_and_ignore | |
| 744 | + ( | |
| 745 | + Connection connection, // to client | |
| 746 | + Int32 dead_line, | |
| 747 | + Int32 number_of_characters, // number of characters to read and ignore | |
| 748 | + DenialOfService dos | |
| 749 | + ) = | |
| 750 | + if number_of_characters =< 0 then ok(unique) else | |
| 751 | + if next_char(connection,dead_line,dos) is | |
| 752 | + { | |
| 753 | + error(msg) then error(msg), | |
| 754 | + ok(c) then read_and_ignore(connection,dead_line,number_of_characters-1,dos) | |
| 755 | + }. | |
| 756 | + | |
| 757 | + | |
| 758 | + | |
| 759 | + | |
| 760 | + | |
| 761 | + | |
| 762 | + | |
| 763 | + *** [2.5] Reading a character string. | |
| 764 | + | |
| 765 | + Sometimes values of HTTP attributes or web args are presented in the form of double | |
| 766 | + quoted strings. The next function handles the reading of such things. The leading | |
| 767 | + double quote is already read in. We must read subsequent characters until the next non | |
| 768 | + backslashed double quote. | |
| 769 | + | |
| 770 | +define Result(Error,String) | |
| 771 | + read_string | |
| 772 | + ( | |
| 773 | + Connection connection, // connection with the client | |
| 774 | + Int32 dead_line, | |
| 775 | + List(Word8) so_far, // characters read so far (in reverse order) | |
| 776 | + DenialOfService dos | |
| 777 | + ) = | |
| 778 | + if next_char(connection,dead_line,dos) is | |
| 779 | + { | |
| 780 | + error(msg) then error(msg), | |
| 781 | + ok(c) then | |
| 782 | + if c = '\\' | |
| 783 | + then if next_char(connection,dead_line,dos) is | |
| 784 | + { | |
| 785 | + error(msg) then error(msg), | |
| 786 | + ok(d) then | |
| 787 | + if d = '\"' | |
| 788 | + then read_string(connection,dead_line,['\"' . so_far],dos) | |
| 789 | + else read_string(connection,dead_line,[d, c . so_far],dos) | |
| 790 | + } | |
| 791 | + else if c = '\"' | |
| 792 | + then ok(implode(reverse(so_far))) | |
| 793 | + else read_string(connection,dead_line,[c . so_far],dos) | |
| 794 | + }. | |
| 795 | + | |
| 796 | + | |
| 797 | + | |
| 798 | + | |
| 799 | + | |
| 800 | + | |
| 801 | + | |
| 802 | + *** [2.6] Padding integers with zeros. | |
| 803 | + | |
| 804 | + 'zero_pad_2' transforms an integer (which is assumed to be between 0 and 99) into a | |
| 805 | + string with exactly two digits. This is used for formating days, hours, minutes and | |
| 806 | + seconds. | |
| 807 | + | |
| 808 | +define String | |
| 809 | + zero_pad_2 | |
| 810 | + ( | |
| 811 | + Int32 n | |
| 812 | + ) = | |
| 813 | + with s = integer_to_string(n), | |
| 814 | + if length(s) < 2 | |
| 815 | + then "0"+s | |
| 816 | + else s. | |
| 817 | + | |
| 818 | + | |
| 819 | + | |
| 820 | + | |
| 821 | + | |
| 822 | + | |
| 823 | + | |
| 824 | + *** [2.7] Converting web arguments to ASCII. | |
| 825 | + | |
| 826 | + The function 'web_to_ascii' gets a character string and replaces web encoding by normal | |
| 827 | + ASCII encoding. This amounts to replacing: | |
| 828 | + | |
| 829 | + + by blank | |
| 830 | + %xx by the character whose ASCII code is xx in hexadecimal | |
| 831 | + | |
| 832 | + Note: We assume that '9' < 'A' (which is the case for ASCII code). | |
| 833 | + | |
| 834 | + | |
| 835 | + | |
| 836 | +define Word8 | |
| 837 | + web_decode | |
| 838 | + ( | |
| 839 | + Word8 x1, | |
| 840 | + Word8 x2 | |
| 841 | + ) = | |
| 842 | + with z1 = word8_to_int32(x1), | |
| 843 | + n1 = if z1 =< '9' then (z1 - '0') else (z1 - 'A' + 10), | |
| 844 | + z2 = word8_to_int32(x2), | |
| 845 | + n2 = if z2 =< '9' then (z2 - '0') else (z2 - 'A' + 10), | |
| 846 | + n = (n1 << 4) + n2, | |
| 847 | + truncate_to_word8(n). | |
| 848 | + | |
| 849 | + | |
| 850 | + | |
| 851 | +define String | |
| 852 | + web_to_ascii | |
| 853 | + ( | |
| 854 | + String web_string, | |
| 855 | + Int32 n, // current position in web_string | |
| 856 | + List(Word8) so_far | |
| 857 | + ) = | |
| 858 | + if nth(n,web_string) is | |
| 859 | + { | |
| 860 | + failure then implode(reverse(so_far)), | |
| 861 | + success(c) then | |
| 862 | + if c = '+' | |
| 863 | + then web_to_ascii(web_string,n+1,[' ' . so_far]) | |
| 864 | + else if c = '%' | |
| 865 | + then if nth(n+1,web_string) is | |
| 866 | + { | |
| 867 | + failure then implode(reverse(so_far)), | |
| 868 | + success(x1) then if nth(n+2,web_string) is | |
| 869 | + { | |
| 870 | + failure then implode(reverse(so_far)), | |
| 871 | + success(x2) then web_to_ascii(web_string,n+3,[web_decode(x1,x2) . so_far]) | |
| 872 | + } | |
| 873 | + } | |
| 874 | + else web_to_ascii(web_string,n+1,[c . so_far]) | |
| 875 | + }. | |
| 876 | + | |
| 877 | + | |
| 878 | + | |
| 879 | + | |
| 880 | + | |
| 881 | + | |
| 882 | + | |
| 883 | + | |
| 884 | + *** [3] Managing the journal. | |
| 885 | + | |
| 886 | + Concurrently working machines should not try to access the same file at the same | |
| 887 | + time. This problem may be solved by using the 'protect' mecanism. | |
| 888 | + | |
| 889 | + | |
| 890 | + | |
| 891 | + *** [3.1] Naming journal files. | |
| 892 | + | |
| 893 | + Since journal messages are rather prolific, we should have at least one file per | |
| 894 | + hour. Hence, the name of a journal file must be constructed from the current year, | |
| 895 | + month, day and hour. For example, it may be: | |
| 896 | + | |
| 897 | + 2003_03_12_19 | |
| 898 | + | |
| 899 | + (this is for the journal of 7 PM to 8 PM, 2003/mar/12). | |
| 900 | + | |
| 901 | +define String | |
| 902 | + make_current_journal_file_name | |
| 903 | + = | |
| 904 | + if convert_time(now) is date_and_time(y,m,d,h,_,_,_,_,_) then | |
| 905 | + integer_to_string(y)+"_"+ | |
| 906 | + zero_pad_2(m)+"_"+ | |
| 907 | + zero_pad_2(d)+"_"+ | |
| 908 | + zero_pad_2(h). | |
| 909 | + | |
| 910 | + | |
| 911 | + | |
| 912 | + | |
| 913 | + | |
| 914 | + | |
| 915 | + | |
| 916 | + *** [3.2] Formating HTTP headers. | |
| 917 | + | |
| 918 | + HTTP headers may be shown on the console or written in the journal. The function below | |
| 919 | + formats a list of HTTP headers. | |
| 920 | + | |
| 921 | +define String | |
| 922 | + show_format | |
| 923 | + ( | |
| 924 | + Web_Site_Description desc, | |
| 925 | + List(HTTP_header) headers, | |
| 926 | + ) = | |
| 927 | + if headers is | |
| 928 | + { | |
| 929 | + [ ] then "", | |
| 930 | + [h . t] then if h is http_header(name,value) then | |
| 931 | + if member(journal_headers(desc),name) | |
| 932 | + then " | "+name+": "+value+"\n"+show_format(desc,t) | |
| 933 | + else show_format(desc,t) | |
| 934 | + }. | |
| 935 | + | |
| 936 | + | |
| 937 | + | |
| 938 | + | |
| 939 | + | |
| 940 | + | |
| 941 | + *** [3.3] Formating web arguments. | |
| 942 | + | |
| 943 | + The same thing for web arguments. | |
| 944 | + | |
| 945 | +define String | |
| 946 | + show_format | |
| 947 | + ( | |
| 948 | + List(Web_arg) lwa | |
| 949 | + ) = | |
| 950 | + if lwa is | |
| 951 | + { | |
| 952 | + [ ] then "", | |
| 953 | + [h . t] then if h is | |
| 954 | + { | |
| 955 | + web_arg(n,v) then | |
| 956 | + " | "+n+"="+(if nth(0,n) = success('p') then "<not shown>" else v)+"\n"+show_format(t), | |
| 957 | + upload(n,fn,tfn) then | |
| 958 | + " | "+n+"="+fn+" (uploaded as '"+tfn+"')\n"+show_format(t) | |
| 959 | + } | |
| 960 | + }. | |
| 961 | + | |
| 962 | + | |
| 963 | + | |
| 964 | + | |
| 965 | + | |
| 966 | + | |
| 967 | + *** [3.4] Formating the whole request. | |
| 968 | + | |
| 969 | + It is cheap to transform month numbers into abbreviated month names. This enhances the | |
| 970 | + readability of the journal. | |
| 971 | + | |
| 972 | +define String | |
| 973 | + format_month | |
| 974 | + ( | |
| 975 | + Int32 m | |
| 976 | + ) = | |
| 977 | + if m = 1 then "jan" else | |
| 978 | + if m = 2 then "feb" else | |
| 979 | + if m = 3 then "mar" else | |
| 980 | + if m = 4 then "apr" else | |
| 981 | + if m = 5 then "may" else | |
| 982 | + if m = 6 then "jun" else | |
| 983 | + if m = 7 then "jul" else | |
| 984 | + if m = 8 then "aug" else | |
| 985 | + if m = 9 then "sep" else | |
| 986 | + if m = 10 then "oct" else | |
| 987 | + if m = 11 then "nov" else | |
| 988 | + if m = 12 then "dec" else | |
| 989 | + "???". | |
| 990 | + | |
| 991 | + | |
| 992 | + Below we format a whole HTTP request. This may give this (actually, it depends on how | |
| 993 | + you defined the values of 'journal_headers' and 'journal_extensions'): | |
| 994 | + | |
| 995 | + [3] 2003/mar/10 10:06:57 from 123.456.123.456: /homepage.awp | |
| 996 | + | host: www.the-best-one.com | |
| 997 | + | user-agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.0.1) Gecko/20020823 Netscape/7.0 | |
| 998 | + | |
| 999 | + The leading number between brackets is the number of the virtual machine which served | |
| 1000 | + the URI. | |
| 1001 | + | |
| 1002 | +define String | |
| 1003 | + format_request | |
| 1004 | + ( | |
| 1005 | + Web_Site_Description desc, | |
| 1006 | + Connection client_connection, | |
| 1007 | + HTTP_RequestLine request_line, | |
| 1008 | + List(HTTP_header) headers, | |
| 1009 | + List(Web_arg) web_args | |
| 1010 | + ) = | |
| 1011 | + with dt = convert_time(now), | |
| 1012 | + if remote_IP_address_and_port(client_connection) is (addr,port) then | |
| 1013 | + integer_to_string(year(dt))+"/"+format_month(month(dt))+"/"+zero_pad_2(day(dt))+" "+ | |
| 1014 | + zero_pad_2(hour(dt))+":"+zero_pad_2(minute(dt))+":"+zero_pad_2(second(dt))+ | |
| 1015 | + " from "+ip_addr_to_string(addr)+ | |
| 1016 | + ": "+uri(request_line)+"\n"+ | |
| 1017 | + show_format(desc,headers)+ | |
| 1018 | + show_format(web_args). | |
| 1019 | + | |
| 1020 | + | |
| 1021 | + | |
| 1022 | + | |
| 1023 | + | |
| 1024 | + | |
| 1025 | + | |
| 1026 | + *** [3.5] Putting it in the journal file (and on the console). | |
| 1027 | + | |
| 1028 | + We must not forget to 'protect' this operation, so that the messages of two machines | |
| 1029 | + (working for the same site) will not be mixed together. | |
| 1030 | + | |
| 1031 | +define One | |
| 1032 | + log_journal_msg | |
| 1033 | + ( | |
| 1034 | + Web_Site_Description desc, | |
| 1035 | + String msg, | |
| 1036 | + ) = | |
| 1037 | + with msg = to_byte_array("["+virtual_machine_id+"] "+msg+"\n"), | |
| 1038 | + protect | |
| 1039 | + ( | |
| 1040 | + if file(site_directory(desc)+"/journal/"+make_current_journal_file_name,append) is | |
| 1041 | + { | |
| 1042 | + failure then unique, | |
| 1043 | + success(journal_file) then | |
| 1044 | + forget(reliable_write(file(journal_file),msg)) | |
| 1045 | + }; | |
| 1046 | + forget(reliable_write(file(stdout),msg)) | |
| 1047 | + ). | |
| 1048 | + | |
| 1049 | + | |
| 1050 | + | |
| 1051 | + | |
| 1052 | + | |
| 1053 | + | |
| 1054 | + | |
| 1055 | + *** [4] Reading the HTTP request. | |
| 1056 | + | |
| 1057 | + | |
| 1058 | + *** [4.1] Skipping leading blanks. | |
| 1059 | + | |
| 1060 | + One of the peculiarities of HTTP is that the characters 13 (carriage return) and 10 | |
| 1061 | + (line feed) followed by either a space (32) or a tab (9), is considered as a blank not | |
| 1062 | + containing any new line. 'skip_http_blanks' must skip all blanks characters until the | |
| 1063 | + first non blank character, which should not be read in. Obviously, because of the above | |
| 1064 | + peculiarity, we need at least 3 characters of lookahead to do this. In other words, we | |
| 1065 | + must be able to unput at least 3 characters (hopefully we are). | |
| 1066 | + | |
| 1067 | + Strictly blanks characters are 'space' and 'tab'. | |
| 1068 | + | |
| 1069 | +define Bool | |
| 1070 | + is_strict_blank | |
| 1071 | + ( | |
| 1072 | + Word8 c | |
| 1073 | + ) = | |
| 1074 | + if c = ' ' then true else c = '\t'. | |
| 1075 | + | |
| 1076 | + | |
| 1077 | + On the contrary, blanks include 13 and 10. | |
| 1078 | + | |
| 1079 | +define Bool | |
| 1080 | + is_blank | |
| 1081 | + ( | |
| 1082 | + Word8 c | |
| 1083 | + ) = | |
| 1084 | + if c = ' ' then true else | |
| 1085 | + if c = '\t' then true else | |
| 1086 | + if c = 13 then true else | |
| 1087 | + c = 10. | |
| 1088 | + | |
| 1089 | + | |
| 1090 | + Skipping HTTP blanks. | |
| 1091 | + | |
| 1092 | +define Result(Error,One) | |
| 1093 | + skip_http_blanks | |
| 1094 | + ( | |
| 1095 | + Connection connection, | |
| 1096 | + Int32 dead_line, | |
| 1097 | + DenialOfService dos | |
| 1098 | + ) = | |
| 1099 | + if next_char(connection,dead_line,dos) is | |
| 1100 | + { | |
| 1101 | + error(msg) then error(msg), | |
| 1102 | + ok(c) then | |
| 1103 | + if is_strict_blank(c) | |
| 1104 | + then skip_http_blanks(connection,dead_line,dos) | |
| 1105 | + else if c = 13 | |
| 1106 | + then if next_char(connection,dead_line,dos) is | |
| 1107 | + { | |
| 1108 | + error(msg) then error(msg), // (unput(c); ok(unique)), | |
| 1109 | + ok(d) then | |
| 1110 | + if d = 10 | |
| 1111 | + then if next_char(connection,dead_line,dos) is | |
| 1112 | + { | |
| 1113 | + error(msg) then error(msg), // (unput(d); unput(c); ok(unique)), | |
| 1114 | + ok(e) then | |
| 1115 | + if is_strict_blank(e) | |
| 1116 | + then skip_http_blanks(connection,dead_line,dos) | |
| 1117 | + else (unput(e); unput(d); unput(c); ok(unique)) | |
| 1118 | + } | |
| 1119 | + else (unput(d); unput(c); ok(unique)) | |
| 1120 | + } | |
| 1121 | + else (unput(c); ok(unique)) | |
| 1122 | + }. | |
| 1123 | + | |
| 1124 | + | |
| 1125 | + | |
| 1126 | + | |
| 1127 | + | |
| 1128 | + | |
| 1129 | + | |
| 1130 | + | |
| 1131 | + *** [4.2] Reading a new line. | |
| 1132 | + | |
| 1133 | + Normally in HTTP a new line is the sequence 13 10 (carriage return line feed), not | |
| 1134 | + followed by a space or tabulator. If it is followed by a space or tabulator, the three | |
| 1135 | + characters are considered blanks, and no new line has been read. Before trying to read | |
| 1136 | + a new line, we first skip leading spaces and tabs. Then we try to read 13 and 10, and | |
| 1137 | + we read another character. if this character is space or tab, we consider we have read | |
| 1138 | + only blanks and we continue reading in order to find our new line. Otherwise, we unput | |
| 1139 | + this character (which may be for example the first character of the name of the next | |
| 1140 | + header), and answer that we have seen a new line. | |
| 1141 | + | |
| 1142 | + Warning: we must not use this function for reading the last pair (13,10) before the | |
| 1143 | + beginning of the body, because if the body is empty, there is no character to read | |
| 1144 | + after this pair, so that the server could wait for a character which will never | |
| 1145 | + come. This is the reason for 'read_and_ignore' above, which is used precisely for | |
| 1146 | + reading that last (13,10) pair. | |
| 1147 | + | |
| 1148 | +define Result(Error,One) | |
| 1149 | + read_new_line | |
| 1150 | + ( | |
| 1151 | + Connection connection, | |
| 1152 | + Int32 dead_line, | |
| 1153 | + DenialOfService dos | |
| 1154 | + ) = | |
| 1155 | + if skip_http_blanks(connection,dead_line,dos) is | |
| 1156 | + { | |
| 1157 | + error(msg) then error(msg), | |
| 1158 | + ok(_) then | |
| 1159 | + if next_char(connection,dead_line,dos) is | |
| 1160 | + { | |
| 1161 | + error(msg) then error(msg), | |
| 1162 | + ok(c) then | |
| 1163 | + if c = 13 | |
| 1164 | + then if next_char(connection,dead_line,dos) is | |
| 1165 | + { | |
| 1166 | + error(msg) then error(msg), | |
| 1167 | + ok(d) then | |
| 1168 | + if d = 10 | |
| 1169 | + then ok(unique) | |
| 1170 | + else (unput(d); | |
| 1171 | + unput(c); | |
| 1172 | + error(end_of_line_expected)) | |
| 1173 | + } | |
| 1174 | + else (unput(c); | |
| 1175 | + error(end_of_line_expected)) | |
| 1176 | + }}. | |
| 1177 | + | |
| 1178 | + | |
| 1179 | + | |
| 1180 | + | |
| 1181 | + | |
| 1182 | + | |
| 1183 | + | |
| 1184 | + | |
| 1185 | + *** [4.3] Reading a 'word'. | |
| 1186 | + | |
| 1187 | + A 'word' is a sequence of characters which begins either by a double quote or not by a | |
| 1188 | + double quote. (However, any leading blanks are read in and ignored. This is | |
| 1189 | + accomplished by 'skip_http_blanks'.) If it begins by a double quote, it is read like a | |
| 1190 | + string, i.e. it ends at the next (non backslashed) double quote. Otherwise, it is | |
| 1191 | + right delimited by any character which may be considered as 'blank'. If the word is | |
| 1192 | + double quoted, the closing double quote is read in. On the contrary, if the word is not | |
| 1193 | + double quoted, the right delimiting blank character is not read in (it is 'unput' back | |
| 1194 | + into the connection), and may be read in again. This is needed because carriage return | |
| 1195 | + or line feed which are 'blank', also have a meaning in HTTP. | |
| 1196 | + | |
| 1197 | +define Result(Error,String) | |
| 1198 | + read_word_aux | |
| 1199 | + ( | |
| 1200 | + Connection connection, | |
| 1201 | + Int32 dead_line, | |
| 1202 | + List(Word8) so_far, | |
| 1203 | + DenialOfService dos | |
| 1204 | + ) = | |
| 1205 | + if next_char(connection,dead_line,dos) is | |
| 1206 | + { | |
| 1207 | + error(msg) then error(msg), | |
| 1208 | + ok(c) then | |
| 1209 | + if is_blank(c) | |
| 1210 | + then (unput(c); | |
| 1211 | + ok(implode(reverse(so_far)))) | |
| 1212 | + else read_word_aux(connection,dead_line,[c . so_far],dos) | |
| 1213 | + }. | |
| 1214 | + | |
| 1215 | +define Result(Error,String) | |
| 1216 | + read_word | |
| 1217 | + ( | |
| 1218 | + Connection connection, | |
| 1219 | + Int32 dead_line, | |
| 1220 | + DenialOfService dos | |
| 1221 | + ) = | |
| 1222 | + if skip_http_blanks(connection,dead_line,dos) is | |
| 1223 | + { | |
| 1224 | + error(msg) then error(msg), | |
| 1225 | + ok(_) then | |
| 1226 | + if next_char(connection,dead_line,dos) is | |
| 1227 | + { | |
| 1228 | + error(msg) then error(msg), | |
| 1229 | + ok(c) then | |
| 1230 | + if c = '\"' | |
| 1231 | + then read_string(connection,dead_line,[],dos) | |
| 1232 | + else read_word_aux(connection,dead_line,[c],dos) | |
| 1233 | + } | |
| 1234 | + }. | |
| 1235 | + | |
| 1236 | + | |
| 1237 | + | |
| 1238 | + | |
| 1239 | + | |
| 1240 | + | |
| 1241 | + | |
| 1242 | + | |
| 1243 | + *** [4.4] Separating the URI from the query string. | |
| 1244 | + | |
| 1245 | + A 'query string' may be postfixed to the URI, just after a question mark. For example, | |
| 1246 | + the client may send the following request: | |
| 1247 | + | |
| 1248 | + GET /catalog.awp?item=3&color=blue | |
| 1249 | + | |
| 1250 | + We separate this into an URI: "/catalog.awp" and the string: "item=3&color=blue" which | |
| 1251 | + will be later transformed into the list: | |
| 1252 | + | |
| 1253 | + [web_arg("item","3"),web_arg("color","blue")] | |
| 1254 | + | |
| 1255 | + | |
| 1256 | +define (String,String) | |
| 1257 | + separate_uri_from_query_string | |
| 1258 | + ( | |
| 1259 | + String uri_and_query_string, | |
| 1260 | + Int32 n | |
| 1261 | + ) = | |
| 1262 | + if nth(n,uri_and_query_string) is | |
| 1263 | + { | |
| 1264 | + failure then (uri_and_query_string,""), | |
| 1265 | + success(c) then | |
| 1266 | + if c = '?' | |
| 1267 | + then (substr(uri_and_query_string,0,n), | |
| 1268 | + substr(uri_and_query_string,n+1,length(uri_and_query_string)-(n+1))) | |
| 1269 | + else separate_uri_from_query_string(uri_and_query_string,n+1) | |
| 1270 | + }. | |
| 1271 | + | |
| 1272 | + | |
| 1273 | + | |
| 1274 | + | |
| 1275 | + | |
| 1276 | + | |
| 1277 | + | |
| 1278 | + | |
| 1279 | + | |
| 1280 | + *** [4.5] Reading the web arguments. | |
| 1281 | + | |
| 1282 | + HTTP/HTTPS requests are sent in one of two formats: | |
| 1283 | + | |
| 1284 | + (1) www-url encoded | |
| 1285 | + (2) multipart/form-data encoded | |
| 1286 | + | |
| 1287 | + The first one is the normal (historical) way of encoding. The second one is required | |
| 1288 | + for uploading files. A server which is supposed to accept upload of files must handle | |
| 1289 | + both formats. The first thing to do is to decide the format of the request. This is | |
| 1290 | + easily done by examining the HTTP headers. If we find the header: | |
| 1291 | + | |
| 1292 | + Content-Type: multipart/form-data | |
| 1293 | + | |
| 1294 | + the request is multipart/form-data encoded. Otherwise, it is 'www-url' encoded. We | |
| 1295 | + first consider 'www-url' encoded requests. | |
| 1296 | + | |
| 1297 | + For a 'www-url' encoded request, the web argument are either in the query string or in | |
| 1298 | + the body of the request, or both. The format is the same for both: | |
| 1299 | + | |
| 1300 | + name=value&name=value&... | |
| 1301 | + | |
| 1302 | + However, we may also have | |
| 1303 | + | |
| 1304 | + name | |
| 1305 | + name= | |
| 1306 | + name=&... | |
| 1307 | + name&... | |
| 1308 | + | |
| 1309 | + i.e. some parts may be missing. Hence, we must be careful. | |
| 1310 | + | |
| 1311 | + Furthermore, web arguments must be translated from web to ASCII when www-url encoded. | |
| 1312 | + | |
| 1313 | +define Bool | |
| 1314 | + is_ampersand_or_equal | |
| 1315 | + ( | |
| 1316 | + Word8 c | |
| 1317 | + ) = | |
| 1318 | + if c = '&' then true else c = '='. | |
| 1319 | + | |
| 1320 | + | |
| 1321 | + | |
| 1322 | + The function 'read_name_or_value' reads the string 's' starting at position 'n' until | |
| 1323 | + either the end of the string or the first '&' or '='. | |
| 1324 | + | |
| 1325 | +define String | |
| 1326 | + read_name_or_value | |
| 1327 | + ( | |
| 1328 | + String s, | |
| 1329 | + Int32 start, | |
| 1330 | + Int32 i | |
| 1331 | + ) = | |
| 1332 | + if nth(i,s) is | |
| 1333 | + { | |
| 1334 | + failure then substr(s,start,i - start), | |
| 1335 | + success(c) then | |
| 1336 | + if is_ampersand_or_equal(c) | |
| 1337 | + then substr(s,start,i-start) // the separator is not included | |
| 1338 | + else read_name_or_value(s,start,i+1) | |
| 1339 | + }. | |
| 1340 | + | |
| 1341 | + | |
| 1342 | +define List(Web_arg) | |
| 1343 | + read_www_url_encoded_web_args | |
| 1344 | + ( | |
| 1345 | + String s, | |
| 1346 | + Int32 start, | |
| 1347 | + ) = | |
| 1348 | + with first = read_name_or_value(s,start,start), | |
| 1349 | + if first = "" | |
| 1350 | + then [] | |
| 1351 | + else with i = start+length(first), | |
| 1352 | + if nth(i,s) is | |
| 1353 | + { | |
| 1354 | + failure then [web_arg(first,"")], | |
| 1355 | + success(c) then | |
| 1356 | + if c = '&' | |
| 1357 | + then [web_arg(first,"") . read_www_url_encoded_web_args(s,i+1)] | |
| 1358 | + else if c = '=' | |
| 1359 | + then with second1 = read_name_or_value(s,i+1,i+1), | |
| 1360 | + // print("\""+second1+"\"\n"); | |
| 1361 | + with second = web_to_ascii(second1,0,[]), | |
| 1362 | + [web_arg(first,second) . read_www_url_encoded_web_args(s,i+length(second1)+2)] | |
| 1363 | + else alert | |
| 1364 | + }. | |
| 1365 | + | |
| 1366 | + | |
| 1367 | + | |
| 1368 | + | |
| 1369 | + | |
| 1370 | + *** [4.7] Reading the request line. | |
| 1371 | + | |
| 1372 | + 'read_request_line' reads three words and a new line from the connection. It tries to | |
| 1373 | + recognize "GET" or "POST" in the first word, separates the URI from the query string in | |
| 1374 | + the second word, transforms the query string into a list of 'Web_arg', and finally | |
| 1375 | + returns a datum of type 'HTTP_RequestLine' if no error arose. | |
| 1376 | + | |
| 1377 | + | |
| 1378 | +define Result(Error,HTTP_RequestType) | |
| 1379 | + identify_get_or_post | |
| 1380 | + ( | |
| 1381 | + String s | |
| 1382 | + ) = | |
| 1383 | + with s = to_lower(s), | |
| 1384 | + if s = "get" then ok(get) else | |
| 1385 | + if s = "post" then ok(post) else | |
| 1386 | + error(not_get_or_post_request(s)). | |
| 1387 | + | |
| 1388 | +define Result(Error,HTTP_RequestLine) | |
| 1389 | + read_request_line | |
| 1390 | + ( | |
| 1391 | + Connection connection, | |
| 1392 | + Int32 dead_line, | |
| 1393 | + DenialOfService dos | |
| 1394 | + ) = | |
| 1395 | + if read_word(connection,dead_line,dos) is | |
| 1396 | + { | |
| 1397 | + error(msg) then error(msg), | |
| 1398 | + ok(get_or_post) then if read_word(connection,dead_line,dos) is | |
| 1399 | + { | |
| 1400 | + error(msg) then error(msg), | |
| 1401 | + ok(uri_and_query_string) then if read_word(connection,dead_line,dos) is | |
| 1402 | + { | |
| 1403 | + error(msg) then error(msg), | |
| 1404 | + ok(http_version) then if read_new_line(connection,dead_line,dos) is | |
| 1405 | + { | |
| 1406 | + error(msg) then error(msg), | |
| 1407 | + ok(_) then if separate_uri_from_query_string(uri_and_query_string,0) is | |
| 1408 | + (uri,query_string) then if identify_get_or_post(get_or_post) is | |
| 1409 | + { | |
| 1410 | + error(msg) then error(msg), | |
| 1411 | + ok(request_type) then | |
| 1412 | + ok(request_line(request_type,uri,read_www_url_encoded_web_args(query_string,0))) | |
| 1413 | + } | |
| 1414 | + } | |
| 1415 | + } | |
| 1416 | + } | |
| 1417 | + }. | |
| 1418 | + | |
| 1419 | + | |
| 1420 | + | |
| 1421 | + | |
| 1422 | + | |
| 1423 | + | |
| 1424 | + | |
| 1425 | + *** [4.8] Reading the HTTP headers. | |
| 1426 | + | |
| 1427 | + Each header is made of a name (containing only letters, the underscore, digits and the | |
| 1428 | + minus sign), a colon, a value, and a new line. The first empty line ends the headers. | |
| 1429 | + | |
| 1430 | + | |
| 1431 | + The next function tests characters acceptable in a header name. | |
| 1432 | + | |
| 1433 | +define Bool | |
| 1434 | + is_header_name_char | |
| 1435 | + ( | |
| 1436 | + Word8 c | |
| 1437 | + ) = | |
| 1438 | + with n = word8_to_int32(c), | |
| 1439 | + if ('a' =< n & n =< 'z') then true else | |
| 1440 | + if ('A' =< n & n =< 'Z') then true else | |
| 1441 | + if ('0' =< n & n =< '9') then true else | |
| 1442 | + if c = '-' then true else | |
| 1443 | + c = '_'. | |
| 1444 | + | |
| 1445 | +define Result(Error,String) | |
| 1446 | + read_header_name | |
| 1447 | + ( | |
| 1448 | + Connection connection, | |
| 1449 | + Int32 dead_line, | |
| 1450 | + List(Word8) so_far, | |
| 1451 | + DenialOfService dos | |
| 1452 | + ) = | |
| 1453 | + if next_char(connection,dead_line,dos) is | |
| 1454 | + { | |
| 1455 | + error(msg) then error(msg), | |
| 1456 | + ok(c) then | |
| 1457 | + if is_header_name_char(c) | |
| 1458 | + then read_header_name(connection,dead_line,[to_lower(c) . so_far],dos) | |
| 1459 | + else unput(c); ok(implode(reverse(so_far))) | |
| 1460 | + }. | |
| 1461 | + | |
| 1462 | +define Result(Error,One) | |
| 1463 | + skip_colon | |
| 1464 | + ( | |
| 1465 | + Connection connection, | |
| 1466 | + Int32 dead_line, | |
| 1467 | + DenialOfService dos | |
| 1468 | + ) = | |
| 1469 | + if skip_http_blanks(connection,dead_line,dos) is | |
| 1470 | + { | |
| 1471 | + error(msg) then error(msg), | |
| 1472 | + ok(_) then | |
| 1473 | + if next_char(connection,dead_line,dos) is | |
| 1474 | + { | |
| 1475 | + error(msg) then error(msg), | |
| 1476 | + ok(c) then | |
| 1477 | + if c = ':' | |
| 1478 | + then ok(unique) | |
| 1479 | + else error(colon_expected) | |
| 1480 | + }}. | |
| 1481 | + | |
| 1482 | + | |
| 1483 | +define Result(Error,String) | |
| 1484 | + read_header_value | |
| 1485 | + ( | |
| 1486 | + Connection connection, | |
| 1487 | + Int32 dead_line, | |
| 1488 | + List(Word8) so_far, | |
| 1489 | + DenialOfService dos | |
| 1490 | + ) = | |
| 1491 | + if next_char(connection,dead_line,dos) is | |
| 1492 | + { | |
| 1493 | + error(msg) then error(msg), | |
| 1494 | + ok(c) then | |
| 1495 | + if c = 13 | |
| 1496 | + then if next_char(connection,dead_line,dos) is | |
| 1497 | + { | |
| 1498 | + error(msg) then error(msg), | |
| 1499 | + ok(d) then | |
| 1500 | + if d = 10 | |
| 1501 | + then if next_char(connection,dead_line,dos) is | |
| 1502 | + { | |
| 1503 | + error(msg) then error(msg), | |
| 1504 | + ok(e) then | |
| 1505 | + if is_strict_blank(e) | |
| 1506 | + then read_header_value(connection,dead_line,[e . so_far],dos) | |
| 1507 | + else (unput(e); ok(implode(reverse(so_far)))) | |
| 1508 | + } | |
| 1509 | + else read_header_value(connection,dead_line,[d, c . so_far],dos) | |
| 1510 | + } | |
| 1511 | + else read_header_value(connection,dead_line,[c . so_far],dos) | |
| 1512 | + }. | |
| 1513 | + | |
| 1514 | + | |
| 1515 | + Reading a single header. | |
| 1516 | + | |
| 1517 | +define Result(Error,Maybe(HTTP_header)) | |
| 1518 | + read_header | |
| 1519 | + ( | |
| 1520 | + Connection connection, | |
| 1521 | + Int32 dead_line, | |
| 1522 | + DenialOfService dos | |
| 1523 | + ) = | |
| 1524 | + if read_header_name(connection,dead_line,[],dos) is | |
| 1525 | + { | |
| 1526 | + error(msg) then error(msg), | |
| 1527 | + ok(name) then | |
| 1528 | + if name = "" then | |
| 1529 | + if read_and_ignore(connection,dead_line,2,dos) /* 13 and 10 */ is | |
| 1530 | + { | |
| 1531 | + error(msg) then error(msg), | |
| 1532 | + ok(_) then // this is the blank line | |
| 1533 | + ok(failure) // end of headers | |
| 1534 | + } | |
| 1535 | + else if skip_colon(connection,dead_line,dos) is | |
| 1536 | + { | |
| 1537 | + error(msg) then error(msg), | |
| 1538 | + ok(_) then if skip_http_blanks(connection,dead_line,dos) is | |
| 1539 | + { | |
| 1540 | + error(msg) then error(msg), | |
| 1541 | + ok(_) then if read_header_value(connection,dead_line,[],dos) is | |
| 1542 | + { | |
| 1543 | + error(msg) then error(msg), | |
| 1544 | + ok(value) then | |
| 1545 | + ok(success(http_header(name,value))) | |
| 1546 | + } | |
| 1547 | + } | |
| 1548 | + } | |
| 1549 | + }. | |
| 1550 | + | |
| 1551 | + | |
| 1552 | + | |
| 1553 | + Reading all the headers. | |
| 1554 | + | |
| 1555 | +define Result(Error,List(HTTP_header)) | |
| 1556 | + read_http_headers | |
| 1557 | + ( | |
| 1558 | + Connection connection, | |
| 1559 | + Int32 dead_line, | |
| 1560 | + DenialOfService dos | |
| 1561 | + ) = | |
| 1562 | + if read_header(connection,dead_line,dos) is | |
| 1563 | + { | |
| 1564 | + error(msg) then error(msg), | |
| 1565 | + ok(mbh) then if mbh is | |
| 1566 | + { | |
| 1567 | + failure then ok([ ]), | |
| 1568 | + success(header) then | |
| 1569 | + if read_http_headers(connection,dead_line,dos) is | |
| 1570 | + { | |
| 1571 | + error(msg) then error(msg), | |
| 1572 | + ok(others) then ok([header . others]) | |
| 1573 | + } | |
| 1574 | + } | |
| 1575 | + }. | |
| 1576 | + | |
| 1577 | + | |
| 1578 | + | |
| 1579 | + | |
| 1580 | + | |
| 1581 | + | |
| 1582 | + | |
| 1583 | + *** [4.9] Getting the size of the request's body. | |
| 1584 | + | |
| 1585 | + The size of the body of the request is given under the 'Content-Length' header. If this | |
| 1586 | + header is not present, the size is assumed to be zero. | |
| 1587 | + | |
| 1588 | +define Result(Error,Int32) | |
| 1589 | + get_body_size | |
| 1590 | + ( | |
| 1591 | + List(HTTP_header) headers | |
| 1592 | + ) = | |
| 1593 | + if headers is | |
| 1594 | + { | |
| 1595 | + [ ] then ok(0), | |
| 1596 | + [h . t] then if h is http_header(name,value) then | |
| 1597 | + if name = "content-length" | |
| 1598 | + then if string_to_integer(value) is | |
| 1599 | + { | |
| 1600 | + failure then error(incorrect_content_length_value), | |
| 1601 | + success(n) then ok(n) | |
| 1602 | + } | |
| 1603 | + else get_body_size(t) | |
| 1604 | + }. | |
| 1605 | + | |
| 1606 | + | |
| 1607 | + | |
| 1608 | + | |
| 1609 | + | |
| 1610 | + | |
| 1611 | + | |
| 1612 | + | |
| 1613 | + | |
| 1614 | + | |
| 1615 | + *** [4.10] Reading the body of the request. | |
| 1616 | + | |
| 1617 | + The body of the request may be very big (it contains uploaded files, if any). We read | |
| 1618 | + it using the primitive 'read', which returns the number of bytes read, which may be | |
| 1619 | + less than the number of bytes we wanted to read. This is not an error, but simply due | |
| 1620 | + to the fact the buffer associated with the connection in the Linux (or MS-Windows) | |
| 1621 | + kernel has a limited size. Hence, we must read bytes again until we have read the | |
| 1622 | + required number of bytes. However, if the number of bytes read is zero, the connection | |
| 1623 | + may be broken. In that case, we must not try to read indefinitely. On the contrary, we | |
| 1624 | + make at most 10 retries, with a small sleeping time between any two of them. | |
| 1625 | + | |
| 1626 | +define Result(Error,ByteArray) | |
| 1627 | + read_http_body | |
| 1628 | + ( | |
| 1629 | + Connection connection, | |
| 1630 | + Int32 body_size, | |
| 1631 | + ByteArray so_far, // when calling this function, 'so_far' is the empty byte array | |
| 1632 | + Int32 retries // this function is called with retries = 10 | |
| 1633 | + ) = | |
| 1634 | + if body_size = 0 then ok(constant_byte_array(0,0)) else | |
| 1635 | + if retries =< 0 then error(cannot_read_from_connection) else | |
| 1636 | + if read(connection,body_size,60) is | |
| 1637 | + { | |
| 1638 | + error then error(cannot_read_from_connection), | |
| 1639 | + timeout then error(timeout(60)), | |
| 1640 | + ok(new_bytes) then with | |
| 1641 | + ba = so_far + new_bytes, // contains all the bytes read so far | |
| 1642 | + nr = length(ba), // total read since the beginning | |
| 1643 | + nn = length(new_bytes), // number of bytes just read | |
| 1644 | + if nr < body_size // must read more bytes | |
| 1645 | + then if nn > 0 // if connection seems to work | |
| 1646 | + then read_http_body(connection,body_size,ba,1000) // continue reading | |
| 1647 | + else sleep(100); // otherwise, sleep 1/10 of second | |
| 1648 | + read_http_body(connection,body_size,ba, // and retry reading | |
| 1649 | + retries-1) // but no more than 10 times | |
| 1650 | + else ok(ba) // required number of bytes has been read | |
| 1651 | + }. | |
| 1652 | + | |
| 1653 | + | |
| 1654 | + Note: During sleeping, 'anbexec' runs other machines. Actually, calling 'sleep', even | |
| 1655 | + for one millisecond, is some way of giving up explicitly, so that other virtual | |
| 1656 | + machines may work. | |
| 1657 | + | |
| 1658 | + | |
| 1659 | + | |
| 1660 | + | |
| 1661 | + | |
| 1662 | + | |
| 1663 | + | |
| 1664 | + | |
| 1665 | + | |
| 1666 | + | |
| 1667 | + | |
| 1668 | + | |
| 1669 | + *** [5] Making the HTTP answer. | |
| 1670 | + | |
| 1671 | + At that point we have read the request line, the headers and the body of the | |
| 1672 | + request, and we must decide what to do. | |
| 1673 | + | |
| 1674 | + Actually, we can do one of the following: | |
| 1675 | + | |
| 1676 | + - send a file, | |
| 1677 | + - execute 'tickets_and_web_page' in case of an ".awp" URI. | |
| 1678 | + | |
| 1679 | + The uploaded file (which are in the body of the request) are saved into temporary files | |
| 1680 | + below. | |
| 1681 | + | |
| 1682 | + | |
| 1683 | + | |
| 1684 | + | |
| 1685 | + | |
| 1686 | + *** [5.1] Avoiding illegal URIs. | |
| 1687 | + | |
| 1688 | + For security reasons, we must avoid illegal URIs, for example those which may climb up | |
| 1689 | + in the file hierarchy. First we accept only few characters in URIs. | |
| 1690 | + | |
| 1691 | +define Bool | |
| 1692 | + is_legal_uri_char | |
| 1693 | + ( | |
| 1694 | + Word8 c | |
| 1695 | + ) = | |
| 1696 | + with n = word8_to_int32(c), | |
| 1697 | + if ('a' =< n & n =< 'z') then true else // accept 'a' to 'z' | |
| 1698 | + if ('A' =< n & n =< 'Z') then true else // accept 'A' to 'Z' | |
| 1699 | + if ('0' =< n & n =< '9') then true else // accept '0' to '9' | |
| 1700 | + if c = '.' then true else // accept '.' '-' '/' and '_' | |
| 1701 | + if c = '-' then true else | |
| 1702 | + if c = '/' then true else | |
| 1703 | + c = '_'. | |
| 1704 | + | |
| 1705 | + We do not accept ~ which is some way of climbing. Of course, we cannot disallow single | |
| 1706 | + dots, which are most often present in legal URIs, but we must avoid double dots .. | |
| 1707 | + which mean 'climb up'. | |
| 1708 | + | |
| 1709 | +define Bool | |
| 1710 | + is_illegal_uri | |
| 1711 | + ( | |
| 1712 | + String uri, | |
| 1713 | + Int32 n | |
| 1714 | + ) = | |
| 1715 | + if nth(n,uri) is | |
| 1716 | + { | |
| 1717 | + failure then false, | |
| 1718 | + success(c) then | |
| 1719 | + if c = '.' // first dot | |
| 1720 | + then if nth(n+1,uri) is | |
| 1721 | + { | |
| 1722 | + failure then false, | |
| 1723 | + success(d) then | |
| 1724 | + if d = '.' // second dot | |
| 1725 | + then true | |
| 1726 | + else is_illegal_uri(uri,n+1) | |
| 1727 | + } | |
| 1728 | + else is_illegal_uri(uri,n+1) | |
| 1729 | + }. | |
| 1730 | + | |
| 1731 | + | |
| 1732 | + | |
| 1733 | + | |
| 1734 | + | |
| 1735 | + | |
| 1736 | + *** [5.2] Managing authorizations for downloading private files. | |
| 1737 | + | |
| 1738 | + Computing the authorization and making the authorization file (containing the absolute | |
| 1739 | + path of the file on the server). | |
| 1740 | + | |
| 1741 | + | |
| 1742 | +define String | |
| 1743 | + compute_authorization | |
| 1744 | + ( | |
| 1745 | + String authorization_secret, | |
| 1746 | + String absolute_path | |
| 1747 | + ) = | |
| 1748 | + to_ascii(sha1((authorization_secret, | |
| 1749 | + absolute_path))). | |
| 1750 | + | |
| 1751 | + | |
| 1752 | +public define String | |
| 1753 | + make_authorization | |
| 1754 | + ( | |
| 1755 | + String site_directory, | |
| 1756 | + String authorization_secret, | |
| 1757 | + String absolute_path | |
| 1758 | + ) = | |
| 1759 | + with private_download_dir = site_directory+"/private_download", | |
| 1760 | + auth = compute_authorization(authorization_secret, | |
| 1761 | + absolute_path), | |
| 1762 | + forget(save(absolute_path, | |
| 1763 | + private_download_dir+"/z"+auth)); | |
| 1764 | + auth. | |
| 1765 | + | |
| 1766 | + | |
| 1767 | + The function 'send_file' defined below handles the recognition of authorizations. | |
| 1768 | + | |
| 1769 | + | |
| 1770 | + | |
| 1771 | + | |
| 1772 | + | |
| 1773 | + *** [5.3] Recognizing MIME types. | |
| 1774 | + | |
| 1775 | + The extension of the (redirected) URI must be either ".awp" or recognized as associated | |
| 1776 | + to a MIME type. Otherwise, the server will not send the file. This is for security, but | |
| 1777 | + also because, we must generate a 'Content-Type' header in the answer, with the right | |
| 1778 | + MIME type. | |
| 1779 | + | |
| 1780 | +define String | |
| 1781 | + get_uri_extension_aux | |
| 1782 | + ( | |
| 1783 | + String uri, | |
| 1784 | + Int32 n // used for searching backwards | |
| 1785 | + ) = | |
| 1786 | + if nth(n,uri) is | |
| 1787 | + { | |
| 1788 | + failure then "", | |
| 1789 | + success(c) then | |
| 1790 | + if c = '.' then substr(uri,n,length(uri)-n) | |
| 1791 | + else if c = '/' then "" | |
| 1792 | + else get_uri_extension_aux(uri,n-1) | |
| 1793 | + }. | |
| 1794 | + | |
| 1795 | +public define String | |
| 1796 | + get_uri_extension | |
| 1797 | + ( | |
| 1798 | + String uri | |
| 1799 | + ) = | |
| 1800 | + get_uri_extension_aux(uri, | |
| 1801 | + length(uri)-1). // search starts at the right end | |
| 1802 | + | |
| 1803 | + | |
| 1804 | + | |
| 1805 | +define Maybe(String) | |
| 1806 | + recognize_mime_type_from_ext | |
| 1807 | + ( | |
| 1808 | + String ext, | |
| 1809 | + List(MIME) l | |
| 1810 | + ) = | |
| 1811 | + if l is | |
| 1812 | + { | |
| 1813 | + [ ] then success("application/octet-stream"), // failure, | |
| 1814 | + [h . t] then if h is mime(mime_type,extension) then | |
| 1815 | + if ext = extension | |
| 1816 | + then success(mime_type) | |
| 1817 | + else recognize_mime_type_from_ext(ext,t) | |
| 1818 | + }. | |
| 1819 | + | |
| 1820 | +define Maybe(String) | |
| 1821 | + recognize_mime_type_from_uri | |
| 1822 | + ( | |
| 1823 | + Web_Site_Description desc, | |
| 1824 | + String uri | |
| 1825 | + ) = | |
| 1826 | + recognize_mime_type_from_ext(get_uri_extension(uri),known_mime_types(desc)). | |
| 1827 | + | |
| 1828 | + | |
| 1829 | + | |
| 1830 | + | |
| 1831 | + | |
| 1832 | + | |
| 1833 | + | |
| 1834 | + | |
| 1835 | + *** [5.4] Formating HTTP headers. | |
| 1836 | + | |
| 1837 | + This is the formating for sending to the client (hence, it has nothing to do with the | |
| 1838 | + component 'journal_headers' in the web site description). | |
| 1839 | + | |
| 1840 | +define Printable_tree | |
| 1841 | + format_headers | |
| 1842 | + ( | |
| 1843 | + List(HTTP_header) headers | |
| 1844 | + ) = | |
| 1845 | + if headers is | |
| 1846 | + { | |
| 1847 | + [ ] then [ ], | |
| 1848 | + [h . t] then if h is http_header(name,value) then | |
| 1849 | + [name,": ",value,crlf . format_headers(t)] | |
| 1850 | + }. | |
| 1851 | + | |
| 1852 | + | |
| 1853 | + | |
| 1854 | + | |
| 1855 | + | |
| 1856 | + | |
| 1857 | + *** [5.5] Sending a file. | |
| 1858 | + | |
| 1859 | + We send 2 headers 'Content-Type' and 'Content-Length'. | |
| 1860 | + | |
| 1861 | +define List(HTTP_header) | |
| 1862 | + headers_for_send_file | |
| 1863 | + ( | |
| 1864 | + String mime_type, | |
| 1865 | + Int32 size, | |
| 1866 | + String etag, | |
| 1867 | + ) = | |
| 1868 | + [ | |
| 1869 | + http_header("Content-Type",mime_type), | |
| 1870 | + http_header("Etag", etag), | |
| 1871 | + http_header("Content-Length",integer_to_string(size)), | |
| 1872 | + ]. | |
| 1873 | + | |
| 1874 | + | |
| 1875 | + | |
| 1876 | + Sending the body of the answer (i.e. the file itself). | |
| 1877 | + | |
| 1878 | +define One | |
| 1879 | + send_file_body | |
| 1880 | + ( | |
| 1881 | + Web_Site_Description desc, | |
| 1882 | + Connection connection, // connection with the client | |
| 1883 | + Connection file, // file to be sent already opened | |
| 1884 | + Int32 size, // size of file | |
| 1885 | + Int32 sent, // bytes already sent | |
| 1886 | + String filename // name of file | |
| 1887 | + ) = | |
| 1888 | + if sent >= size then unique else | |
| 1889 | + if read(file,min(10000,size-sent),60) is | |
| 1890 | + { | |
| 1891 | + error then log_journal_msg(desc,"Cannot read from file '"+filename+"'.\n"), | |
| 1892 | + timeout then log_journal_msg(desc,"Cannot read from file timeoput'"+filename+"'.\n"), | |
| 1893 | + ok(ba) then | |
| 1894 | + with nr = length(ba), // get the number of bytes read | |
| 1895 | + if reliable_write(connection,ba) is | |
| 1896 | + { | |
| 1897 | + failure then log_journal_msg(desc,"Cannot write into connection.\n"), | |
| 1898 | + success(nw) then | |
| 1899 | + send_file_body(desc,connection,file,size,sent+nw,filename) | |
| 1900 | + } | |
| 1901 | + }. | |
| 1902 | + | |
| 1903 | + | |
| 1904 | +define String | |
| 1905 | + compute_etag | |
| 1906 | + ( | |
| 1907 | + String filename, | |
| 1908 | + Int32 size, | |
| 1909 | + ) = | |
| 1910 | + if get_file_times(filename) is | |
| 1911 | + { | |
| 1912 | + failure then println("FAILED to get get_file_times() for file '" + filename + "'"); to_ascii(sha1((filename, size))), | |
| 1913 | + success(ftimes) then to_ascii(sha1((filename, ftimes, size))) | |
| 1914 | + }. | |
| 1915 | + | |
| 1916 | +define Bool | |
| 1917 | + are_same_etag | |
| 1918 | + ( | |
| 1919 | + Maybe(String) input_etag, | |
| 1920 | + String current_etag | |
| 1921 | + ) = | |
| 1922 | + if input_etag is | |
| 1923 | + { | |
| 1924 | + failure then println("input etag not found"); false, | |
| 1925 | + success(etag) then println("Input etag = " + etag); etag = current_etag | |
| 1926 | + }. | |
| 1927 | + | |
| 1928 | + Sending the answer line, the headers and the body. | |
| 1929 | + | |
| 1930 | +define One | |
| 1931 | + send_file | |
| 1932 | + ( | |
| 1933 | + Web_Site_Description desc, | |
| 1934 | + Connection connection, | |
| 1935 | + List(HTTP_header) input_headers, | |
| 1936 | + List(HTTP_header) headers, | |
| 1937 | + Int32 size, | |
| 1938 | + Connection file, | |
| 1939 | + String filename, | |
| 1940 | + String full_path, | |
| 1941 | + String mime_type, | |
| 1942 | + One -> One action_before_send_file | |
| 1943 | + ) = | |
| 1944 | + action_before_send_file(unique); | |
| 1945 | + with input_etag = http_header_value(input_headers, "If-None-Match"), | |
| 1946 | + current_etag = compute_etag(full_path, size), | |
| 1947 | + if are_same_etag(input_etag, current_etag) is | |
| 1948 | + { | |
| 1949 | + false then | |
| 1950 | + forget(reliable_write(connection,to_byte_array("HTTP/1.1 200 OK"+crlf))); | |
| 1951 | + forget(reliable_write(connection,[format_headers(headers + headers_for_send_file(mime_type, size, current_etag)) , crlf])); | |
| 1952 | + send_file_body(desc,connection,file,size,0,filename), | |
| 1953 | + true then | |
| 1954 | + forget(reliable_write(connection,to_byte_array("HTTP/1.1 304 Not Modified"+crlf))); | |
| 1955 | + forget(reliable_write(connection,[format_headers([http_header("Etag", current_etag) . headers]) , crlf])) | |
| 1956 | + //send_file_body(desc,connection,file,size,0,filename) | |
| 1957 | + }. | |
| 1958 | + | |
| 1959 | + | |
| 1960 | + | |
| 1961 | + Checking if a connection is under SSL. | |
| 1962 | + | |
| 1963 | +define Bool | |
| 1964 | + is_SSL | |
| 1965 | + ( | |
| 1966 | + Connection c | |
| 1967 | + ) = | |
| 1968 | + if c is | |
| 1969 | + { | |
| 1970 | + file_r(_) then false, | |
| 1971 | + file_w(_) then false, | |
| 1972 | + file_rw(_) then false, | |
| 1973 | + tcp(_) then false, | |
| 1974 | + ssl(_) then true | |
| 1975 | + }. | |
| 1976 | + | |
| 1977 | + | |
| 1978 | + | |
| 1979 | + Before opening and sending a file, we check the MIME type. It must be recognized, | |
| 1980 | + except if there is a valid authorization for private download. | |
| 1981 | + | |
| 1982 | +define One | |
| 1983 | + send_file | |
| 1984 | + ( | |
| 1985 | + Web_Site_Description desc, | |
| 1986 | + Connection connection, | |
| 1987 | + String uri, | |
| 1988 | + List(HTTP_header) input_headers, | |
| 1989 | + Maybe(String) mbauthorization, | |
| 1990 | + One -> One action_before_send_file | |
| 1991 | + ) = | |
| 1992 | + if mbauthorization is | |
| 1993 | + { | |
| 1994 | + //--- file without authorization: take it from public --- | |
| 1995 | + failure then if recognize_mime_type_from_uri(desc,uri) is | |
| 1996 | + { | |
| 1997 | + failure then log_journal_msg(desc,"No MIME type found for '"+uri+"'.\n"), | |
| 1998 | + success(mime_type) then | |
| 1999 | + with path = site_directory(desc)+"/public"+uri, | |
| 2000 | + if (Maybe(RStream))connect to file path is | |
| 2001 | + { | |
| 2002 | + failure then log_journal_msg(desc,"Cannot find file '"+path+"'.\n"), | |
| 2003 | + success(f) then with size = file_size(f), | |
| 2004 | + send_file(desc, | |
| 2005 | + connection, | |
| 2006 | + input_headers, | |
| 2007 | + [], | |
| 2008 | + size, | |
| 2009 | + file(f), | |
| 2010 | + uri, | |
| 2011 | + path, | |
| 2012 | + mime_type, | |
| 2013 | + action_before_send_file) | |
| 2014 | + } | |
| 2015 | + }, | |
| 2016 | + | |
| 2017 | + //--- file with authorization: apply 'private download' mecanism --- | |
| 2018 | + success(authorization) then | |
| 2019 | + with private_download_dir = site_directory(desc)+"/private_download", | |
| 2020 | + if (RetrieveResult(String))retrieve(private_download_dir+"/z"+authorization) | |
| 2021 | + is ok(absolute_path) | |
| 2022 | + then ( | |
| 2023 | + with new_hash = compute_authorization(authorization_secret(desc), | |
| 2024 | + absolute_path), | |
| 2025 | + if (Maybe(RStream))connect to file absolute_path is | |
| 2026 | + { | |
| 2027 | + failure then log_journal_msg(desc,"Cannot find file '"+absolute_path+"'.\n"), | |
| 2028 | + success(f) then with size = file_size(f), | |
| 2029 | + mime_type = if recognize_mime_type_from_uri(desc,uri) is | |
| 2030 | + { | |
| 2031 | + failure then "application/octet-stream" | |
| 2032 | + success(mime_type) then mime_type | |
| 2033 | + }, | |
| 2034 | + send_file(desc, | |
| 2035 | + connection, | |
| 2036 | + input_headers, | |
| 2037 | + [], | |
| 2038 | + size, | |
| 2039 | + file(f), | |
| 2040 | + uri, | |
| 2041 | + absolute_path, | |
| 2042 | + mime_type, | |
| 2043 | + action_before_send_file) | |
| 2044 | + } | |
| 2045 | + ) | |
| 2046 | + else log_journal_msg(desc,"Cannot find or read authorization file.\n") | |
| 2047 | + }. | |
| 2048 | + | |
| 2049 | + | |
| 2050 | + | |
| 2051 | + | |
| 2052 | + | |
| 2053 | + | |
| 2054 | + | |
| 2055 | + | |
| 2056 | + *** [5.6] Answering a www-url encoded request. | |
| 2057 | + | |
| 2058 | + Standard headers are for answering ".awp" requests. | |
| 2059 | + | |
| 2060 | +define List(HTTP_header) | |
| 2061 | + standard_headers | |
| 2062 | + ( | |
| 2063 | + Int32 answer_body_size, | |
| 2064 | + String charset | |
| 2065 | + ) = | |
| 2066 | + [ | |
| 2067 | + //http_header("Content-Type","text/html"), | |
| 2068 | + http_header("Content-Type","text/html; charset="+charset), | |
| 2069 | + http_header("Content-length",integer_to_string(answer_body_size)) | |
| 2070 | + ]. | |
| 2071 | + | |
| 2072 | + | |
| 2073 | +define One | |
| 2074 | + www_url_answer | |
| 2075 | + ( | |
| 2076 | + String host_name, | |
| 2077 | + Web_Site_Description desc, | |
| 2078 | + Connection connection, // with the client | |
| 2079 | + Int32 ip_addr, // of the client | |
| 2080 | + HTTP_RequestLine request_line, | |
| 2081 | + List(HTTP_header) headers, | |
| 2082 | + ByteArray body, | |
| 2083 | + One -> String generate_tt // trust ticket generation | |
| 2084 | + ) = | |
| 2085 | + with all_web_args = query_string(request_line) + | |
| 2086 | + read_www_url_encoded_web_args(to_string(body),0), | |
| 2087 | + uri = uri(request_line), | |
| 2088 | + ext = get_uri_extension(uri), | |
| 2089 | + (if member(journal_extensions(desc),ext) | |
| 2090 | + then log_journal_msg(desc, | |
| 2091 | + format_request(desc,connection,request_line,headers,all_web_args)) | |
| 2092 | + else unique); | |
| 2093 | + if is_illegal_uri(uri,0) | |
| 2094 | + then log_journal_msg(desc,"Received illegal URI: "+uri+"\n") | |
| 2095 | + else (if (ext = ".awp" | ext = "") | |
| 2096 | + then (with answer_headers_body = awp_handler(desc)(host_name, | |
| 2097 | + http_info(ip_addr,uri,headers,generate_tt), | |
| 2098 | + all_web_args, | |
| 2099 | + is_SSL(connection)), | |
| 2100 | + if answer_headers_body is (additional_headers,answer_body) then | |
| 2101 | + forget(reliable_write(connection, | |
| 2102 | + [ "HTTP/1.1 200 OK", crlf, | |
| 2103 | + format_headers(standard_headers(length(answer_body),charset(desc))), | |
| 2104 | + format_headers(additional_headers), | |
| 2105 | + crlf . | |
| 2106 | + answer_body]))) | |
| 2107 | + else (send_file(desc, | |
| 2108 | + connection, | |
| 2109 | + uri, | |
| 2110 | + headers, | |
| 2111 | + if web_arg_value(all_web_args,"zauth") is | |
| 2112 | + { | |
| 2113 | + not_found then failure, | |
| 2114 | + found(v) then success(v) | |
| 2115 | + }, | |
| 2116 | + (One u) |-> before_send_file(desc)(all_web_args)))). | |
| 2117 | + | |
| 2118 | + | |
| 2119 | + | |
| 2120 | + | |
| 2121 | + | |
| 2122 | + | |
| 2123 | + | |
| 2124 | + *** [5.7] Answering a multipart/form-data encoded request. | |
| 2125 | + | |
| 2126 | + In order to support upload of files, we must be able to read web arguments which are | |
| 2127 | + encoded in a multipart/form-data body. The first thing to do is to find the | |
| 2128 | + boundary. The boundary is a special string which delimits the various parts of the | |
| 2129 | + 'multipart' body. It is found within the value of the 'Content-Type' HTTP header, as | |
| 2130 | + the value of the 'boundary' attribute. | |
| 2131 | + | |
| 2132 | + | |
| 2133 | + | |
| 2134 | + | |
| 2135 | + | |
| 2136 | + *** [5.7.1] Finding the boundary. | |
| 2137 | + | |
| 2138 | + Hence, we just have to find the string 'boundary=' within the value of the | |
| 2139 | + 'Content-Type' header, and read the value of the boundary from there. | |
| 2140 | + | |
| 2141 | +define Bool | |
| 2142 | + delimits_boundary | |
| 2143 | + ( | |
| 2144 | + Word8 c | |
| 2145 | + ) = | |
| 2146 | + if c = ' ' then true else | |
| 2147 | + if c = 13 then true else | |
| 2148 | + if c = 10 then true else | |
| 2149 | + if c = 0 then true else | |
| 2150 | + if c = ',' then true else | |
| 2151 | + c = ';'. | |
| 2152 | + | |
| 2153 | + | |
| 2154 | +define Maybe(String) | |
| 2155 | + get_boundary_value_3 | |
| 2156 | + ( | |
| 2157 | + String s, | |
| 2158 | + Int32 i, | |
| 2159 | + List(Word8) so_far | |
| 2160 | + ) = | |
| 2161 | + if nth(i,s) is | |
| 2162 | + { | |
| 2163 | + failure then success(implode(reverse(so_far))), | |
| 2164 | + success(c) then | |
| 2165 | + if delimits_boundary(c) | |
| 2166 | + then success(implode(reverse(so_far))) | |
| 2167 | + else get_boundary_value_3(s,i+1,[c . so_far]) | |
| 2168 | + }. | |
| 2169 | + | |
| 2170 | + | |
| 2171 | + | |
| 2172 | +define Maybe(String) | |
| 2173 | + get_boundary_value_2 | |
| 2174 | + ( | |
| 2175 | + String s, | |
| 2176 | + Int32 i, | |
| 2177 | + ) = | |
| 2178 | + if nth(i,s) is | |
| 2179 | + { | |
| 2180 | + failure then failure, | |
| 2181 | + success(c) then | |
| 2182 | + if is_blank(c) | |
| 2183 | + then get_boundary_value_2(s,i+1) | |
| 2184 | + else get_boundary_value_3(s,i+1,[c]) | |
| 2185 | + }. | |
| 2186 | + | |
| 2187 | +define Maybe(String) | |
| 2188 | + get_boundary_value_1 | |
| 2189 | + ( | |
| 2190 | + String s, // string into which we must find '= ...' | |
| 2191 | + Int32 i // position of start of search | |
| 2192 | + ) = | |
| 2193 | + if nth(i,s) is | |
| 2194 | + { | |
| 2195 | + failure then failure, | |
| 2196 | + success(c) then | |
| 2197 | + if is_blank(c) | |
| 2198 | + then get_boundary_value_1(s,i+1) | |
| 2199 | + else if c = '=' | |
| 2200 | + then get_boundary_value_2(s,i+1) | |
| 2201 | + else failure | |
| 2202 | + }. | |
| 2203 | + | |
| 2204 | + | |
| 2205 | +define Maybe(String) | |
| 2206 | + get_boundary | |
| 2207 | + ( | |
| 2208 | + String content_type_header_value | |
| 2209 | + ) = | |
| 2210 | + if find("boundary",content_type_header_value,0) is | |
| 2211 | + { | |
| 2212 | + failure then failure, | |
| 2213 | + success(n) then // 'boundary' has been found at position n | |
| 2214 | + get_boundary_value_1(content_type_header_value,n+8) | |
| 2215 | + }. | |
| 2216 | + | |
| 2217 | +define Maybe(String) | |
| 2218 | + get_boundary | |
| 2219 | + ( | |
| 2220 | + List(HTTP_header) headers | |
| 2221 | + ) = | |
| 2222 | + if headers is | |
| 2223 | + { | |
| 2224 | + [ ] then failure, | |
| 2225 | + [h . t] then if h is http_header(name,value) then | |
| 2226 | + if name = "content-type" | |
| 2227 | + then get_boundary(value) | |
| 2228 | + else get_boundary(t) | |
| 2229 | + }. | |
| 2230 | + | |
| 2231 | + | |
| 2232 | + | |
| 2233 | + | |
| 2234 | + | |
| 2235 | + | |
| 2236 | + | |
| 2237 | + | |
| 2238 | + *** [5.7.2] Reading attributes from a multipart entity. | |
| 2239 | + | |
| 2240 | + Entities in a multipart/form-data body are separated by instances of the string: | |
| 2241 | + | |
| 2242 | + --bbbbb | |
| 2243 | + | |
| 2244 | + where bbbbb is the boundary computed above. Actually, the body has the form: | |
| 2245 | + | |
| 2246 | + --bbbbb | |
| 2247 | + <entity 1> | |
| 2248 | + --bbbbb | |
| 2249 | + <entity 2> | |
| 2250 | + --bbbbb | |
| 2251 | + ... | |
| 2252 | + --bbbbb | |
| 2253 | + <last entity> | |
| 2254 | + --bbbbb | |
| 2255 | + | |
| 2256 | + | |
| 2257 | + We have to extract an entity which is in the body between offsets 'start' and 'end' | |
| 2258 | + (computed when boundaries have been localized). The entity itself is made of two parts: | |
| 2259 | + headers and body. The body is separated from the headers by a blank line. This blank | |
| 2260 | + line (a double crlf) marks the beginning of the body of the entity. Within the headers | |
| 2261 | + of the entity, we look for a 'Content-Disposition' header, which should look like this: | |
| 2262 | + | |
| 2263 | + Content-Disposition: form-data; name="..."; filename="..." crlf | |
| 2264 | + | |
| 2265 | + We are just interested in the name and the file name. Hence we first search | |
| 2266 | + 'Content-Disposition', then we search 'name' and read the value, and we do the same for | |
| 2267 | + 'filename'. | |
| 2268 | + | |
| 2269 | + If the 'filename' attribute is not present, the web arg is an ordinary one, otherwise, | |
| 2270 | + it is an uploaded file. | |
| 2271 | + | |
| 2272 | + | |
| 2273 | + Below is a variant of 'find' (see 'tools/findstring.anubis'), with an extra 'end' | |
| 2274 | + argument. | |
| 2275 | + | |
| 2276 | +define Maybe(Int32) | |
| 2277 | + find | |
| 2278 | + ( | |
| 2279 | + String what, | |
| 2280 | + ByteArray where, | |
| 2281 | + Int32 start, | |
| 2282 | + Int32 end | |
| 2283 | + ) = | |
| 2284 | + if find(to_byte_array(what),where,start) is | |
| 2285 | + { | |
| 2286 | + failure then failure, | |
| 2287 | + success(n) then | |
| 2288 | + if n+length(what) >= end | |
| 2289 | + then failure | |
| 2290 | + else success(n) | |
| 2291 | + }. | |
| 2292 | + | |
| 2293 | + | |
| 2294 | +define String | |
| 2295 | + read_attribute_value | |
| 2296 | + ( | |
| 2297 | + ByteArray where, | |
| 2298 | + Int32 start, | |
| 2299 | + Int32 end, | |
| 2300 | + List(Word8) so_far | |
| 2301 | + ) = | |
| 2302 | + if start >= end then implode(reverse(so_far)) else | |
| 2303 | + if nth(start,where) is | |
| 2304 | + { | |
| 2305 | + failure then implode(reverse(so_far)), | |
| 2306 | + success(c) then | |
| 2307 | + if c = '\"' | |
| 2308 | + then implode(reverse(so_far)) | |
| 2309 | + else read_attribute_value(where,start+1,end,[c . so_far]) | |
| 2310 | + }. | |
| 2311 | + | |
| 2312 | +define Maybe(String) | |
| 2313 | + find_attribute | |
| 2314 | + ( | |
| 2315 | + String name, | |
| 2316 | + ByteArray where, | |
| 2317 | + Int32 start, | |
| 2318 | + Int32 end | |
| 2319 | + ) = | |
| 2320 | + with name = name+"=\"", | |
| 2321 | + if find(to_byte_array(name),where,start) is | |
| 2322 | + { | |
| 2323 | + failure then failure, | |
| 2324 | + success(n) then | |
| 2325 | + if n+length(name) >= end | |
| 2326 | + then failure | |
| 2327 | + else success(read_attribute_value(where,n+length(name),end,[])) | |
| 2328 | + }. | |
| 2329 | + | |
| 2330 | + | |
| 2331 | + | |
| 2332 | +define Maybe((String,Maybe(String))) | |
| 2333 | + find_name_and_filename | |
| 2334 | + ( | |
| 2335 | + ByteArray body, | |
| 2336 | + Int32 start, | |
| 2337 | + Int32 end | |
| 2338 | + ) = | |
| 2339 | + if find(to_byte_array("Content-Disposition"),body,start) is | |
| 2340 | + { | |
| 2341 | + failure then failure, | |
| 2342 | + success(n) then | |
| 2343 | + if find_attribute("name",body,n+19,end) is | |
| 2344 | + { | |
| 2345 | + failure then failure, | |
| 2346 | + success(name_value) then if find_attribute("filename",body,n+19,end) is | |
| 2347 | + { | |
| 2348 | + failure then success((name_value,failure)), | |
| 2349 | + success(filename_value) then success((name_value,success(filename_value))) | |
| 2350 | + } | |
| 2351 | + } | |
| 2352 | + }. | |
| 2353 | + | |
| 2354 | + | |
| 2355 | + | |
| 2356 | + | |
| 2357 | + | |
| 2358 | + | |
| 2359 | + | |
| 2360 | + | |
| 2361 | + | |
| 2362 | + | |
| 2363 | + *** [5.7.3] Creating a temporary filename for an uploaded file. | |
| 2364 | + | |
| 2365 | +variable Int32 uploaded_file_count = 0. | |
| 2366 | + | |
| 2367 | + This variable is local to the virtual machine. Hence, its value is 0 each time a new | |
| 2368 | + requests arrives. Temporary uploaded files are stored in the directory represented by | |
| 2369 | + 'upload_temporary_directory'. The filenames have the form: | |
| 2370 | + | |
| 2371 | + _m_n | |
| 2372 | + | |
| 2373 | + where 'm' is the number of the virtual machine, and 'n' a number obtained by | |
| 2374 | + incrementing 'uploaded_file_count'. Notice that the program must do something with this | |
| 2375 | + file (move it to some directory/name), otherwise, it will probably be overwritten the | |
| 2376 | + next time the same machine works. | |
| 2377 | + | |
| 2378 | + | |
| 2379 | + | |
| 2380 | + | |
| 2381 | + | |
| 2382 | + | |
| 2383 | + *** [5.7.4] Saving an uploaded file under a temporary filename. | |
| 2384 | + | |
| 2385 | +define Maybe(String) // returns the temporary file name | |
| 2386 | + save_uploaded_file | |
| 2387 | + ( | |
| 2388 | + Web_Site_Description desc, | |
| 2389 | + ByteArray body, | |
| 2390 | + Int32 start, | |
| 2391 | + Int32 end | |
| 2392 | + ) = | |
| 2393 | + uploaded_file_count <- 1 + *uploaded_file_count; | |
| 2394 | + with tfn = "_"+integer_to_string(virtual_machine_id)+"_"+integer_to_string(*uploaded_file_count), | |
| 2395 | + if (Maybe(WStream))connect to file site_directory(desc)+"/upload_temporary/"+tfn is | |
| 2396 | + { | |
| 2397 | + failure then failure, | |
| 2398 | + success(f) then | |
| 2399 | + if reliable_write(file(f),extract(body,start,end)) is | |
| 2400 | + { | |
| 2401 | + failure then failure, | |
| 2402 | + success(nw) then | |
| 2403 | + if nw = end - start | |
| 2404 | + then success(tfn) | |
| 2405 | + else failure | |
| 2406 | + } | |
| 2407 | + }. | |
| 2408 | + | |
| 2409 | + | |
| 2410 | + | |
| 2411 | + | |
| 2412 | + | |
| 2413 | + | |
| 2414 | + | |
| 2415 | + | |
| 2416 | + *** [5.7.5] Removing the path from a file name. | |
| 2417 | + | |
| 2418 | + When a file is uploaded, the browser sends the complete path of the file on the client | |
| 2419 | + machine as the file name. Actually, this is not quite normal. Nevertheless, we need to | |
| 2420 | + remove the path, and keep only the file name. This is achieved by 'remove_path' below. | |
| 2421 | + | |
| 2422 | +define Int32 | |
| 2423 | + file_name_begin | |
| 2424 | + ( | |
| 2425 | + String full_name, | |
| 2426 | + Int32 i | |
| 2427 | + ) = | |
| 2428 | + if nth(i,full_name) is | |
| 2429 | + { | |
| 2430 | + failure then 0, | |
| 2431 | + success(c) then | |
| 2432 | + if c = '/' then i+1 else | |
| 2433 | + if c = '\\' then i+1 else | |
| 2434 | + file_name_begin(full_name,i-1) | |
| 2435 | + }. | |
| 2436 | + | |
| 2437 | +define String | |
| 2438 | + remove_path | |
| 2439 | + ( | |
| 2440 | + String full_name | |
| 2441 | + ) = | |
| 2442 | + with l = length(full_name), | |
| 2443 | + b = file_name_begin(full_name,l-1), | |
| 2444 | + substr(full_name,b,l-b). | |
| 2445 | + | |
| 2446 | + | |
| 2447 | + | |
| 2448 | + | |
| 2449 | + | |
| 2450 | + *** [5.7.6] Reading a multipart entity. | |
| 2451 | + | |
| 2452 | +define Maybe(Web_arg) | |
| 2453 | + get_multipart_entity | |
| 2454 | + ( | |
| 2455 | + Web_Site_Description desc, | |
| 2456 | + ByteArray body, | |
| 2457 | + Int32 start, | |
| 2458 | + Int32 end | |
| 2459 | + ) = | |
| 2460 | + if find(to_byte_array(crlf+crlf),body,start) is | |
| 2461 | + { | |
| 2462 | + failure then failure, | |
| 2463 | + success(k) then | |
| 2464 | + if k >= end // must be within this entity, not the next one | |
| 2465 | + then failure | |
| 2466 | + else if find_name_and_filename(body,start,k) is | |
| 2467 | + { | |
| 2468 | + failure then failure, | |
| 2469 | + success(n_mbfn) then if n_mbfn is (name,mbfn) then | |
| 2470 | + if mbfn is | |
| 2471 | + { | |
| 2472 | + failure then | |
| 2473 | + success(web_arg(name,to_string(extract(body,k+4,end-2)))), | |
| 2474 | + // we must substract 2 to end because of crlf just before the boundary | |
| 2475 | + | |
| 2476 | + success(fn) then | |
| 2477 | + if save_uploaded_file(desc,body,k+4,end-2) is | |
| 2478 | + { | |
| 2479 | + failure then failure, | |
| 2480 | + success(tfn) then | |
| 2481 | + success(upload(name,remove_path(fn), | |
| 2482 | + site_directory(desc)+"/upload_temporary/"+tfn)) | |
| 2483 | + | |
| 2484 | + } | |
| 2485 | + } | |
| 2486 | + } | |
| 2487 | + }. | |
| 2488 | + | |
| 2489 | + | |
| 2490 | + | |
| 2491 | +define List(Web_arg) | |
| 2492 | + read_multipart_form_data_encoded_web_args | |
| 2493 | + ( | |
| 2494 | + Web_Site_Description desc, | |
| 2495 | + ByteArray body, | |
| 2496 | + ByteArray __boundary, | |
| 2497 | + Int32 i, | |
| 2498 | + ) = | |
| 2499 | + if find(__boundary,body,i) is | |
| 2500 | + { | |
| 2501 | + failure then [ ], | |
| 2502 | + success(n) then | |
| 2503 | + if find(__boundary,body,n+length(__boundary)) is | |
| 2504 | + { | |
| 2505 | + failure then [ ], | |
| 2506 | + success(m) then | |
| 2507 | + if get_multipart_entity(desc,body,n+length(__boundary),m) is | |
| 2508 | + { | |
| 2509 | + failure then [ ], | |
| 2510 | + success(wa) then | |
| 2511 | + [wa . read_multipart_form_data_encoded_web_args(desc,body,__boundary,m)] | |
| 2512 | + } | |
| 2513 | + } | |
| 2514 | + }. | |
| 2515 | + | |
| 2516 | + | |
| 2517 | + | |
| 2518 | +define One | |
| 2519 | + multipart_form_data_answer | |
| 2520 | + ( | |
| 2521 | + String host_name, | |
| 2522 | + Web_Site_Description desc, | |
| 2523 | + Connection connection, | |
| 2524 | + Int32 ip_addr, | |
| 2525 | + HTTP_RequestLine request_line, | |
| 2526 | + List(HTTP_header) headers, | |
| 2527 | + ByteArray body, | |
| 2528 | + One -> String generate_tt | |
| 2529 | + ) = | |
| 2530 | + if get_boundary(headers) is | |
| 2531 | + { | |
| 2532 | + failure then unique, | |
| 2533 | + success(boundary) then | |
| 2534 | + with all_web_args = query_string(request_line) + | |
| 2535 | + read_multipart_form_data_encoded_web_args(desc, | |
| 2536 | + body, | |
| 2537 | + to_byte_array("--"+boundary), | |
| 2538 | + 0), | |
| 2539 | + uri = uri(request_line), | |
| 2540 | + ext = get_uri_extension(uri), | |
| 2541 | + log_journal_msg(desc, | |
| 2542 | + format_request(desc,connection,request_line,headers,all_web_args)); | |
| 2543 | + if is_illegal_uri(uri,0) | |
| 2544 | + then log_journal_msg(desc,"Received illegal URI: "+uri+"\n") | |
| 2545 | + else | |
| 2546 | + if (ext = ".awp" | ext = "") then | |
| 2547 | + (with answer_headers_body = awp_handler(desc)(host_name, | |
| 2548 | + http_info(ip_addr,uri,headers,generate_tt), | |
| 2549 | + all_web_args, | |
| 2550 | + is_SSL(connection)), | |
| 2551 | + if answer_headers_body is (additional_headers,answer_body) then | |
| 2552 | + forget(reliable_write(connection, | |
| 2553 | + [ "HTTP/1.1 200 OK",crlf, | |
| 2554 | + format_headers(standard_headers(length(answer_body),charset(desc))), | |
| 2555 | + format_headers(additional_headers), | |
| 2556 | + crlf . | |
| 2557 | + answer_body]))) | |
| 2558 | + else unique | |
| 2559 | + }. | |
| 2560 | + | |
| 2561 | + | |
| 2562 | + | |
| 2563 | + | |
| 2564 | + | |
| 2565 | + | |
| 2566 | + | |
| 2567 | + | |
| 2568 | + *** [5.8] Handling redirections. | |
| 2569 | + | |
| 2570 | + 'redirections' (of type 'List(Redirection)') contains redirection directives. Each one | |
| 2571 | + has the form: | |
| 2572 | + | |
| 2573 | + redirect(required_uri,required_host,corresponding_uri). | |
| 2574 | + | |
| 2575 | + The host required by the client may be found in the 'Host' HTTP header. The URI | |
| 2576 | + required by the client is given below as 'uri'. We just have to find the required host | |
| 2577 | + in the headers, and to find the corresponding redirection directive. | |
| 2578 | + | |
| 2579 | + | |
| 2580 | + In the next fonction, the required host and URI are known. We just have to search in | |
| 2581 | + the 'redirections' list. | |
| 2582 | + | |
| 2583 | +define String | |
| 2584 | + handle_redirection | |
| 2585 | + ( | |
| 2586 | + String required_uri, | |
| 2587 | + String required_host, | |
| 2588 | + List(Redirection) redirections | |
| 2589 | + ) = | |
| 2590 | + if redirections is | |
| 2591 | + { | |
| 2592 | + [ ] then required_uri, | |
| 2593 | + [h . t] then if h is redirect(uri,host,target) then | |
| 2594 | + if host = required_host | |
| 2595 | + then if uri = required_uri | |
| 2596 | + then target | |
| 2597 | + else handle_redirection(required_uri,required_host,t) | |
| 2598 | + else handle_redirection(required_uri,required_host,t) | |
| 2599 | + }. | |
| 2600 | + | |
| 2601 | + | |
| 2602 | + | |
| 2603 | + The host name may be encumbered by a port number, like | |
| 2604 | + | |
| 2605 | + www.our-business.com:1607 | |
| 2606 | + | |
| 2607 | + We must remove this port number, otherwise the host name may not be recognized. | |
| 2608 | + | |
| 2609 | +define String | |
| 2610 | + strip_port | |
| 2611 | + ( | |
| 2612 | + String name, | |
| 2613 | + Int32 i | |
| 2614 | + ) = | |
| 2615 | + if nth(i,name) is | |
| 2616 | + { | |
| 2617 | + failure then name, | |
| 2618 | + success(c) then | |
| 2619 | + if c = ':' | |
| 2620 | + then substr(name,0,i) | |
| 2621 | + else strip_port(name,i+1) | |
| 2622 | + }. | |
| 2623 | + | |
| 2624 | + | |
| 2625 | + | |
| 2626 | + | |
| 2627 | + | |
| 2628 | + Finding the 'Host' header. No redirection is performed if this header is not found. | |
| 2629 | + | |
| 2630 | +define String | |
| 2631 | + handle_redirection // returns the redirected URI | |
| 2632 | + ( | |
| 2633 | + List(Redirection) redirections, | |
| 2634 | + String uri, // original URI | |
| 2635 | + List(HTTP_header) headers | |
| 2636 | + ) = | |
| 2637 | + if headers is | |
| 2638 | + { | |
| 2639 | + [ ] then uri, | |
| 2640 | + [h . t] then if h is http_header(name,value) then | |
| 2641 | + if name = "host" | |
| 2642 | + then handle_redirection(uri,strip_port(value,0),redirections) | |
| 2643 | + else handle_redirection(redirections,uri,t) | |
| 2644 | + }. | |
| 2645 | + | |
| 2646 | + | |
| 2647 | + | |
| 2648 | + | |
| 2649 | + | |
| 2650 | + | |
| 2651 | + | |
| 2652 | + | |
| 2653 | + *** [5.9] Answering both sorts of requests. | |
| 2654 | + | |
| 2655 | + We must decide if the request is www-url encoded or multipart/form-data encoded. This | |
| 2656 | + is achieved through the header 'Content-Type'. | |
| 2657 | + | |
| 2658 | +define EncodingType | |
| 2659 | + get_encoding_type | |
| 2660 | + ( | |
| 2661 | + List(HTTP_header) headers | |
| 2662 | + ) = | |
| 2663 | + if headers is | |
| 2664 | + { | |
| 2665 | + [ ] then www_url, // this is the default | |
| 2666 | + [h . t] then if h is http_header(name,value) then | |
| 2667 | + if name = "content-type" | |
| 2668 | + then if find("multipart/form-data",value,0) is | |
| 2669 | + { | |
| 2670 | + failure then www_url, | |
| 2671 | + success(_) then multipart_form_data | |
| 2672 | + } | |
| 2673 | + else get_encoding_type(t) | |
| 2674 | + }. | |
| 2675 | + | |
| 2676 | + | |
| 2677 | + | |
| 2678 | +define One | |
| 2679 | + send_answer | |
| 2680 | + ( | |
| 2681 | + String host_name, | |
| 2682 | + Web_Site_Description desc, | |
| 2683 | + Connection connection, | |
| 2684 | + HTTP_RequestLine rqline, | |
| 2685 | + List(HTTP_header) headers, | |
| 2686 | + ByteArray body, | |
| 2687 | + One -> String generate_tt | |
| 2688 | + ) = | |
| 2689 | + if rqline is request_line(type,uri,qstring) then | |
| 2690 | + with rqline = request_line(type,handle_redirection(redirections(desc),uri,headers),qstring), | |
| 2691 | + if remote_IP_address_and_port(connection) is (ip_addr,_) then | |
| 2692 | + if get_encoding_type(headers) is | |
| 2693 | + { | |
| 2694 | + www_url then | |
| 2695 | + www_url_answer(host_name,desc,connection,ip_addr,rqline,headers,body,generate_tt), | |
| 2696 | + multipart_form_data then | |
| 2697 | + multipart_form_data_answer(host_name,desc,connection,ip_addr,rqline,headers,body,generate_tt) | |
| 2698 | + }. | |
| 2699 | + | |
| 2700 | + | |
| 2701 | + | |
| 2702 | + | |
| 2703 | + | |
| 2704 | + | |
| 2705 | + | |
| 2706 | + *** [6] The HTTP/HTTPS server. | |
| 2707 | + | |
| 2708 | + The command 'start_server' (declared in 'predefined.anubis') starts a virtual machine | |
| 2709 | + which opens a server TCP/IP connection, and which continuously listens to this | |
| 2710 | + connection. When a request arrives, this machine delegates the work of deciphering and | |
| 2711 | + answering the request to another virtual machine, and continues to listen. The job of | |
| 2712 | + the delegated machine is defined by the HTTP request handler below. | |
| 2713 | + | |
| 2714 | + | |
| 2715 | + | |
| 2716 | + | |
| 2717 | + | |
| 2718 | + *** [6.1] Determining the requested host. | |
| 2719 | + | |
| 2720 | + When a request arrives to one of our two servers, we must decide which site (host) is | |
| 2721 | + requested. | |
| 2722 | + | |
| 2723 | +define Maybe(String) | |
| 2724 | + get_host_header_value | |
| 2725 | + ( | |
| 2726 | + List(HTTP_header) headers | |
| 2727 | + ) = | |
| 2728 | + if headers is | |
| 2729 | + { | |
| 2730 | + [ ] then failure, | |
| 2731 | + [h . t] then if h is http_header(name,value) then | |
| 2732 | + if name = "host" | |
| 2733 | + then success(strip_port(value,0)) | |
| 2734 | + else get_host_header_value(t) | |
| 2735 | + }. | |
| 2736 | + | |
| 2737 | +define Maybe((String,Web_Site_Description)) | |
| 2738 | + get_site | |
| 2739 | + ( | |
| 2740 | + String requested_host, | |
| 2741 | + List(Web_Site_Description) sites | |
| 2742 | + ) = | |
| 2743 | + if sites is | |
| 2744 | + { | |
| 2745 | + [ ] then print("Requested host '"+requested_host+"' does not exist.\n"); failure, | |
| 2746 | + [site1 . others] then | |
| 2747 | + if site1 is web_site_description(common_names,_,_,_,_,_,_,_,_,_) then | |
| 2748 | + if member(common_names,requested_host) | |
| 2749 | + then success((requested_host,site1)) | |
| 2750 | + else get_site(requested_host,others) | |
| 2751 | + }. | |
| 2752 | + | |
| 2753 | + | |
| 2754 | +define Maybe((String,Web_Site_Description)) | |
| 2755 | + get_site | |
| 2756 | + ( | |
| 2757 | + List(HTTP_header) headers, | |
| 2758 | + List(Web_Site_Description) sites | |
| 2759 | + ) = | |
| 2760 | + if get_host_header_value(headers) is | |
| 2761 | + { | |
| 2762 | + failure then print("No 'Host' HTTP header.\n"); failure, | |
| 2763 | + success(requested_host) then | |
| 2764 | + //here we treat the case with only one site. hence we accept any host request | |
| 2765 | + //print("*** there is " +length(sites) + " sites \n"); | |
| 2766 | + if length(sites) = 1 then | |
| 2767 | + with site = force_nth(0, sites), | |
| 2768 | + //print("ONE server OK\n"); | |
| 2769 | + success((requested_host, site)) | |
| 2770 | + else | |
| 2771 | + get_site(requested_host,sites) | |
| 2772 | + }. | |
| 2773 | + | |
| 2774 | + | |
| 2775 | + | |
| 2776 | + | |
| 2777 | + | |
| 2778 | + *** [6.2] The HTTP request handler. | |
| 2779 | + | |
| 2780 | + Here is the HTTP/HTTPS handler. It is called at each new request in a separate virtual | |
| 2781 | + machine. It reads the headers of the HTTP request, determines the host, determines body | |
| 2782 | + size, reads the body of the HTTP request, and answers the request. | |
| 2783 | + | |
| 2784 | + | |
| 2785 | + | |
| 2786 | +define One -> String make_generate_trust_ticket(DenialOfService dos). | |
| 2787 | + | |
| 2788 | + | |
| 2789 | +define One | |
| 2790 | + http_https_handler | |
| 2791 | + ( | |
| 2792 | + List(Web_Site_Description) sites, | |
| 2793 | + Connection connection, | |
| 2794 | + Bool is_https, | |
| 2795 | + DenialOfService dos | |
| 2796 | + ) = | |
| 2797 | + with start_time = (Int32)now, | |
| 2798 | + sttm <- start_time; | |
| 2799 | + if dos is denial_of_service(mc_v,rld_v,hd_v,ad_v,ld_v,ra_v) then | |
| 2800 | + if remote_IP_address_and_port(connection) is (ip_addr,port) then | |
| 2801 | + if read_request_line(connection,start_time+*rld_v,dos) is | |
| 2802 | + { | |
| 2803 | + error(msg) then print(format(msg)), | |
| 2804 | + ok(request_line) then | |
| 2805 | + if read_http_headers(connection,start_time+*hd_v,dos) is | |
| 2806 | + { | |
| 2807 | + error(msg) then print(format(msg)), | |
| 2808 | + ok(headers) then if get_site(headers,sites) is | |
| 2809 | + { | |
| 2810 | + failure then unique, | |
| 2811 | + success(p) then if p is (host_name,desc) then | |
| 2812 | + if get_body_size(headers) is | |
| 2813 | + { | |
| 2814 | + error(msg) then log_journal_msg(desc,format(msg)), | |
| 2815 | + ok(body_size) then | |
| 2816 | + if read_http_body(connection,body_size,constant_byte_array(0,0),1000) is | |
| 2817 | + { | |
| 2818 | + error(msg) then log_journal_msg(desc,format(msg)), | |
| 2819 | + ok(body) then | |
| 2820 | + send_answer(host_name,desc,connection,request_line,headers,body, | |
| 2821 | + make_generate_trust_ticket(dos)) | |
| 2822 | + } | |
| 2823 | + } | |
| 2824 | + } | |
| 2825 | + } | |
| 2826 | + }. | |
| 2827 | + | |
| 2828 | + | |
| 2829 | + Below are the two tools for constructing the handlers required by 'start_server' and | |
| 2830 | + 'start_ssl_server' (see 'predefined.anubis'). | |
| 2831 | + | |
| 2832 | +define Bool is_dubious_IP(Int32 ip, DenialOfService dos). | |
| 2833 | + | |
| 2834 | +define Server -> ((RWStream) -> One) | |
| 2835 | + make_http_handler | |
| 2836 | + ( | |
| 2837 | + List(Web_Site_Description) sites, | |
| 2838 | + DenialOfService dos | |
| 2839 | + ) = | |
| 2840 | + (Server server) |-> (RWStream connection) |-> | |
| 2841 | + if remote_IP_address_and_port(connection) is (addr,_) then | |
| 2842 | + if is_dubious_IP(addr,dos) | |
| 2843 | + then print("Rejecting dubious IP address "+ip_addr_to_string(addr)+"\n") | |
| 2844 | + else http_https_handler(sites,tcp(connection),false,dos). | |
| 2845 | + | |
| 2846 | +define Server -> (SSL_Connection -> One) | |
| 2847 | + make_https_handler | |
| 2848 | + ( | |
| 2849 | + List(Web_Site_Description) sites, | |
| 2850 | + DenialOfService dos | |
| 2851 | + ) = | |
| 2852 | + (Server server) |-> (SSL_Connection connection) |-> | |
| 2853 | + http_https_handler(sites,ssl(connection),true,dos). | |
| 2854 | + | |
| 2855 | + | |
| 2856 | + | |
| 2857 | + | |
| 2858 | + *** [6.3] Server's tasks. | |
| 2859 | + | |
| 2860 | + Some tasks must be executed periodically, for example for cleaning up directories from | |
| 2861 | + short life time files. | |
| 2862 | + | |
| 2863 | + The next function removes from the given directory (and recursively from its | |
| 2864 | + subdirectories) all the files which are more than 10 minutes old. | |
| 2865 | + | |
| 2866 | +define One | |
| 2867 | + cleanup_directory_10mn | |
| 2868 | + ( | |
| 2869 | + String dir // path of private download directory (or subdirectory) with trailing slash | |
| 2870 | + ) = | |
| 2871 | + forget(map((FileDescription fd) |-> if fd is | |
| 2872 | + { | |
| 2873 | + no_info(name) then forget(remove(dir+name)), | |
| 2874 | + file(name,_,_,d) then if d+600 < now then forget(remove(dir+name)) else unique, | |
| 2875 | + link(name,_,_,d) then if d+600 < now then forget(remove(dir+name)) else unique, | |
| 2876 | + directory(name,_,_) then cleanup_directory_10mn(dir+name+"/"), | |
| 2877 | + }, | |
| 2878 | + directory_full_list(dir,"*","*","*"))). | |
| 2879 | + | |
| 2880 | + | |
| 2881 | +define One | |
| 2882 | + http_servers_tasks | |
| 2883 | + ( | |
| 2884 | + List(Web_Site_Description) sites, | |
| 2885 | + List(Server) servers, | |
| 2886 | + Int32 period, | |
| 2887 | + Int32 next_time, | |
| 2888 | + ) = | |
| 2889 | + if mapand(is_down,servers) | |
| 2890 | + then unique | |
| 2891 | + else if now > next_time | |
| 2892 | + then | |
| 2893 | + ( | |
| 2894 | + /* | |
| 2895 | + forget(map((Web_Site_Description wsd) |-> | |
| 2896 | + cleanup_directory_10mn(site_directory(wsd)+"/private_download/"), | |
| 2897 | + sites)); | |
| 2898 | + */ | |
| 2899 | + http_servers_tasks(sites,servers,period,next_time+period) | |
| 2900 | + ) | |
| 2901 | + else | |
| 2902 | + ( | |
| 2903 | + sleep(1000); | |
| 2904 | + http_servers_tasks(sites,servers,period,next_time) | |
| 2905 | + ). | |
| 2906 | + | |
| 2907 | + | |
| 2908 | +public define One | |
| 2909 | + start_http_servers_tasks | |
| 2910 | + ( | |
| 2911 | + List(Web_Site_Description) sites, | |
| 2912 | + List(Server) servers, | |
| 2913 | + Int32 period | |
| 2914 | + ) = | |
| 2915 | + delegate http_servers_tasks(sites,servers,period,now), | |
| 2916 | + unique. | |
| 2917 | + | |
| 2918 | + | |
| 2919 | + | |
| 2920 | + | |
| 2921 | + *** [6.4] Protection against 'denial of service' attacks. | |
| 2922 | + | |
| 2923 | + | |
| 2924 | + *** [6.4.1] Counting connections. | |
| 2925 | + | |
| 2926 | +define Bool // returns false if the counter cannot be incremented (too many connections) | |
| 2927 | + increment_connections_counter | |
| 2928 | + ( | |
| 2929 | + Var(Int32) counter | |
| 2930 | + ) = | |
| 2931 | + protect with n = *counter, | |
| 2932 | + if n >= 100 | |
| 2933 | + then false | |
| 2934 | + else (counter <- (*counter)+1); true. | |
| 2935 | + | |
| 2936 | +define One | |
| 2937 | + decrement_connections_counter | |
| 2938 | + ( | |
| 2939 | + Var(Int32) counter | |
| 2940 | + ) = | |
| 2941 | + protect counter <- (*counter)-1. | |
| 2942 | + | |
| 2943 | + | |
| 2944 | + | |
| 2945 | + | |
| 2946 | + | |
| 2947 | + *** [6.4.2] Recording dubious IP addresses. | |
| 2948 | + | |
| 2949 | + | |
| 2950 | +define List(DubiousIP) | |
| 2951 | + record_dubious_IP | |
| 2952 | + ( | |
| 2953 | + Int32 ip, | |
| 2954 | + List(DubiousIP) l | |
| 2955 | + ) = | |
| 2956 | + if l is | |
| 2957 | + { | |
| 2958 | + [ ] then [dubious_ip(ip,now)], | |
| 2959 | + [h . t] then if h is dubious_ip(addr,time) then | |
| 2960 | + if addr = ip | |
| 2961 | + then [dubious_ip(addr,now) . t] | |
| 2962 | + else [h . record_dubious_IP(ip,t)] | |
| 2963 | + }. | |
| 2964 | + | |
| 2965 | + | |
| 2966 | +define One | |
| 2967 | + record_dubious_IP | |
| 2968 | + ( | |
| 2969 | + Int32 dubious_IP, | |
| 2970 | + Var(List(DubiousIP)) v | |
| 2971 | + ) = | |
| 2972 | + protect v <- record_dubious_IP(dubious_IP,*v). | |
| 2973 | + | |
| 2974 | + | |
| 2975 | +define One | |
| 2976 | + record_dubious_IP | |
| 2977 | + ( | |
| 2978 | + Int32 addr, | |
| 2979 | + DenialOfService dos | |
| 2980 | + ) = | |
| 2981 | + record_dubious_IP(addr,list_of_dubious(dos)). | |
| 2982 | + | |
| 2983 | + | |
| 2984 | +public define DenialOfService | |
| 2985 | + load_denial_of_service_info | |
| 2986 | + = | |
| 2987 | + if (RetrieveResult(DenialOfService))retrieve(my_anubis_directory+"/web_sites/dos_info") is | |
| 2988 | + ok(dos) then dos else denial_of_service( | |
| 2989 | + var(100), | |
| 2990 | + var(1000), | |
| 2991 | + var(1500), | |
| 2992 | + var(2000), | |
| 2993 | + var([]), | |
| 2994 | + var([])). | |
| 2995 | + | |
| 2996 | + | |
| 2997 | + | |
| 2998 | + | |
| 2999 | + *** [6.4.3] Testing if an address is dubious. | |
| 3000 | + | |
| 3001 | +define Bool | |
| 3002 | + is_dubious_IP | |
| 3003 | + ( | |
| 3004 | + Int32 ip, | |
| 3005 | + List(DubiousIP) l | |
| 3006 | + ) = | |
| 3007 | + if l is | |
| 3008 | + { | |
| 3009 | + [ ] then false, | |
| 3010 | + [h . t] then if h is dubious_ip(addr,time) then | |
| 3011 | + if ip = addr | |
| 3012 | + then true | |
| 3013 | + else is_dubious_IP(ip,t) | |
| 3014 | + }. | |
| 3015 | + | |
| 3016 | + | |
| 3017 | +define Bool | |
| 3018 | + is_dubious_IP | |
| 3019 | + ( | |
| 3020 | + Int32 ip, | |
| 3021 | + DenialOfService dos | |
| 3022 | + ) = | |
| 3023 | + if dos is | |
| 3024 | + { | |
| 3025 | + denial_of_service(mc_v,rld_v,hd_v,ad_v,ld_v,ra_v) then | |
| 3026 | + if member(*ra_v,ip) then false else | |
| 3027 | + is_dubious_IP(ip,*ld_v) | |
| 3028 | + }. | |
| 3029 | + | |
| 3030 | + | |
| 3031 | + | |
| 3032 | + | |
| 3033 | + *** [6.4.4] Removing inactive dubious IP addresses. | |
| 3034 | + | |
| 3035 | +define List(DubiousIP) | |
| 3036 | + remove_inactive_dubious_IP | |
| 3037 | + ( | |
| 3038 | + List(DubiousIP) l, | |
| 3039 | + Int32 ref_time, | |
| 3040 | + ) = | |
| 3041 | + if l is | |
| 3042 | + { | |
| 3043 | + [ ] then [ ], | |
| 3044 | + [h . t] then if h is dubious_ip(addr,time) then | |
| 3045 | + if time < ref_time | |
| 3046 | + then (print(ip_addr_to_string(addr)+" removed from dubious addresses list.\n"); | |
| 3047 | + remove_inactive_dubious_IP(t,ref_time)) | |
| 3048 | + else [h . remove_inactive_dubious_IP(t,ref_time)] | |
| 3049 | + }. | |
| 3050 | + | |
| 3051 | +define One | |
| 3052 | + remove_inactive_dubious_IP | |
| 3053 | + ( | |
| 3054 | + Var(List(DubiousIP)) v | |
| 3055 | + ) = | |
| 3056 | + protect | |
| 3057 | + with ref_time = now - 600, // 10 minutes | |
| 3058 | + v <- remove_inactive_dubious_IP(*v,ref_time). | |
| 3059 | + | |
| 3060 | + | |
| 3061 | + The above function will be executed periodically by the servers's tasks machine. | |
| 3062 | + | |
| 3063 | + | |
| 3064 | + | |
| 3065 | + *** [6.4.5] Making the function for generating trust tickets. | |
| 3066 | + | |
| 3067 | +define One -> String | |
| 3068 | + make_generate_trust_ticket | |
| 3069 | + ( | |
| 3070 | + DenialOfService dos | |
| 3071 | + ) = | |
| 3072 | + (One _) |-> "". | |
| 3073 | + | |
| 3074 | + | |
| 3075 | + | |
| 3076 | + | |
| 3077 | + | |
| 3078 | + | |
| 3079 | + | |
| 3080 | + *** [6.5] Starting the HTTP/HTTPS server. | |
| 3081 | + | |
| 3082 | + | |
| 3083 | + The next function creates the directories for all sites (if they don't already exist). | |
| 3084 | + | |
| 3085 | +define One | |
| 3086 | + create_directories | |
| 3087 | + ( | |
| 3088 | + List(Web_Site_Description) sites | |
| 3089 | + ) = | |
| 3090 | + if sites is | |
| 3091 | + { | |
| 3092 | + [ ] then unique, | |
| 3093 | + [s1 . others] then | |
| 3094 | + with site_dir = site_directory(s1), | |
| 3095 | + forget(make_directory(site_dir+"/public",default_directory_mode)); | |
| 3096 | + forget(make_directory(site_dir+"/upload_temporary",default_directory_mode)); | |
| 3097 | + forget(make_directory(site_dir+"/private_download",default_directory_mode)); | |
| 3098 | + forget(make_directory(site_dir+"/journal",default_directory_mode)); | |
| 3099 | + create_directories(others) | |
| 3100 | + }. | |
| 3101 | + | |
| 3102 | + | |
| 3103 | + | |
| 3104 | + | |
| 3105 | + | |
| 3106 | + Below are the commands for starting an HTTP server and an HTTPS server. | |
| 3107 | + | |
| 3108 | + | |
| 3109 | +define StartServerResult | |
| 3110 | + start_http_server | |
| 3111 | + ( | |
| 3112 | + Int32 ip_address, | |
| 3113 | + Int32 port, | |
| 3114 | + Server -> ((RWStream) -> One) handler, | |
| 3115 | + Int32 retries, | |
| 3116 | + DenialOfService dos | |
| 3117 | + ) = | |
| 3118 | + if start_server(ip_address, | |
| 3119 | + port, | |
| 3120 | + handler, | |
| 3121 | + identity) is ok(server) | |
| 3122 | + then print(" \r"); | |
| 3123 | + ok(server) | |
| 3124 | + else print("Port "+port+": retry number "+retries+"\r"); | |
| 3125 | + sleep(1000); | |
| 3126 | + start_http_server(ip_address,port,handler,retries+1,dos). | |
| 3127 | + | |
| 3128 | +public define StartServerResult | |
| 3129 | + start_http_server | |
| 3130 | + ( | |
| 3131 | + Int32 ip_address, | |
| 3132 | + Int32 port, | |
| 3133 | + List(Web_Site_Description) sites, | |
| 3134 | + DenialOfService dos | |
| 3135 | + ) = | |
| 3136 | + create_directories(sites); | |
| 3137 | + start_http_server(ip_address,port, | |
| 3138 | + make_http_handler(sites,dos), | |
| 3139 | + 0, | |
| 3140 | + dos). | |
| 3141 | + | |
| 3142 | + | |
| 3143 | + For the HTTPS server, we have a problem which is due to the fact that 'anbexec' is not | |
| 3144 | + yet able to manipulate several SSL server certificates. 'anbexec' and | |
| 3145 | + 'predefined.anubis' must be changed. Sorry ! This will be done as soon as possible. The | |
| 3146 | + 'solution' for the time being is to provide the common name of the unique SSL server | |
| 3147 | + certificate. | |
| 3148 | + | |
| 3149 | + | |
| 3150 | +define StartServerResult | |
| 3151 | + start_https_server | |
| 3152 | + ( | |
| 3153 | + Int32 ip_address, | |
| 3154 | + Int32 port, | |
| 3155 | + String certificate_common_name, | |
| 3156 | + Server -> (SSL_Connection -> One) handler, | |
| 3157 | + Int32 retries, | |
| 3158 | + DenialOfService dos | |
| 3159 | + ) = | |
| 3160 | + if start_ssl_server(ip_address, | |
| 3161 | + port, | |
| 3162 | + certificate_common_name, | |
| 3163 | + handler, | |
| 3164 | + identity) is ok(server) | |
| 3165 | + then print(" \r"); | |
| 3166 | + ok(server) | |
| 3167 | + else print("Port "+port+": retry number "+retries+"\r"); | |
| 3168 | + sleep(1000); | |
| 3169 | + start_https_server(ip_address,port, | |
| 3170 | + certificate_common_name, | |
| 3171 | + handler,retries+1, | |
| 3172 | + dos). | |
| 3173 | + | |
| 3174 | + | |
| 3175 | +public define StartServerResult | |
| 3176 | + start_https_server | |
| 3177 | + ( | |
| 3178 | + Int32 ip_address, | |
| 3179 | + Int32 port, | |
| 3180 | + String certificate_common_name, // of SSL server certificate | |
| 3181 | + List(Web_Site_Description) sites, | |
| 3182 | + DenialOfService dos | |
| 3183 | + ) = | |
| 3184 | + create_directories(sites); | |
| 3185 | + start_https_server(ip_address,port,certificate_common_name, | |
| 3186 | + make_https_handler(sites,dos), | |
| 3187 | + 0,dos). | |
| 3188 | + | |
| 3189 | + | |
| 3190 | + | |
| 3191 | + | |
| 3192 | + | |
| 3193 | + | |
| 3194 | + | |
| 3195 | + | |
| 3196 | + | |
| 3197 | + *** [7] The web dispatcher. | |
| 3198 | + | |
| 3199 | + | |
| 3200 | + *** [7.1] The dispatcher server. | |
| 3201 | + | |
| 3202 | +define One | |
| 3203 | + send_dispatching_page | |
| 3204 | + ( | |
| 3205 | + RWStream conn, | |
| 3206 | + String common_name, | |
| 3207 | + Int32 port | |
| 3208 | + ) = | |
| 3209 | + print("Dispatching '"+common_name+"' to port "+port+"\n"); | |
| 3210 | + forget(reliable_write(conn,to_byte_array( | |
| 3211 | + "<html><head><meta http-equiv=\"Refresh\" content=\"0;URL="+ | |
| 3212 | + "http://"+common_name+":"+port+"/"+ | |
| 3213 | + "\"></head><body></body></html>" | |
| 3214 | + ))). | |
| 3215 | + | |
| 3216 | + | |
| 3217 | + | |
| 3218 | +define Maybe(DispatcherInfo) | |
| 3219 | + find_host | |
| 3220 | + ( | |
| 3221 | + List(DispatcherInfo) l, | |
| 3222 | + String host | |
| 3223 | + ) = | |
| 3224 | + if l is | |
| 3225 | + { | |
| 3226 | + [ ] then failure, | |
| 3227 | + [h . t] then if h is site(name,port) then | |
| 3228 | + if name = host | |
| 3229 | + then success(h) | |
| 3230 | + else find_host(t,host) | |
| 3231 | + }. | |
| 3232 | + | |
| 3233 | + | |
| 3234 | + | |
| 3235 | +define Server -> ((RWStream) -> One) | |
| 3236 | + make_dispatcher_handler | |
| 3237 | + ( | |
| 3238 | + Var(List(DispatcherInfo)) info_v, | |
| 3239 | + DenialOfService dos | |
| 3240 | + ) = | |
| 3241 | + (Server server) |-> (RWStream conn) |-> | |
| 3242 | + with start_time = (Int32)now, | |
| 3243 | + if read_request_line(tcp(conn),start_time+*request_line_delay(dos),dos) is | |
| 3244 | + { | |
| 3245 | + error(msg) then print(format(msg)), | |
| 3246 | + ok(request_line) then | |
| 3247 | + if read_http_headers(tcp(conn),start_time+*headers_delay(dos),dos) is | |
| 3248 | + { | |
| 3249 | + error(msg) then print(format(msg)), | |
| 3250 | + ok(headers) then if get_host_header_value(headers) is | |
| 3251 | + { | |
| 3252 | + failure then print("No 'HOST' HTTP header.\n"), | |
| 3253 | + success(host) then | |
| 3254 | + if find_host(*info_v,host) is | |
| 3255 | + { | |
| 3256 | + failure then print("Host: '"+host+"' not registered.\n"), | |
| 3257 | + success(s) then if s is site(common_name,ip_port) then | |
| 3258 | + send_dispatching_page(conn,common_name,ip_port) | |
| 3259 | + } | |
| 3260 | + } | |
| 3261 | + } | |
| 3262 | + }. | |
| 3263 | + | |
| 3264 | + | |
| 3265 | +define One | |
| 3266 | + dispatcher_update_error | |
| 3267 | + ( | |
| 3268 | + String file_path | |
| 3269 | + ) = | |
| 3270 | + print("web_dispatcher: unable to reread file: '"+file_path+"'.\n"). | |
| 3271 | + | |
| 3272 | + | |
| 3273 | +define Bool | |
| 3274 | + dispatcher_update_data | |
| 3275 | + ( | |
| 3276 | + String info_file_path, | |
| 3277 | + Var(List(DispatcherInfo)) info_v, | |
| 3278 | + Var(Int32) info_date_v | |
| 3279 | + ) = | |
| 3280 | + if directory_full_list(my_anubis_directory+"/web_sites","dispatcher.info","","") is | |
| 3281 | + { | |
| 3282 | + [ ] then false, | |
| 3283 | + [h . t] then if h is | |
| 3284 | + { | |
| 3285 | + no_info(n) then false, | |
| 3286 | + file(n,_,_,d) then if n = "dispatcher.info" | |
| 3287 | + then (info_date_v <- d; | |
| 3288 | + if (RetrieveResult(List(DispatcherInfo)))retrieve(info_file_path) is | |
| 3289 | + { | |
| 3290 | + cannot_find_file then false, | |
| 3291 | + read_error then false, | |
| 3292 | + type_error then false, | |
| 3293 | + ok(info) then info_v <- info; true | |
| 3294 | + }) | |
| 3295 | + else false, | |
| 3296 | + link(_,_,_,_) then false, | |
| 3297 | + directory(_,_,_) then false | |
| 3298 | + } | |
| 3299 | + }. | |
| 3300 | + | |
| 3301 | + | |
| 3302 | + | |
| 3303 | + The loop within which the dispatcher updates its data every 3 seconds: | |
| 3304 | + | |
| 3305 | +define One | |
| 3306 | + dispatcher_update_task | |
| 3307 | + ( | |
| 3308 | + String info_file_path, | |
| 3309 | + Var(List(DispatcherInfo)) info_v, | |
| 3310 | + Var(Int32) info_date_v | |
| 3311 | + ) = | |
| 3312 | + sleep(3000); | |
| 3313 | + (if dispatcher_update_data(info_file_path,info_v,info_date_v) | |
| 3314 | + then unique | |
| 3315 | + else dispatcher_update_error(info_file_path)); | |
| 3316 | + dispatcher_update_task(info_file_path,info_v,info_date_v). | |
| 3317 | + | |
| 3318 | + | |
| 3319 | +public define One | |
| 3320 | + start_web_dispatcher | |
| 3321 | + ( | |
| 3322 | + Int32 ip_address, // address for listening (typically 0: listen on all interfaces) | |
| 3323 | + Int32 http_port, // typically 80 | |
| 3324 | + DenialOfService dos | |
| 3325 | + ) = | |
| 3326 | + with info_file_path = my_anubis_directory+"/web_sites/dispatcher.info", | |
| 3327 | + info_v = var((List(DispatcherInfo))[]), | |
| 3328 | + info_date_v = var((Int32)0), | |
| 3329 | + if dispatcher_update_data(info_file_path,info_v,info_date_v) | |
| 3330 | + then if start_server(ip_address, | |
| 3331 | + http_port, | |
| 3332 | + make_dispatcher_handler(info_v,dos), | |
| 3333 | + (One u)|->u) is | |
| 3334 | + { | |
| 3335 | + cannot_create_the_socket then | |
| 3336 | + print("Cannot create the socket for HTTP server.\n"), | |
| 3337 | + cannot_bind_to_port then | |
| 3338 | + print("Cannot bind HTTP server to port "+http_port+".\n"), | |
| 3339 | + cannot_listen_on_port then | |
| 3340 | + print("HTTP server cannot listen on port "+http_port+".\n"), | |
| 3341 | + ok(http_server) then | |
| 3342 | + dispatcher_update_task(info_file_path,info_v,info_date_v) | |
| 3343 | + } | |
| 3344 | + else dispatcher_update_error(info_file_path). | |
| 3345 | + | |
| 3346 | + | |
| 3347 | + | |
| 3348 | + *** [7.2] The dispatcher web site. | |
| 3349 | + | |
| 3350 | + global define One | |
| 3351 | + web_dispatcher | |
| 3352 | + ( | |
| 3353 | + List(String) args | |
| 3354 | + ) = | |
| 3355 | + start_web_dispatcher(0,80,load_denial_of_service_info). | |
| 3356 | + | |
| 3357 | + | |
| 3358 | + | |
| 3359 | + | |
| 3360 | + | |
| 3361 | + | |
| 3362 | + *** [7.3] Managing the info file. | |
| 3363 | + | |
| 3364 | +define Int32 | |
| 3365 | + register_ip_address | |
| 3366 | + = | |
| 3367 | + if ip_address(prompt(" numerical IP address (for HTTP): ")) is | |
| 3368 | + { | |
| 3369 | + failure then print(" *** Error: incorrect IP address.\n"); | |
| 3370 | + register_ip_address, | |
| 3371 | + success(n) then n | |
| 3372 | + }. | |
| 3373 | + | |
| 3374 | + | |
| 3375 | +define Int32 | |
| 3376 | + register_ip_port | |
| 3377 | + = | |
| 3378 | + if string_to_integer(prompt(" IP port (for HTTP): ")) is | |
| 3379 | + { | |
| 3380 | + failure then print(" *** Error: incorrect IP port.\n"); | |
| 3381 | + register_ip_port, | |
| 3382 | + success(p) then if (0 =< p & p =< 65535) | |
| 3383 | + then p | |
| 3384 | + else print(" *** Error: IP port out of bounds.\n"); | |
| 3385 | + register_ip_port | |
| 3386 | + }. | |
| 3387 | + | |
| 3388 | + | |
| 3389 | +define One | |
| 3390 | + register_new_site | |
| 3391 | + ( | |
| 3392 | + Var(List(DispatcherInfo)) info_v | |
| 3393 | + ) = | |
| 3394 | + print("\n"); | |
| 3395 | + print(" Registering a new site:\n"); | |
| 3396 | + with name = prompt(" Site name: "), | |
| 3397 | + with addr = register_ip_address, | |
| 3398 | + with port = register_ip_port, | |
| 3399 | + (protect info_v <- [site(name,port) . *info_v]); | |
| 3400 | + print(" Site "+name+" at "+ip_addr_to_string(addr)+":"+port+" added\n (but not saved to disk).\n"). | |
| 3401 | + | |
| 3402 | + | |
| 3403 | +define List(DispatcherInfo) | |
| 3404 | + find_sites | |
| 3405 | + ( | |
| 3406 | + List(DispatcherInfo) l, | |
| 3407 | + String name | |
| 3408 | + ) = | |
| 3409 | + if l is | |
| 3410 | + { | |
| 3411 | + [ ] then [ ], | |
| 3412 | + [h . t] then if h is site(n,_) then | |
| 3413 | + if find(name,n,0) is | |
| 3414 | + { | |
| 3415 | + failure then find_sites(t,name), | |
| 3416 | + success(_) then [h . find_sites(t,name)] | |
| 3417 | + } | |
| 3418 | + }. | |
| 3419 | + | |
| 3420 | + | |
| 3421 | +define String | |
| 3422 | + pad | |
| 3423 | + ( | |
| 3424 | + String s, | |
| 3425 | + Int32 l | |
| 3426 | + ) = | |
| 3427 | + if length(s) >= l | |
| 3428 | + then s | |
| 3429 | + else s+constant_string(l-length(s),' '). | |
| 3430 | + | |
| 3431 | + | |
| 3432 | + | |
| 3433 | +define One | |
| 3434 | + show_sites_1 | |
| 3435 | + ( | |
| 3436 | + List(DispatcherInfo) l, | |
| 3437 | + Int32 i | |
| 3438 | + ) = | |
| 3439 | + if l is | |
| 3440 | + { | |
| 3441 | + [ ] then unique, | |
| 3442 | + [h . t] then if h is site(name,port) then | |
| 3443 | + print(" ["+i+"] "+pad(name,40)+" "+" "+port+"\n"); | |
| 3444 | + show_sites_1(t,i+1) | |
| 3445 | + }. | |
| 3446 | + | |
| 3447 | + | |
| 3448 | +define One | |
| 3449 | + show_sites | |
| 3450 | + ( | |
| 3451 | + List(DispatcherInfo) l, | |
| 3452 | + Int32 i | |
| 3453 | + ) = | |
| 3454 | + print(" Name Port\n"); | |
| 3455 | + print(" --------------------------------------------------------\n"); | |
| 3456 | + show_sites_1(l,i). | |
| 3457 | + | |
| 3458 | +define List(DispatcherInfo) | |
| 3459 | + replace_info | |
| 3460 | + ( | |
| 3461 | + List(DispatcherInfo) l, | |
| 3462 | + String site_name, | |
| 3463 | + Int32 new_port | |
| 3464 | + ) = | |
| 3465 | + if l is | |
| 3466 | + { | |
| 3467 | + [ ] then alert, | |
| 3468 | + [h . t] then if h is site(n,_) then | |
| 3469 | + if n = site_name | |
| 3470 | + then [site(n,new_port) . t] | |
| 3471 | + else [h . replace_info(t,site_name,new_port)] | |
| 3472 | + }. | |
| 3473 | + | |
| 3474 | +define List(DispatcherInfo) | |
| 3475 | + delete_info | |
| 3476 | + ( | |
| 3477 | + List(DispatcherInfo) l, | |
| 3478 | + String site_name, | |
| 3479 | + ) = | |
| 3480 | + if l is | |
| 3481 | + { | |
| 3482 | + [ ] then alert, | |
| 3483 | + [h . t] then if h is site(n,_) then | |
| 3484 | + if n = site_name | |
| 3485 | + then t | |
| 3486 | + else [h . delete_info(t,site_name)] | |
| 3487 | + }. | |
| 3488 | + | |
| 3489 | + | |
| 3490 | +define One | |
| 3491 | + update_site | |
| 3492 | + ( | |
| 3493 | + Var(List(DispatcherInfo)) info_v, | |
| 3494 | + String site_name, | |
| 3495 | + Int32 old_port | |
| 3496 | + ) = | |
| 3497 | + print("\n"); | |
| 3498 | + print(" Updating site '"+site_name+"': (currently: "+old_port+")\n"); | |
| 3499 | + with new_port = register_ip_port, | |
| 3500 | + answer = prompt(" Update '"+site_name+"' as: "+new_port+" [Y/N] ? "), | |
| 3501 | + if (answer = "Y" | answer = "y") | |
| 3502 | + then info_v <- replace_info(*info_v,site_name,new_port) | |
| 3503 | + else unique. | |
| 3504 | + | |
| 3505 | + | |
| 3506 | + | |
| 3507 | +define Bool | |
| 3508 | + compare | |
| 3509 | + ( | |
| 3510 | + DispatcherInfo d1, | |
| 3511 | + DispatcherInfo d2 | |
| 3512 | + ) = | |
| 3513 | + if d1 is site(n1,_) then | |
| 3514 | + if d2 is site(n2,_) then | |
| 3515 | + string_less(n1,n2). | |
| 3516 | + | |
| 3517 | + | |
| 3518 | + | |
| 3519 | +define One | |
| 3520 | + update_site | |
| 3521 | + ( | |
| 3522 | + Var(List(DispatcherInfo)) info_v | |
| 3523 | + ) = | |
| 3524 | + print("\n"); | |
| 3525 | + with prefix = prompt(" Search for site to update: "), | |
| 3526 | + if find_sites(*info_v,prefix) is | |
| 3527 | + { | |
| 3528 | + [ ] then print(" No site found.\n"); | |
| 3529 | + update_site(info_v), | |
| 3530 | + [h . t] then | |
| 3531 | + show_sites(qsort([h . t],compare),1); | |
| 3532 | + with i1 = prompt(" Choose a site to update [1/.../"+(length(t)+1)+"]: "), | |
| 3533 | + if string_to_integer(i1) is | |
| 3534 | + { | |
| 3535 | + failure then print(" *** Error: site number not recognized.\n"); | |
| 3536 | + update_site(info_v), | |
| 3537 | + success(ii1) then if nth(ii1-1,*info_v) is | |
| 3538 | + { | |
| 3539 | + failure then print(" *** Error: site number "+i1+" does not exist.\n"); | |
| 3540 | + update_site(info_v), | |
| 3541 | + success(site_info) then if site_info is site(name,old_port) then | |
| 3542 | + update_site(info_v,name,old_port) | |
| 3543 | + } | |
| 3544 | + } | |
| 3545 | + }. | |
| 3546 | + | |
| 3547 | + | |
| 3548 | +define One | |
| 3549 | + delete_site | |
| 3550 | + ( | |
| 3551 | + Var(List(DispatcherInfo)) info_v, | |
| 3552 | + String site_name, | |
| 3553 | + Int32 old_port | |
| 3554 | + ) = | |
| 3555 | + print("\n"); | |
| 3556 | + print(" Deleting site '"+site_name+"': (currently: "+old_port+")\n"); | |
| 3557 | + with answer = prompt(" Are you sure you want to delete site: '"+site_name+"' [Y/N] ? "), | |
| 3558 | + if (answer = "Y" | answer = "y") | |
| 3559 | + then info_v <- delete_info(*info_v,site_name) | |
| 3560 | + else print(" Site '"+site_name+"' not deleted.\n"). | |
| 3561 | + | |
| 3562 | + | |
| 3563 | +define One | |
| 3564 | + delete_site | |
| 3565 | + ( | |
| 3566 | + Var(List(DispatcherInfo)) info_v | |
| 3567 | + ) = | |
| 3568 | + print("\n"); | |
| 3569 | + with prefix = prompt(" Search for site to delete: "), | |
| 3570 | + if find_sites(*info_v,prefix) is | |
| 3571 | + { | |
| 3572 | + [ ] then print(" No site found.\n"); | |
| 3573 | + delete_site(info_v), | |
| 3574 | + [h . t] then | |
| 3575 | + show_sites(qsort([h . t],compare),1); | |
| 3576 | + with i1 = prompt(" Choose a site to delete [1/.../"+(length(t)+1)+"]: "), | |
| 3577 | + if string_to_integer(i1) is | |
| 3578 | + { | |
| 3579 | + failure then print(" *** Error: site number not recognized.\n"); | |
| 3580 | + delete_site(info_v), | |
| 3581 | + success(ii1) then if nth(ii1-1,*info_v) is | |
| 3582 | + { | |
| 3583 | + failure then print(" *** Error: site number "+i1+" does not exist.\n"); | |
| 3584 | + delete_site(info_v), | |
| 3585 | + success(site_info) then if site_info is site(name,old_port) then | |
| 3586 | + delete_site(info_v,name,old_port) | |
| 3587 | + } | |
| 3588 | + } | |
| 3589 | + }. | |
| 3590 | + | |
| 3591 | + | |
| 3592 | +define One | |
| 3593 | + manager | |
| 3594 | + ( | |
| 3595 | + Var(List(DispatcherInfo)) info_v, | |
| 3596 | + String file_path | |
| 3597 | + ) = | |
| 3598 | + print("\n"); | |
| 3599 | + print(" --- Welcome to the Web Dispatcher Manager ---\n"); | |
| 3600 | + with l = length(*info_v), | |
| 3601 | + print(" "+l+" site"+(if l>1 then "s" else "")+" currently registred.\n"); | |
| 3602 | + print(" [L] List registered sites.\n"); | |
| 3603 | + print(" [R] Register a new site.\n"); | |
| 3604 | + print(" [U] Update a registred site.\n"); | |
| 3605 | + print(" [D] Delete a registred site.\n"); | |
| 3606 | + with propose_write_v = var((Bool)true), | |
| 3607 | + action = prompt(" Choose an action [L/R/U/D]: "), | |
| 3608 | + (if (action = "L" | action = "l") then (show_sites(*info_v,1); propose_write_v <- false) else | |
| 3609 | + if (action = "R" | action = "r") then register_new_site(info_v) else | |
| 3610 | + if (action = "U" | action = "u") then update_site(info_v) else | |
| 3611 | + if (action = "D" | action = "d") then delete_site(info_v) else | |
| 3612 | + print("Action not recognized.\n")); | |
| 3613 | + print("\n"); | |
| 3614 | + if *propose_write_v then | |
| 3615 | + with result = prompt(" Write modifications to data base [Y/N] ?"), | |
| 3616 | + if (result = "Y" | result = "y") | |
| 3617 | + then if save(*info_v,file_path) is | |
| 3618 | + { | |
| 3619 | + cannot_open_file then print(" File '"+file_path+"' not found.\n"), | |
| 3620 | + write_error then print(" Error while writing file '"+file_path+"'.\n"), | |
| 3621 | + ok then print(" Data base has been modified.\n") | |
| 3622 | + } | |
| 3623 | + else print(" Data base not modified.\n") | |
| 3624 | + else unique. | |
| 3625 | + | |
| 3626 | + | |
| 3627 | + | |
| 3628 | +global define One | |
| 3629 | + manage_web_dispatcher | |
| 3630 | + ( | |
| 3631 | + List(String) args | |
| 3632 | + ) = | |
| 3633 | + with info_v = var((List(DispatcherInfo))[]), | |
| 3634 | + with file_path = my_anubis_directory+"/web_sites/dispatcher.info", | |
| 3635 | + if (RetrieveResult(List(DispatcherInfo)))retrieve(file_path) is | |
| 3636 | + { | |
| 3637 | + cannot_find_file then print("File '"+file_path+"' does not exist.\n"); | |
| 3638 | + with answer = prompt("Create it [Y/N] ? "), | |
| 3639 | + if (answer = "Y" | answer = "y") | |
| 3640 | + then if save((List(DispatcherInfo))[],file_path) is | |
| 3641 | + { | |
| 3642 | + cannot_open_file then | |
| 3643 | + print("Cannot create file '"+file_path+"'.\n"), | |
| 3644 | + write_error then | |
| 3645 | + print("Error while creating file '"+file_path+"'.\n"), | |
| 3646 | + ok then manager(info_v,file_path) | |
| 3647 | + } | |
| 3648 | + else unique, | |
| 3649 | + read_error then print("Error while reading file '"+file_path+"'.\n"), | |
| 3650 | + type_error then print("File '"+file_path+"' is corrupted.\n"), | |
| 3651 | + ok(info) then info_v <- info; | |
| 3652 | + manager(info_v,file_path) | |
| 3653 | + }. | |
| 3654 | + | |
| 3655 | + | |
| 3656 | + | |
| 3657 | + | |
| 3658 | + | ... | ... |