1<?xml version="1.0" encoding="ISO-8859-1"?> 2<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> 3<html xmlns="http://www.w3.org/1999/xhtml" lang="en" xml:lang="en"><head><!-- 4 XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX 5 This file is generated from xml source: DO NOT EDIT 6 XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX 7 --> 8<title>Apache 1.3 API notes - Apache HTTP Server</title> 9<link href="/style/css/manual.css" rel="stylesheet" media="all" type="text/css" title="Main stylesheet" /> 10<link href="/style/css/manual-loose-100pc.css" rel="alternate stylesheet" media="all" type="text/css" title="No Sidebar - Default font size" /> 11<link href="/style/css/manual-print.css" rel="stylesheet" media="print" type="text/css" /><link rel="stylesheet" type="text/css" href="/style/css/prettify.css" /> 12<script src="/style/scripts/prettify.js" type="text/javascript"> 13</script> 14 15<link href="/images/favicon.ico" rel="shortcut icon" /></head> 16<body id="manual-page"><div id="page-header"> 17<p class="menu"><a href="/mod/">Modules</a> | <a href="/mod/directives.html">Directives</a> | <a href="http://wiki.apache.org/httpd/FAQ">FAQ</a> | <a href="/glossary.html">Glossary</a> | <a href="/sitemap.html">Sitemap</a></p> 18<p class="apache">Apache HTTP Server Version 2.2</p> 19<img alt="" src="/images/feather.gif" /></div> 20<div class="up"><a href="./"><img title="<-" alt="<-" src="/images/left.gif" /></a></div> 21<div id="path"> 22<a href="http://www.apache.org/">Apache</a> > <a href="http://httpd.apache.org/">HTTP Server</a> > <a href="http://httpd.apache.org/docs/">Documentation</a> > <a href="../">Version 2.2</a> > <a href="./">Developer Documentation</a></div><div id="page-content"><div id="preamble"><h1>Apache 1.3 API notes</h1> 23<div class="toplang"> 24<p><span>Available Languages: </span><a href="/en/developer/API.html" title="English"> en </a></p> 25</div> 26 27 <div class="warning"><h3>Warning</h3> 28 <p>This document has not been updated to take into account changes made 29 in the 2.0 version of the Apache HTTP Server. Some of the information may 30 still be relevant, but please use it with care.</p> 31 </div> 32 33 <p>These are some notes on the Apache API and the data structures you have 34 to deal with, <em>etc.</em> They are not yet nearly complete, but hopefully, 35 they will help you get your bearings. Keep in mind that the API is still 36 subject to change as we gain experience with it. (See the TODO file for 37 what <em>might</em> be coming). However, it will be easy to adapt modules 38 to any changes that are made. (We have more modules to adapt than you 39 do).</p> 40 41 <p>A few notes on general pedagogical style here. In the interest of 42 conciseness, all structure declarations here are incomplete -- the real 43 ones have more slots that I'm not telling you about. For the most part, 44 these are reserved to one component of the server core or another, and 45 should be altered by modules with caution. However, in some cases, they 46 really are things I just haven't gotten around to yet. Welcome to the 47 bleeding edge.</p> 48 49 <p>Finally, here's an outline, to give you some bare idea of what's coming 50 up, and in what order:</p> 51 52 <ul> 53 <li> 54 <a href="#basics">Basic concepts.</a> 55 56 <ul> 57 <li><a href="#HMR">Handlers, Modules, and 58 Requests</a></li> 59 60 <li><a href="#moduletour">A brief tour of a 61 module</a></li> 62 </ul> 63 </li> 64 65 <li> 66 <a href="#handlers">How handlers work</a> 67 68 <ul> 69 <li><a href="#req_tour">A brief tour of the 70 <code>request_rec</code></a></li> 71 72 <li><a href="#req_orig">Where request_rec structures come 73 from</a></li> 74 75 <li><a href="#req_return">Handling requests, declining, 76 and returning error codes</a></li> 77 78 <li><a href="#resp_handlers">Special considerations for 79 response handlers</a></li> 80 81 <li><a href="#auth_handlers">Special considerations for 82 authentication handlers</a></li> 83 84 <li><a href="#log_handlers">Special considerations for 85 logging handlers</a></li> 86 </ul> 87 </li> 88 89 <li><a href="#pools">Resource allocation and resource 90 pools</a></li> 91 92 <li> 93 <a href="#config">Configuration, commands and the like</a> 94 95 <ul> 96 <li><a href="#per-dir">Per-directory configuration 97 structures</a></li> 98 99 <li><a href="#commands">Command handling</a></li> 100 101 <li><a href="#servconf">Side notes --- per-server 102 configuration, virtual servers, <em>etc</em>.</a></li> 103 </ul> 104 </li> 105 </ul> 106</div> 107<div id="quickview"><ul id="toc"><li><img alt="" src="/images/down.gif" /> <a href="#basics">Basic concepts</a></li> 108<li><img alt="" src="/images/down.gif" /> <a href="#handlers">How handlers work</a></li> 109<li><img alt="" src="/images/down.gif" /> <a href="#pools">Resource allocation and resource pools</a></li> 110<li><img alt="" src="/images/down.gif" /> <a href="#config">Configuration, commands and the like</a></li> 111</ul><ul class="seealso"><li><a href="#comments_section">Comments</a></li></ul></div> 112<div class="top"><a href="#page-header"><img alt="top" src="/images/up.gif" /></a></div> 113<div class="section"> 114<h2><a name="basics" id="basics">Basic concepts</a></h2> 115 <p>We begin with an overview of the basic concepts behind the API, and how 116 they are manifested in the code.</p> 117 118 <h3><a name="HMR" id="HMR">Handlers, Modules, and Requests</a></h3> 119 <p>Apache breaks down request handling into a series of steps, more or 120 less the same way the Netscape server API does (although this API has a 121 few more stages than NetSite does, as hooks for stuff I thought might be 122 useful in the future). These are:</p> 123 124 <ul> 125 <li>URI -> Filename translation</li> 126 <li>Auth ID checking [is the user who they say they are?]</li> 127 <li>Auth access checking [is the user authorized <em>here</em>?]</li> 128 <li>Access checking other than auth</li> 129 <li>Determining MIME type of the object requested</li> 130 <li>`Fixups' -- there aren't any of these yet, but the phase is intended 131 as a hook for possible extensions like <code class="directive"><a href="/mod/mod_env.html#setenv">SetEnv</a></code>, which don't really fit well elsewhere.</li> 132 <li>Actually sending a response back to the client.</li> 133 <li>Logging the request</li> 134 </ul> 135 136 <p>These phases are handled by looking at each of a succession of 137 <em>modules</em>, looking to see if each of them has a handler for the 138 phase, and attempting invoking it if so. The handler can typically do one 139 of three things:</p> 140 141 <ul> 142 <li><em>Handle</em> the request, and indicate that it has done so by 143 returning the magic constant <code>OK</code>.</li> 144 145 <li><em>Decline</em> to handle the request, by returning the magic integer 146 constant <code>DECLINED</code>. In this case, the server behaves in all 147 respects as if the handler simply hadn't been there.</li> 148 149 <li>Signal an error, by returning one of the HTTP error codes. This 150 terminates normal handling of the request, although an ErrorDocument may 151 be invoked to try to mop up, and it will be logged in any case.</li> 152 </ul> 153 154 <p>Most phases are terminated by the first module that handles them; 155 however, for logging, `fixups', and non-access authentication checking, 156 all handlers always run (barring an error). Also, the response phase is 157 unique in that modules may declare multiple handlers for it, via a 158 dispatch table keyed on the MIME type of the requested object. Modules may 159 declare a response-phase handler which can handle <em>any</em> request, 160 by giving it the key <code>*/*</code> (<em>i.e.</em>, a wildcard MIME type 161 specification). However, wildcard handlers are only invoked if the server 162 has already tried and failed to find a more specific response handler for 163 the MIME type of the requested object (either none existed, or they all 164 declined).</p> 165 166 <p>The handlers themselves are functions of one argument (a 167 <code>request_rec</code> structure. vide infra), which returns an integer, 168 as above.</p> 169 170 171 <h3><a name="moduletour" id="moduletour">A brief tour of a module</a></h3> 172 <p>At this point, we need to explain the structure of a module. Our 173 candidate will be one of the messier ones, the CGI module -- this handles 174 both CGI scripts and the <code class="directive"><a href="/mod/mod_alias.html#scriptalias">ScriptAlias</a></code> config file command. It's actually a great deal 175 more complicated than most modules, but if we're going to have only one 176 example, it might as well be the one with its fingers in every place.</p> 177 178 <p>Let's begin with handlers. In order to handle the CGI scripts, the 179 module declares a response handler for them. Because of <code class="directive"><a href="/mod/mod_alias.html#scriptalias">ScriptAlias</a></code>, it also has handlers for the 180 name translation phase (to recognize <code class="directive"><a href="/mod/mod_alias.html#scriptalias">ScriptAlias</a></code>ed URIs), the type-checking phase (any 181 <code class="directive"><a href="/mod/mod_alias.html#scriptalias">ScriptAlias</a></code>ed request is typed 182 as a CGI script).</p> 183 184 <p>The module needs to maintain some per (virtual) server information, 185 namely, the <code class="directive"><a href="/mod/mod_alias.html#scriptalias">ScriptAlias</a></code>es in 186 effect; the module structure therefore contains pointers to a functions 187 which builds these structures, and to another which combines two of them 188 (in case the main server and a virtual server both have <code class="directive"><a href="/mod/mod_alias.html#scriptalias">ScriptAlias</a></code>es declared).</p> 189 190 <p>Finally, this module contains code to handle the <code class="directive"><a href="/mod/mod_alias.html#scriptalias">ScriptAlias</a></code> command itself. This particular 191 module only declares one command, but there could be more, so modules have 192 <em>command tables</em> which declare their commands, and describe where 193 they are permitted, and how they are to be invoked.</p> 194 195 <p>A final note on the declared types of the arguments of some of these 196 commands: a <code>pool</code> is a pointer to a <em>resource pool</em> 197 structure; these are used by the server to keep track of the memory which 198 has been allocated, files opened, <em>etc.</em>, either to service a 199 particular request, or to handle the process of configuring itself. That 200 way, when the request is over (or, for the configuration pool, when the 201 server is restarting), the memory can be freed, and the files closed, 202 <em>en masse</em>, without anyone having to write explicit code to track 203 them all down and dispose of them. Also, a <code>cmd_parms</code> 204 structure contains various information about the config file being read, 205 and other status information, which is sometimes of use to the function 206 which processes a config-file command (such as <code class="directive"><a href="/mod/mod_alias.html#scriptalias">ScriptAlias</a></code>). With no further ado, the 207 module itself:</p> 208 209 <div class="example"><p><code> 210 /* Declarations of handlers. */<br /> 211 <br /> 212 int translate_scriptalias (request_rec *);<br /> 213 int type_scriptalias (request_rec *);<br /> 214 int cgi_handler (request_rec *);<br /> 215 <br /> 216 /* Subsidiary dispatch table for response-phase <br /> 217 * handlers, by MIME type */<br /> 218 <br /> 219 handler_rec cgi_handlers[] = {<br /> 220 <span class="indent"> 221 { "application/x-httpd-cgi", cgi_handler },<br /> 222 { NULL }<br /> 223 </span> 224 };<br /> 225 <br /> 226 /* Declarations of routines to manipulate the <br /> 227 * module's configuration info. Note that these are<br /> 228 * returned, and passed in, as void *'s; the server<br /> 229 * core keeps track of them, but it doesn't, and can't,<br /> 230 * know their internal structure.<br /> 231 */<br /> 232 <br /> 233 void *make_cgi_server_config (pool *);<br /> 234 void *merge_cgi_server_config (pool *, void *, void *);<br /> 235 <br /> 236 /* Declarations of routines to handle config-file commands */<br /> 237 <br /> 238 extern char *script_alias(cmd_parms *, void *per_dir_config, char *fake, 239 char *real);<br /> 240 <br /> 241 command_rec cgi_cmds[] = {<br /> 242 <span class="indent"> 243 { "ScriptAlias", script_alias, NULL, RSRC_CONF, TAKE2,<br /> 244 <span class="indent">"a fakename and a realname"},<br /></span> 245 { NULL }<br /> 246 </span> 247 };<br /> 248 <br /> 249 module cgi_module = { 250</code></p><pre> STANDARD_MODULE_STUFF, 251 NULL, /* initializer */ 252 NULL, /* dir config creator */ 253 NULL, /* dir merger */ 254 make_cgi_server_config, /* server config */ 255 merge_cgi_server_config, /* merge server config */ 256 cgi_cmds, /* command table */ 257 cgi_handlers, /* handlers */ 258 translate_scriptalias, /* filename translation */ 259 NULL, /* check_user_id */ 260 NULL, /* check auth */ 261 NULL, /* check access */ 262 type_scriptalias, /* type_checker */ 263 NULL, /* fixups */ 264 NULL, /* logger */ 265 NULL /* header parser */ 266};</pre></div> 267 268</div><div class="top"><a href="#page-header"><img alt="top" src="/images/up.gif" /></a></div> 269<div class="section"> 270<h2><a name="handlers" id="handlers">How handlers work</a></h2> 271 <p>The sole argument to handlers is a <code>request_rec</code> structure. 272 This structure describes a particular request which has been made to the 273 server, on behalf of a client. In most cases, each connection to the 274 client generates only one <code>request_rec</code> structure.</p> 275 276 <h3><a name="req_tour" id="req_tour">A brief tour of the request_rec</a></h3> 277 <p>The <code>request_rec</code> contains pointers to a resource pool 278 which will be cleared when the server is finished handling the request; 279 to structures containing per-server and per-connection information, and 280 most importantly, information on the request itself.</p> 281 282 <p>The most important such information is a small set of character strings 283 describing attributes of the object being requested, including its URI, 284 filename, content-type and content-encoding (these being filled in by the 285 translation and type-check handlers which handle the request, 286 respectively).</p> 287 288 <p>Other commonly used data items are tables giving the MIME headers on 289 the client's original request, MIME headers to be sent back with the 290 response (which modules can add to at will), and environment variables for 291 any subprocesses which are spawned off in the course of servicing the 292 request. These tables are manipulated using the <code>ap_table_get</code> 293 and <code>ap_table_set</code> routines.</p> 294 295 <div class="note"> 296 <p>Note that the <code>Content-type</code> header value <em>cannot</em> 297 be set by module content-handlers using the <code>ap_table_*()</code> 298 routines. Rather, it is set by pointing the <code>content_type</code> 299 field in the <code>request_rec</code> structure to an appropriate 300 string. <em>e.g.</em>,</p> 301 <div class="example"><p><code> 302 r->content_type = "text/html"; 303 </code></p></div> 304 </div> 305 306 <p>Finally, there are pointers to two data structures which, in turn, 307 point to per-module configuration structures. Specifically, these hold 308 pointers to the data structures which the module has built to describe 309 the way it has been configured to operate in a given directory (via 310 <code>.htaccess</code> files or <code class="directive"><a href="/mod/core.html#directory"><Directory></a></code> sections), for private data it has built in the 311 course of servicing the request (so modules' handlers for one phase can 312 pass `notes' to their handlers for other phases). There is another such 313 configuration vector in the <code>server_rec</code> data structure pointed 314 to by the <code>request_rec</code>, which contains per (virtual) server 315 configuration data.</p> 316 317 <p>Here is an abridged declaration, giving the fields most commonly 318 used:</p> 319 320 <div class="example"><p><code> 321 struct request_rec {<br /> 322 <br /> 323 pool *pool;<br /> 324 conn_rec *connection;<br /> 325 server_rec *server;<br /> 326 <br /> 327 /* What object is being requested */<br /> 328 <br /> 329 char *uri;<br /> 330 char *filename;<br /> 331 char *path_info; 332</code></p><pre>char *args; /* QUERY_ARGS, if any */ 333struct stat finfo; /* Set by server core; 334 * st_mode set to zero if no such file */</pre><p><code> 335 char *content_type;<br /> 336 char *content_encoding;<br /> 337 <br /> 338 /* MIME header environments, in and out. Also, <br /> 339 * an array containing environment variables to<br /> 340 * be passed to subprocesses, so people can write<br /> 341 * modules to add to that environment.<br /> 342 *<br /> 343 * The difference between headers_out and <br /> 344 * err_headers_out is that the latter are printed <br /> 345 * even on error, and persist across internal<br /> 346 * redirects (so the headers printed for <br /> 347 * <code class="directive"><a href="/mod/core.html#errordocument">ErrorDocument</a></code> handlers will have 348 them).<br /> 349 */<br /> 350 <br /> 351 table *headers_in;<br /> 352 table *headers_out;<br /> 353 table *err_headers_out;<br /> 354 table *subprocess_env;<br /> 355 <br /> 356 /* Info about the request itself... */<br /> 357 <br /> 358</code></p><pre>int header_only; /* HEAD request, as opposed to GET */ 359char *protocol; /* Protocol, as given to us, or HTTP/0.9 */ 360char *method; /* GET, HEAD, POST, <em>etc.</em> */ 361int method_number; /* M_GET, M_POST, <em>etc.</em> */ 362 363</pre><p><code> 364 /* Info for logging */<br /> 365 <br /> 366 char *the_request;<br /> 367 int bytes_sent;<br /> 368 <br /> 369 /* A flag which modules can set, to indicate that<br /> 370 * the data being returned is volatile, and clients<br /> 371 * should be told not to cache it.<br /> 372 */<br /> 373 <br /> 374 int no_cache;<br /> 375 <br /> 376 /* Various other config info which may change<br /> 377 * with .htaccess files<br /> 378 * These are config vectors, with one void*<br /> 379 * pointer for each module (the thing pointed<br /> 380 * to being the module's business).<br /> 381 */<br /> 382 <br /> 383</code></p><pre>void *per_dir_config; /* Options set in config files, <em>etc.</em> */ 384void *request_config; /* Notes on *this* request */</pre><p><code> 385 <br /> 386 }; 387 </code></p></div> 388 389 390 <h3><a name="req_orig" id="req_orig">Where request_rec structures come from</a></h3> 391 <p>Most <code>request_rec</code> structures are built by reading an HTTP 392 request from a client, and filling in the fields. However, there are a 393 few exceptions:</p> 394 395 <ul> 396 <li>If the request is to an imagemap, a type map (<em>i.e.</em>, a 397 <code>*.var</code> file), or a CGI script which returned a local 398 `Location:', then the resource which the user requested is going to be 399 ultimately located by some URI other than what the client originally 400 supplied. In this case, the server does an <em>internal redirect</em>, 401 constructing a new <code>request_rec</code> for the new URI, and 402 processing it almost exactly as if the client had requested the new URI 403 directly.</li> 404 405 <li>If some handler signaled an error, and an <code>ErrorDocument</code> 406 is in scope, the same internal redirect machinery comes into play.</li> 407 408 <li><p>Finally, a handler occasionally needs to investigate `what would 409 happen if' some other request were run. For instance, the directory 410 indexing module needs to know what MIME type would be assigned to a 411 request for each directory entry, in order to figure out what icon to 412 use.</p> 413 414 <p>Such handlers can construct a <em>sub-request</em>, using the 415 functions <code>ap_sub_req_lookup_file</code>, 416 <code>ap_sub_req_lookup_uri</code>, and <code>ap_sub_req_method_uri</code>; 417 these construct a new <code>request_rec</code> structure and processes it 418 as you would expect, up to but not including the point of actually sending 419 a response. (These functions skip over the access checks if the 420 sub-request is for a file in the same directory as the original 421 request).</p> 422 423 <p>(Server-side includes work by building sub-requests and then actually 424 invoking the response handler for them, via the function 425 <code>ap_run_sub_req</code>).</p> 426 </li> 427 </ul> 428 429 430 <h3><a name="req_return" id="req_return">Handling requests, declining, and returning 431 error codes</a></h3> 432 <p>As discussed above, each handler, when invoked to handle a particular 433 <code>request_rec</code>, has to return an <code>int</code> to indicate 434 what happened. That can either be</p> 435 436 <ul> 437 <li><code>OK</code> -- the request was handled successfully. This may or 438 may not terminate the phase.</li> 439 440 <li><code>DECLINED</code> -- no erroneous condition exists, but the module 441 declines to handle the phase; the server tries to find another.</li> 442 443 <li>an HTTP error code, which aborts handling of the request.</li> 444 </ul> 445 446 <p>Note that if the error code returned is <code>REDIRECT</code>, then 447 the module should put a <code>Location</code> in the request's 448 <code>headers_out</code>, to indicate where the client should be 449 redirected <em>to</em>.</p> 450 451 452 <h3><a name="resp_handlers" id="resp_handlers">Special considerations for response 453 handlers</a></h3> 454 <p>Handlers for most phases do their work by simply setting a few fields 455 in the <code>request_rec</code> structure (or, in the case of access 456 checkers, simply by returning the correct error code). However, response 457 handlers have to actually send a request back to the client.</p> 458 459 <p>They should begin by sending an HTTP response header, using the 460 function <code>ap_send_http_header</code>. (You don't have to do anything 461 special to skip sending the header for HTTP/0.9 requests; the function 462 figures out on its own that it shouldn't do anything). If the request is 463 marked <code>header_only</code>, that's all they should do; they should 464 return after that, without attempting any further output.</p> 465 466 <p>Otherwise, they should produce a request body which responds to the 467 client as appropriate. The primitives for this are <code>ap_rputc</code> 468 and <code>ap_rprintf</code>, for internally generated output, and 469 <code>ap_send_fd</code>, to copy the contents of some <code>FILE *</code> 470 straight to the client.</p> 471 472 <p>At this point, you should more or less understand the following piece 473 of code, which is the handler which handles <code>GET</code> requests 474 which have no more specific handler; it also shows how conditional 475 <code>GET</code>s can be handled, if it's desirable to do so in a 476 particular response handler -- <code>ap_set_last_modified</code> checks 477 against the <code>If-modified-since</code> value supplied by the client, 478 if any, and returns an appropriate code (which will, if nonzero, be 479 USE_LOCAL_COPY). No similar considerations apply for 480 <code>ap_set_content_length</code>, but it returns an error code for 481 symmetry.</p> 482 483 <div class="example"><p><code> 484 int default_handler (request_rec *r)<br /> 485 {<br /> 486 <span class="indent"> 487 int errstatus;<br /> 488 FILE *f;<br /> 489 <br /> 490 if (r->method_number != M_GET) return DECLINED;<br /> 491 if (r->finfo.st_mode == 0) return NOT_FOUND;<br /> 492 <br /> 493 if ((errstatus = ap_set_content_length (r, r->finfo.st_size))<br /> 494 || 495 (errstatus = ap_set_last_modified (r, r->finfo.st_mtime)))<br /> 496 return errstatus;<br /> 497 <br /> 498 f = fopen (r->filename, "r");<br /> 499 <br /> 500 if (f == NULL) {<br /> 501 <span class="indent"> 502 log_reason("file permissions deny server access", r->filename, r);<br /> 503 return FORBIDDEN;<br /> 504 </span> 505 }<br /> 506 <br /> 507 register_timeout ("send", r);<br /> 508 ap_send_http_header (r);<br /> 509 <br /> 510 if (!r->header_only) send_fd (f, r);<br /> 511 ap_pfclose (r->pool, f);<br /> 512 return OK;<br /> 513 </span> 514 } 515 </code></p></div> 516 517 <p>Finally, if all of this is too much of a challenge, there are a few 518 ways out of it. First off, as shown above, a response handler which has 519 not yet produced any output can simply return an error code, in which 520 case the server will automatically produce an error response. Secondly, 521 it can punt to some other handler by invoking 522 <code>ap_internal_redirect</code>, which is how the internal redirection 523 machinery discussed above is invoked. A response handler which has 524 internally redirected should always return <code>OK</code>.</p> 525 526 <p>(Invoking <code>ap_internal_redirect</code> from handlers which are 527 <em>not</em> response handlers will lead to serious confusion).</p> 528 529 530 <h3><a name="auth_handlers" id="auth_handlers">Special considerations for authentication 531 handlers</a></h3> 532 <p>Stuff that should be discussed here in detail:</p> 533 534 <ul> 535 <li>Authentication-phase handlers not invoked unless auth is 536 configured for the directory.</li> 537 538 <li>Common auth configuration stored in the core per-dir 539 configuration; it has accessors <code>ap_auth_type</code>, 540 <code>ap_auth_name</code>, and <code>ap_requires</code>.</li> 541 542 <li>Common routines, to handle the protocol end of things, at 543 least for HTTP basic authentication 544 (<code>ap_get_basic_auth_pw</code>, which sets the 545 <code>connection->user</code> structure field 546 automatically, and <code>ap_note_basic_auth_failure</code>, 547 which arranges for the proper <code>WWW-Authenticate:</code> 548 header to be sent back).</li> 549 </ul> 550 551 552 <h3><a name="log_handlers" id="log_handlers">Special considerations for logging 553 handlers</a></h3> 554 <p>When a request has internally redirected, there is the question of 555 what to log. Apache handles this by bundling the entire chain of redirects 556 into a list of <code>request_rec</code> structures which are threaded 557 through the <code>r->prev</code> and <code>r->next</code> pointers. 558 The <code>request_rec</code> which is passed to the logging handlers in 559 such cases is the one which was originally built for the initial request 560 from the client; note that the <code>bytes_sent</code> field will only be 561 correct in the last request in the chain (the one for which a response was 562 actually sent).</p> 563 564</div><div class="top"><a href="#page-header"><img alt="top" src="/images/up.gif" /></a></div> 565<div class="section"> 566<h2><a name="pools" id="pools">Resource allocation and resource pools</a></h2> 567 <p>One of the problems of writing and designing a server-pool server is 568 that of preventing leakage, that is, allocating resources (memory, open 569 files, <em>etc.</em>), without subsequently releasing them. The resource 570 pool machinery is designed to make it easy to prevent this from happening, 571 by allowing resource to be allocated in such a way that they are 572 <em>automatically</em> released when the server is done with them.</p> 573 574 <p>The way this works is as follows: the memory which is allocated, file 575 opened, <em>etc.</em>, to deal with a particular request are tied to a 576 <em>resource pool</em> which is allocated for the request. The pool is a 577 data structure which itself tracks the resources in question.</p> 578 579 <p>When the request has been processed, the pool is <em>cleared</em>. At 580 that point, all the memory associated with it is released for reuse, all 581 files associated with it are closed, and any other clean-up functions which 582 are associated with the pool are run. When this is over, we can be confident 583 that all the resource tied to the pool have been released, and that none of 584 them have leaked.</p> 585 586 <p>Server restarts, and allocation of memory and resources for per-server 587 configuration, are handled in a similar way. There is a <em>configuration 588 pool</em>, which keeps track of resources which were allocated while reading 589 the server configuration files, and handling the commands therein (for 590 instance, the memory that was allocated for per-server module configuration, 591 log files and other files that were opened, and so forth). When the server 592 restarts, and has to reread the configuration files, the configuration pool 593 is cleared, and so the memory and file descriptors which were taken up by 594 reading them the last time are made available for reuse.</p> 595 596 <p>It should be noted that use of the pool machinery isn't generally 597 obligatory, except for situations like logging handlers, where you really 598 need to register cleanups to make sure that the log file gets closed when 599 the server restarts (this is most easily done by using the function <code><a href="#pool-files">ap_pfopen</a></code>, which also arranges for the 600 underlying file descriptor to be closed before any child processes, such as 601 for CGI scripts, are <code>exec</code>ed), or in case you are using the 602 timeout machinery (which isn't yet even documented here). However, there are 603 two benefits to using it: resources allocated to a pool never leak (even if 604 you allocate a scratch string, and just forget about it); also, for memory 605 allocation, <code>ap_palloc</code> is generally faster than 606 <code>malloc</code>.</p> 607 608 <p>We begin here by describing how memory is allocated to pools, and then 609 discuss how other resources are tracked by the resource pool machinery.</p> 610 611 <h3>Allocation of memory in pools</h3> 612 <p>Memory is allocated to pools by calling the function 613 <code>ap_palloc</code>, which takes two arguments, one being a pointer to 614 a resource pool structure, and the other being the amount of memory to 615 allocate (in <code>char</code>s). Within handlers for handling requests, 616 the most common way of getting a resource pool structure is by looking at 617 the <code>pool</code> slot of the relevant <code>request_rec</code>; hence 618 the repeated appearance of the following idiom in module code:</p> 619 620 <div class="example"><p><code> 621 int my_handler(request_rec *r)<br /> 622 {<br /> 623 <span class="indent"> 624 struct my_structure *foo;<br /> 625 ...<br /> 626 <br /> 627 foo = (foo *)ap_palloc (r->pool, sizeof(my_structure));<br /> 628 </span> 629 } 630 </code></p></div> 631 632 <p>Note that <em>there is no <code>ap_pfree</code></em> -- 633 <code>ap_palloc</code>ed memory is freed only when the associated resource 634 pool is cleared. This means that <code>ap_palloc</code> does not have to 635 do as much accounting as <code>malloc()</code>; all it does in the typical 636 case is to round up the size, bump a pointer, and do a range check.</p> 637 638 <p>(It also raises the possibility that heavy use of 639 <code>ap_palloc</code> could cause a server process to grow excessively 640 large. There are two ways to deal with this, which are dealt with below; 641 briefly, you can use <code>malloc</code>, and try to be sure that all of 642 the memory gets explicitly <code>free</code>d, or you can allocate a 643 sub-pool of the main pool, allocate your memory in the sub-pool, and clear 644 it out periodically. The latter technique is discussed in the section 645 on sub-pools below, and is used in the directory-indexing code, in order 646 to avoid excessive storage allocation when listing directories with 647 thousands of files).</p> 648 649 650 <h3>Allocating initialized memory</h3> 651 <p>There are functions which allocate initialized memory, and are 652 frequently useful. The function <code>ap_pcalloc</code> has the same 653 interface as <code>ap_palloc</code>, but clears out the memory it 654 allocates before it returns it. The function <code>ap_pstrdup</code> 655 takes a resource pool and a <code>char *</code> as arguments, and 656 allocates memory for a copy of the string the pointer points to, returning 657 a pointer to the copy. Finally <code>ap_pstrcat</code> is a varargs-style 658 function, which takes a pointer to a resource pool, and at least two 659 <code>char *</code> arguments, the last of which must be 660 <code>NULL</code>. It allocates enough memory to fit copies of each of 661 the strings, as a unit; for instance:</p> 662 663 <div class="example"><p><code> 664 ap_pstrcat (r->pool, "foo", "/", "bar", NULL); 665 </code></p></div> 666 667 <p>returns a pointer to 8 bytes worth of memory, initialized to 668 <code>"foo/bar"</code>.</p> 669 670 671 <h3><a name="pools-used" id="pools-used">Commonly-used pools in the Apache Web 672 server</a></h3> 673 <p>A pool is really defined by its lifetime more than anything else. 674 There are some static pools in http_main which are passed to various 675 non-http_main functions as arguments at opportune times. Here they 676 are:</p> 677 678 <dl> 679 <dt><code>permanent_pool</code></dt> 680 <dd>never passed to anything else, this is the ancestor of all pools</dd> 681 682 <dt><code>pconf</code></dt> 683 <dd> 684 <ul> 685 <li>subpool of permanent_pool</li> 686 687 <li>created at the beginning of a config "cycle"; exists 688 until the server is terminated or restarts; passed to all 689 config-time routines, either via cmd->pool, or as the 690 "pool *p" argument on those which don't take pools</li> 691 692 <li>passed to the module init() functions</li> 693 </ul> 694 </dd> 695 696 <dt><code>ptemp</code></dt> 697 <dd> 698 <ul> 699 <li>sorry I lie, this pool isn't called this currently in 700 1.3, I renamed it this in my pthreads development. I'm 701 referring to the use of ptrans in the parent... contrast 702 this with the later definition of ptrans in the 703 child.</li> 704 705 <li>subpool of permanent_pool</li> 706 707 <li>created at the beginning of a config "cycle"; exists 708 until the end of config parsing; passed to config-time 709 routines <em>via</em> cmd->temp_pool. Somewhat of a 710 "bastard child" because it isn't available everywhere. 711 Used for temporary scratch space which may be needed by 712 some config routines but which is deleted at the end of 713 config.</li> 714 </ul> 715 </dd> 716 717 <dt><code>pchild</code></dt> 718 <dd> 719 <ul> 720 <li>subpool of permanent_pool</li> 721 722 <li>created when a child is spawned (or a thread is 723 created); lives until that child (thread) is 724 destroyed</li> 725 726 <li>passed to the module child_init functions</li> 727 728 <li>destruction happens right after the child_exit 729 functions are called... (which may explain why I think 730 child_exit is redundant and unneeded)</li> 731 </ul> 732 </dd> 733 734 <dt><code>ptrans</code></dt> 735 <dd> 736 <ul> 737 <li>should be a subpool of pchild, but currently is a 738 subpool of permanent_pool, see above</li> 739 740 <li>cleared by the child before going into the accept() 741 loop to receive a connection</li> 742 743 <li>used as connection->pool</li> 744 </ul> 745 </dd> 746 747 <dt><code>r->pool</code></dt> 748 <dd> 749 <ul> 750 <li>for the main request this is a subpool of 751 connection->pool; for subrequests it is a subpool of 752 the parent request's pool.</li> 753 754 <li>exists until the end of the request (<em>i.e.</em>, 755 ap_destroy_sub_req, or in child_main after 756 process_request has finished)</li> 757 758 <li>note that r itself is allocated from r->pool; 759 <em>i.e.</em>, r->pool is first created and then r is 760 the first thing palloc()d from it</li> 761 </ul> 762 </dd> 763 </dl> 764 765 <p>For almost everything folks do, <code>r->pool</code> is the pool to 766 use. But you can see how other lifetimes, such as pchild, are useful to 767 some modules... such as modules that need to open a database connection 768 once per child, and wish to clean it up when the child dies.</p> 769 770 <p>You can also see how some bugs have manifested themself, such as 771 setting <code>connection->user</code> to a value from 772 <code>r->pool</code> -- in this case connection exists for the 773 lifetime of <code>ptrans</code>, which is longer than 774 <code>r->pool</code> (especially if <code>r->pool</code> is a 775 subrequest!). So the correct thing to do is to allocate from 776 <code>connection->pool</code>.</p> 777 778 <p>And there was another interesting bug in <code class="module"><a href="/mod/mod_include.html">mod_include</a></code> 779 / <code class="module"><a href="/mod/mod_cgi.html">mod_cgi</a></code>. You'll see in those that they do this test 780 to decide if they should use <code>r->pool</code> or 781 <code>r->main->pool</code>. In this case the resource that they are 782 registering for cleanup is a child process. If it were registered in 783 <code>r->pool</code>, then the code would <code>wait()</code> for the 784 child when the subrequest finishes. With <code class="module"><a href="/mod/mod_include.html">mod_include</a></code> this 785 could be any old <code>#include</code>, and the delay can be up to 3 786 seconds... and happened quite frequently. Instead the subprocess is 787 registered in <code>r->main->pool</code> which causes it to be 788 cleaned up when the entire request is done -- <em>i.e.</em>, after the 789 output has been sent to the client and logging has happened.</p> 790 791 792 <h3><a name="pool-files" id="pool-files">Tracking open files, etc.</a></h3> 793 <p>As indicated above, resource pools are also used to track other sorts 794 of resources besides memory. The most common are open files. The routine 795 which is typically used for this is <code>ap_pfopen</code>, which takes a 796 resource pool and two strings as arguments; the strings are the same as 797 the typical arguments to <code>fopen</code>, <em>e.g.</em>,</p> 798 799 <div class="example"><p><code> 800 ...<br /> 801 FILE *f = ap_pfopen (r->pool, r->filename, "r");<br /> 802 <br /> 803 if (f == NULL) { ... } else { ... }<br /> 804 </code></p></div> 805 806 <p>There is also a <code>ap_popenf</code> routine, which parallels the 807 lower-level <code>open</code> system call. Both of these routines arrange 808 for the file to be closed when the resource pool in question is 809 cleared.</p> 810 811 <p>Unlike the case for memory, there <em>are</em> functions to close files 812 allocated with <code>ap_pfopen</code>, and <code>ap_popenf</code>, namely 813 <code>ap_pfclose</code> and <code>ap_pclosef</code>. (This is because, on 814 many systems, the number of files which a single process can have open is 815 quite limited). It is important to use these functions to close files 816 allocated with <code>ap_pfopen</code> and <code>ap_popenf</code>, since to 817 do otherwise could cause fatal errors on systems such as Linux, which 818 react badly if the same <code>FILE*</code> is closed more than once.</p> 819 820 <p>(Using the <code>close</code> functions is not mandatory, since the 821 file will eventually be closed regardless, but you should consider it in 822 cases where your module is opening, or could open, a lot of files).</p> 823 824 825 <h3>Other sorts of resources -- cleanup functions</h3> 826 <p>More text goes here. Describe the cleanup primitives in terms of 827 which the file stuff is implemented; also, <code>spawn_process</code>.</p> 828 829 <p>Pool cleanups live until <code>clear_pool()</code> is called: 830 <code>clear_pool(a)</code> recursively calls <code>destroy_pool()</code> 831 on all subpools of <code>a</code>; then calls all the cleanups for 832 <code>a</code>; then releases all the memory for <code>a</code>. 833 <code>destroy_pool(a)</code> calls <code>clear_pool(a)</code> and then 834 releases the pool structure itself. <em>i.e.</em>, 835 <code>clear_pool(a)</code> doesn't delete <code>a</code>, it just frees 836 up all the resources and you can start using it again immediately.</p> 837 838 839 <h3>Fine control -- creating and dealing with sub-pools, with 840 a note on sub-requests</h3> 841 <p>On rare occasions, too-free use of <code>ap_palloc()</code> and the 842 associated primitives may result in undesirably profligate resource 843 allocation. You can deal with such a case by creating a <em>sub-pool</em>, 844 allocating within the sub-pool rather than the main pool, and clearing or 845 destroying the sub-pool, which releases the resources which were 846 associated with it. (This really <em>is</em> a rare situation; the only 847 case in which it comes up in the standard module set is in case of listing 848 directories, and then only with <em>very</em> large directories. 849 Unnecessary use of the primitives discussed here can hair up your code 850 quite a bit, with very little gain).</p> 851 852 <p>The primitive for creating a sub-pool is <code>ap_make_sub_pool</code>, 853 which takes another pool (the parent pool) as an argument. When the main 854 pool is cleared, the sub-pool will be destroyed. The sub-pool may also be 855 cleared or destroyed at any time, by calling the functions 856 <code>ap_clear_pool</code> and <code>ap_destroy_pool</code>, respectively. 857 (The difference is that <code>ap_clear_pool</code> frees resources 858 associated with the pool, while <code>ap_destroy_pool</code> also 859 deallocates the pool itself. In the former case, you can allocate new 860 resources within the pool, and clear it again, and so forth; in the 861 latter case, it is simply gone).</p> 862 863 <p>One final note -- sub-requests have their own resource pools, which are 864 sub-pools of the resource pool for the main request. The polite way to 865 reclaim the resources associated with a sub request which you have 866 allocated (using the <code>ap_sub_req_...</code> functions) is 867 <code>ap_destroy_sub_req</code>, which frees the resource pool. Before 868 calling this function, be sure to copy anything that you care about which 869 might be allocated in the sub-request's resource pool into someplace a 870 little less volatile (for instance, the filename in its 871 <code>request_rec</code> structure).</p> 872 873 <p>(Again, under most circumstances, you shouldn't feel obliged to call 874 this function; only 2K of memory or so are allocated for a typical sub 875 request, and it will be freed anyway when the main request pool is 876 cleared. It is only when you are allocating many, many sub-requests for a 877 single main request that you should seriously consider the 878 <code>ap_destroy_...</code> functions).</p> 879 880</div><div class="top"><a href="#page-header"><img alt="top" src="/images/up.gif" /></a></div> 881<div class="section"> 882<h2><a name="config" id="config">Configuration, commands and the like</a></h2> 883 <p>One of the design goals for this server was to maintain external 884 compatibility with the NCSA 1.3 server --- that is, to read the same 885 configuration files, to process all the directives therein correctly, and 886 in general to be a drop-in replacement for NCSA. On the other hand, another 887 design goal was to move as much of the server's functionality into modules 888 which have as little as possible to do with the monolithic server core. The 889 only way to reconcile these goals is to move the handling of most commands 890 from the central server into the modules.</p> 891 892 <p>However, just giving the modules command tables is not enough to divorce 893 them completely from the server core. The server has to remember the 894 commands in order to act on them later. That involves maintaining data which 895 is private to the modules, and which can be either per-server, or 896 per-directory. Most things are per-directory, including in particular access 897 control and authorization information, but also information on how to 898 determine file types from suffixes, which can be modified by 899 <code class="directive"><a href="/mod/mod_mime.html#addtype">AddType</a></code> and <code class="directive"><a href="/mod/core.html#defaulttype">DefaultType</a></code> directives, and so forth. In general, 900 the governing philosophy is that anything which <em>can</em> be made 901 configurable by directory should be; per-server information is generally 902 used in the standard set of modules for information like 903 <code class="directive"><a href="/mod/mod_alias.html#alias">Alias</a></code>es and <code class="directive"><a href="/mod/mod_alias.html#redirect">Redirect</a></code>s which come into play before the 904 request is tied to a particular place in the underlying file system.</p> 905 906 <p>Another requirement for emulating the NCSA server is being able to handle 907 the per-directory configuration files, generally called 908 <code>.htaccess</code> files, though even in the NCSA server they can 909 contain directives which have nothing at all to do with access control. 910 Accordingly, after URI -> filename translation, but before performing any 911 other phase, the server walks down the directory hierarchy of the underlying 912 filesystem, following the translated pathname, to read any 913 <code>.htaccess</code> files which might be present. The information which 914 is read in then has to be <em>merged</em> with the applicable information 915 from the server's own config files (either from the <code class="directive"><a href="/mod/core.html#directory"><Directory></a></code> sections in 916 <code>access.conf</code>, or from defaults in <code>srm.conf</code>, which 917 actually behaves for most purposes almost exactly like <code><Directory 918 /></code>).</p> 919 920 <p>Finally, after having served a request which involved reading 921 <code>.htaccess</code> files, we need to discard the storage allocated for 922 handling them. That is solved the same way it is solved wherever else 923 similar problems come up, by tying those structures to the per-transaction 924 resource pool.</p> 925 926 <h3><a name="per-dir" id="per-dir">Per-directory configuration structures</a></h3> 927 <p>Let's look out how all of this plays out in <code>mod_mime.c</code>, 928 which defines the file typing handler which emulates the NCSA server's 929 behavior of determining file types from suffixes. What we'll be looking 930 at, here, is the code which implements the <code class="directive"><a href="/mod/mod_mime.html#addtype">AddType</a></code> and <code class="directive"><a href="/mod/mod_mime.html#addencoding">AddEncoding</a></code> commands. These commands can appear in 931 <code>.htaccess</code> files, so they must be handled in the module's 932 private per-directory data, which in fact, consists of two separate 933 tables for MIME types and encoding information, and is declared as 934 follows:</p> 935 936 <div class="example"><pre>typedef struct { 937 table *forced_types; /* Additional AddTyped stuff */ 938 table *encoding_types; /* Added with AddEncoding... */ 939} mime_dir_config;</pre></div> 940 941 <p>When the server is reading a configuration file, or <code class="directive"><a href="/mod/core.html#directory"><Directory></a></code> section, which includes 942 one of the MIME module's commands, it needs to create a 943 <code>mime_dir_config</code> structure, so those commands have something 944 to act on. It does this by invoking the function it finds in the module's 945 `create per-dir config slot', with two arguments: the name of the 946 directory to which this configuration information applies (or 947 <code>NULL</code> for <code>srm.conf</code>), and a pointer to a 948 resource pool in which the allocation should happen.</p> 949 950 <p>(If we are reading a <code>.htaccess</code> file, that resource pool 951 is the per-request resource pool for the request; otherwise it is a 952 resource pool which is used for configuration data, and cleared on 953 restarts. Either way, it is important for the structure being created to 954 vanish when the pool is cleared, by registering a cleanup on the pool if 955 necessary).</p> 956 957 <p>For the MIME module, the per-dir config creation function just 958 <code>ap_palloc</code>s the structure above, and a creates a couple of 959 tables to fill it. That looks like this:</p> 960 961 <div class="example"><p><code> 962 void *create_mime_dir_config (pool *p, char *dummy)<br /> 963 {<br /> 964 <span class="indent"> 965 mime_dir_config *new =<br /> 966 <span class="indent"> 967 (mime_dir_config *) ap_palloc (p, sizeof(mime_dir_config));<br /> 968 </span> 969 <br /> 970 new->forced_types = ap_make_table (p, 4);<br /> 971 new->encoding_types = ap_make_table (p, 4);<br /> 972 <br /> 973 return new;<br /> 974 </span> 975 } 976 </code></p></div> 977 978 <p>Now, suppose we've just read in a <code>.htaccess</code> file. We 979 already have the per-directory configuration structure for the next 980 directory up in the hierarchy. If the <code>.htaccess</code> file we just 981 read in didn't have any <code class="directive"><a href="/mod/mod_mime.html#addtype">AddType</a></code> 982 or <code class="directive"><a href="/mod/mod_mime.html#addencoding">AddEncoding</a></code> commands, its 983 per-directory config structure for the MIME module is still valid, and we 984 can just use it. Otherwise, we need to merge the two structures 985 somehow.</p> 986 987 <p>To do that, the server invokes the module's per-directory config merge 988 function, if one is present. That function takes three arguments: the two 989 structures being merged, and a resource pool in which to allocate the 990 result. For the MIME module, all that needs to be done is overlay the 991 tables from the new per-directory config structure with those from the 992 parent:</p> 993 994 <div class="example"><p><code> 995 void *merge_mime_dir_configs (pool *p, void *parent_dirv, void *subdirv)<br /> 996 {<br /> 997 <span class="indent"> 998 mime_dir_config *parent_dir = (mime_dir_config *)parent_dirv;<br /> 999 mime_dir_config *subdir = (mime_dir_config *)subdirv;<br /> 1000 mime_dir_config *new =<br /> 1001 <span class="indent"> 1002 (mime_dir_config *)ap_palloc (p, sizeof(mime_dir_config));<br /> 1003 </span> 1004 <br /> 1005 new->forced_types = ap_overlay_tables (p, subdir->forced_types,<br /> 1006 <span class="indent"> 1007 parent_dir->forced_types);<br /> 1008 </span> 1009 new->encoding_types = ap_overlay_tables (p, subdir->encoding_types,<br /> 1010 <span class="indent"> 1011 parent_dir->encoding_types);<br /> 1012 </span> 1013 <br /> 1014 return new;<br /> 1015 </span> 1016 } 1017 </code></p></div> 1018 1019 <p>As a note -- if there is no per-directory merge function present, the 1020 server will just use the subdirectory's configuration info, and ignore 1021 the parent's. For some modules, that works just fine (<em>e.g.</em>, for 1022 the includes module, whose per-directory configuration information 1023 consists solely of the state of the <code>XBITHACK</code>), and for those 1024 modules, you can just not declare one, and leave the corresponding 1025 structure slot in the module itself <code>NULL</code>.</p> 1026 1027 1028 <h3><a name="commands" id="commands">Command handling</a></h3> 1029 <p>Now that we have these structures, we need to be able to figure out how 1030 to fill them. That involves processing the actual <code class="directive"><a href="/mod/mod_mime.html#addtype">AddType</a></code> and <code class="directive"><a href="/mod/mod_mime.html#addencoding">AddEncoding</a></code> commands. To find commands, the server looks in 1031 the module's command table. That table contains information on how many 1032 arguments the commands take, and in what formats, where it is permitted, 1033 and so forth. That information is sufficient to allow the server to invoke 1034 most command-handling functions with pre-parsed arguments. Without further 1035 ado, let's look at the <code class="directive"><a href="/mod/mod_mime.html#addtype">AddType</a></code> 1036 command handler, which looks like this (the <code class="directive"><a href="/mod/mod_mime.html#addencoding">AddEncoding</a></code> command looks basically the same, and won't be 1037 shown here):</p> 1038 1039 <div class="example"><p><code> 1040 char *add_type(cmd_parms *cmd, mime_dir_config *m, char *ct, char *ext)<br /> 1041 {<br /> 1042 <span class="indent"> 1043 if (*ext == '.') ++ext;<br /> 1044 ap_table_set (m->forced_types, ext, ct);<br /> 1045 return NULL;<br /> 1046 </span> 1047 } 1048 </code></p></div> 1049 1050 <p>This command handler is unusually simple. As you can see, it takes 1051 four arguments, two of which are pre-parsed arguments, the third being the 1052 per-directory configuration structure for the module in question, and the 1053 fourth being a pointer to a <code>cmd_parms</code> structure. That 1054 structure contains a bunch of arguments which are frequently of use to 1055 some, but not all, commands, including a resource pool (from which memory 1056 can be allocated, and to which cleanups should be tied), and the (virtual) 1057 server being configured, from which the module's per-server configuration 1058 data can be obtained if required.</p> 1059 1060 <p>Another way in which this particular command handler is unusually 1061 simple is that there are no error conditions which it can encounter. If 1062 there were, it could return an error message instead of <code>NULL</code>; 1063 this causes an error to be printed out on the server's 1064 <code>stderr</code>, followed by a quick exit, if it is in the main config 1065 files; for a <code>.htaccess</code> file, the syntax error is logged in 1066 the server error log (along with an indication of where it came from), and 1067 the request is bounced with a server error response (HTTP error status, 1068 code 500).</p> 1069 1070 <p>The MIME module's command table has entries for these commands, which 1071 look like this:</p> 1072 1073 <div class="example"><p><code> 1074 command_rec mime_cmds[] = {<br /> 1075 <span class="indent"> 1076 { "AddType", add_type, NULL, OR_FILEINFO, TAKE2,<br /> 1077 <span class="indent">"a mime type followed by a file extension" },<br /></span> 1078 { "AddEncoding", add_encoding, NULL, OR_FILEINFO, TAKE2,<br /> 1079 <span class="indent"> 1080 "an encoding (<em>e.g.</em>, gzip), followed by a file extension" },<br /> 1081 </span> 1082 { NULL }<br /> 1083 </span> 1084 }; 1085 </code></p></div> 1086 1087 <p>The entries in these tables are:</p> 1088 <ul> 1089 <li>The name of the command</li> 1090 <li>The function which handles it</li> 1091 <li>a <code>(void *)</code> pointer, which is passed in the 1092 <code>cmd_parms</code> structure to the command handler --- 1093 this is useful in case many similar commands are handled by 1094 the same function.</li> 1095 1096 <li>A bit mask indicating where the command may appear. There 1097 are mask bits corresponding to each 1098 <code>AllowOverride</code> option, and an additional mask 1099 bit, <code>RSRC_CONF</code>, indicating that the command may 1100 appear in the server's own config files, but <em>not</em> in 1101 any <code>.htaccess</code> file.</li> 1102 1103 <li>A flag indicating how many arguments the command handler 1104 wants pre-parsed, and how they should be passed in. 1105 <code>TAKE2</code> indicates two pre-parsed arguments. Other 1106 options are <code>TAKE1</code>, which indicates one 1107 pre-parsed argument, <code>FLAG</code>, which indicates that 1108 the argument should be <code>On</code> or <code>Off</code>, 1109 and is passed in as a boolean flag, <code>RAW_ARGS</code>, 1110 which causes the server to give the command the raw, unparsed 1111 arguments (everything but the command name itself). There is 1112 also <code>ITERATE</code>, which means that the handler looks 1113 the same as <code>TAKE1</code>, but that if multiple 1114 arguments are present, it should be called multiple times, 1115 and finally <code>ITERATE2</code>, which indicates that the 1116 command handler looks like a <code>TAKE2</code>, but if more 1117 arguments are present, then it should be called multiple 1118 times, holding the first argument constant.</li> 1119 1120 <li>Finally, we have a string which describes the arguments 1121 that should be present. If the arguments in the actual config 1122 file are not as required, this string will be used to help 1123 give a more specific error message. (You can safely leave 1124 this <code>NULL</code>).</li> 1125 </ul> 1126 1127 <p>Finally, having set this all up, we have to use it. This is ultimately 1128 done in the module's handlers, specifically for its file-typing handler, 1129 which looks more or less like this; note that the per-directory 1130 configuration structure is extracted from the <code>request_rec</code>'s 1131 per-directory configuration vector by using the 1132 <code>ap_get_module_config</code> function.</p> 1133 1134 <div class="example"><p><code> 1135 int find_ct(request_rec *r)<br /> 1136 {<br /> 1137 <span class="indent"> 1138 int i;<br /> 1139 char *fn = ap_pstrdup (r->pool, r->filename);<br /> 1140 mime_dir_config *conf = (mime_dir_config *)<br /> 1141 <span class="indent"> 1142 ap_get_module_config(r->per_dir_config, &mime_module);<br /> 1143 </span> 1144 char *type;<br /> 1145 <br /> 1146 if (S_ISDIR(r->finfo.st_mode)) {<br /> 1147 <span class="indent"> 1148 r->content_type = DIR_MAGIC_TYPE;<br /> 1149 return OK;<br /> 1150 </span> 1151 }<br /> 1152 <br /> 1153 if((i=ap_rind(fn,'.')) < 0) return DECLINED;<br /> 1154 ++i;<br /> 1155 <br /> 1156 if ((type = ap_table_get (conf->encoding_types, &fn[i])))<br /> 1157 {<br /> 1158 <span class="indent"> 1159 r->content_encoding = type;<br /> 1160 <br /> 1161 /* go back to previous extension to try to use it as a type */<br /> 1162 fn[i-1] = '\0';<br /> 1163 if((i=ap_rind(fn,'.')) < 0) return OK;<br /> 1164 ++i;<br /> 1165 </span> 1166 }<br /> 1167 <br /> 1168 if ((type = ap_table_get (conf->forced_types, &fn[i])))<br /> 1169 {<br /> 1170 <span class="indent"> 1171 r->content_type = type;<br /> 1172 </span> 1173 }<br /> 1174 <br /> 1175 return OK; 1176 </span> 1177 } 1178 </code></p></div> 1179 1180 1181 <h3><a name="servconf" id="servconf">Side notes -- per-server configuration, 1182 virtual servers, <em>etc</em>.</a></h3> 1183 <p>The basic ideas behind per-server module configuration are basically 1184 the same as those for per-directory configuration; there is a creation 1185 function and a merge function, the latter being invoked where a virtual 1186 server has partially overridden the base server configuration, and a 1187 combined structure must be computed. (As with per-directory configuration, 1188 the default if no merge function is specified, and a module is configured 1189 in some virtual server, is that the base configuration is simply 1190 ignored).</p> 1191 1192 <p>The only substantial difference is that when a command needs to 1193 configure the per-server private module data, it needs to go to the 1194 <code>cmd_parms</code> data to get at it. Here's an example, from the 1195 alias module, which also indicates how a syntax error can be returned 1196 (note that the per-directory configuration argument to the command 1197 handler is declared as a dummy, since the module doesn't actually have 1198 per-directory config data):</p> 1199 1200 <div class="example"><p><code> 1201 char *add_redirect(cmd_parms *cmd, void *dummy, char *f, char *url)<br /> 1202 {<br /> 1203 <span class="indent"> 1204 server_rec *s = cmd->server;<br /> 1205 alias_server_conf *conf = (alias_server_conf *)<br /> 1206 <span class="indent"> 1207 ap_get_module_config(s->module_config,&alias_module);<br /> 1208 </span> 1209 alias_entry *new = ap_push_array (conf->redirects);<br /> 1210 <br /> 1211 if (!ap_is_url (url)) return "Redirect to non-URL";<br /> 1212 <br /> 1213 new->fake = f; new->real = url;<br /> 1214 return NULL;<br /> 1215 </span> 1216 } 1217 </code></p></div> 1218 1219</div></div> 1220<div class="bottomlang"> 1221<p><span>Available Languages: </span><a href="/en/developer/API.html" title="English"> en </a></p> 1222</div><div class="top"><a href="#page-header"><img src="/images/up.gif" alt="top" /></a></div><div class="section"><h2><a id="comments_section" name="comments_section">Comments</a></h2><div class="warning"><strong>Notice:</strong><br />This is not a Q&A section. Comments placed here should be pointed towards suggestions on improving the documentation or server, and may be removed again by our moderators if they are either implemented or considered invalid/off-topic. Questions on how to manage the Apache HTTP Server should be directed at either our IRC channel, #httpd, on Freenode, or sent to our <a href="http://httpd.apache.org/lists.html">mailing lists</a>.</div> 1223<script type="text/javascript"><!--//--><![CDATA[//><!-- 1224var comments_shortname = 'httpd'; 1225var comments_identifier = 'http://httpd.apache.org/docs/2.2/developer/API.html'; 1226(function(w, d) { 1227 if (w.location.hostname.toLowerCase() == "httpd.apache.org") { 1228 d.write('<div id="comments_thread"><\/div>'); 1229 var s = d.createElement('script'); 1230 s.type = 'text/javascript'; 1231 s.async = true; 1232 s.src = 'https://comments.apache.org/show_comments.lua?site=' + comments_shortname + '&page=' + comments_identifier; 1233 (d.getElementsByTagName('head')[0] || d.getElementsByTagName('body')[0]).appendChild(s); 1234 } 1235 else { 1236 d.write('<div id="comments_thread">Comments are disabled for this page at the moment.<\/div>'); 1237 } 1238})(window, document); 1239//--><!]]></script></div><div id="footer"> 1240<p class="apache">Copyright 2013 The Apache Software Foundation.<br />Licensed under the <a href="http://www.apache.org/licenses/LICENSE-2.0">Apache License, Version 2.0</a>.</p> 1241<p class="menu"><a href="/mod/">Modules</a> | <a href="/mod/directives.html">Directives</a> | <a href="http://wiki.apache.org/httpd/FAQ">FAQ</a> | <a href="/glossary.html">Glossary</a> | <a href="/sitemap.html">Sitemap</a></p></div><script type="text/javascript"><!--//--><![CDATA[//><!-- 1242if (typeof(prettyPrint) !== 'undefined') { 1243 prettyPrint(); 1244} 1245//--><!]]></script> 1246</body></html>