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
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">
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 2.5</p>
19 <img alt="" src="../images/feather.gif" /></div>
20 <div class="up"><a href="./"><img title="<-" alt="<-" src="../images/left.gif" /></a></div>
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 2.5</a> > <a href="./">Developer Documentation</a></div><div id="page-content"><div id="preamble"><h1>Apache 1.3 API notes</h1>
24 <p><span>Available Languages: </span><a href="../en/developer/API.html" title="English"> en </a></p>
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>
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
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
49 <p>Finally, here's an outline, to give you some bare idea of what's coming
50 up, and in what order:</p>
54 <a href="#basics">Basic concepts.</a>
57 <li><a href="#HMR">Handlers, Modules, and
60 <li><a href="#moduletour">A brief tour of a
66 <a href="#handlers">How handlers work</a>
69 <li><a href="#req_tour">A brief tour of the
70 <code>request_rec</code></a></li>
72 <li><a href="#req_orig">Where request_rec structures come
75 <li><a href="#req_return">Handling requests, declining,
76 and returning error codes</a></li>
78 <li><a href="#resp_handlers">Special considerations for
79 response handlers</a></li>
81 <li><a href="#auth_handlers">Special considerations for
82 authentication handlers</a></li>
84 <li><a href="#log_handlers">Special considerations for
85 logging handlers</a></li>
89 <li><a href="#pools">Resource allocation and resource
93 <a href="#config">Configuration, commands and the like</a>
96 <li><a href="#per-dir">Per-directory configuration
99 <li><a href="#commands">Command handling</a></li>
101 <li><a href="#servconf">Side notes --- per-server
102 configuration, virtual servers, <em>etc</em>.</a></li>
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>
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>
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>
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
142 <li><em>Handle</em> the request, and indicate that it has done so by
143 returning the magic constant <code>OK</code>.</li>
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>
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>
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
166 <p>The handlers themselves are functions of one argument (a
167 <code>request_rec</code> structure. vide infra), which returns an integer,
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>
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>
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>
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>
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
209 <div class="example"><p><code>
210 /* Declarations of handlers. */<br />
212 int translate_scriptalias (request_rec *);<br />
213 int type_scriptalias (request_rec *);<br />
214 int cgi_handler (request_rec *);<br />
216 /* Subsidiary dispatch table for response-phase <br />
217 * handlers, by MIME type */<br />
219 handler_rec cgi_handlers[] = {<br />
220 <span class="indent">
221 { "application/x-httpd-cgi", cgi_handler },<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 />
233 void *make_cgi_server_config (pool *);<br />
234 void *merge_cgi_server_config (pool *, void *, void *);<br />
236 /* Declarations of routines to handle config-file commands */<br />
238 extern char *script_alias(cmd_parms *, void *per_dir_config, char *fake,
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>
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 */
265 NULL /* header parser */
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>
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>
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,
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>
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";
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>
317 <p>Here is an abridged declaration, giving the fields most commonly
320 <div class="example"><p><code>
321 struct request_rec {<br />
324 conn_rec *connection;<br />
325 server_rec *server;<br />
327 /* What object is being requested */<br />
330 char *filename;<br />
332 </code></p><pre>char *args; /* QUERY_ARGS, if any */
333 struct 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 />
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 />
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
351 table *headers_in;<br />
352 table *headers_out;<br />
353 table *err_headers_out;<br />
354 table *subprocess_env;<br />
356 /* Info about the request itself... */<br />
358 </code></p><pre>int header_only; /* HEAD request, as opposed to GET */
359 char *protocol; /* Protocol, as given to us, or HTTP/0.9 */
360 char *method; /* GET, HEAD, POST, <em>etc.</em> */
361 int method_number; /* M_GET, M_POST, <em>etc.</em> */
364 /* Info for logging */<br />
366 char *the_request;<br />
367 int bytes_sent;<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 />
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 />
383 </code></p><pre>void *per_dir_config; /* Options set in config files, <em>etc.</em> */
384 void *request_config; /* Notes on *this* request */</pre><p><code>
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
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
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>
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
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
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>
430 <h3><a name="req_return" id="req_return">Handling requests, declining, and returning
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>
437 <li><code>OK</code> -- the request was handled successfully. This may or
438 may not terminate the phase.</li>
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>
443 <li>an HTTP error code, which aborts handling of the request.</li>
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>
452 <h3><a name="resp_handlers" id="resp_handlers">Special considerations for response
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>
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>
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>
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
483 <div class="example"><p><code>
484 int default_handler (request_rec *r)<br />
486 <span class="indent">
490 if (r->method_number != M_GET) return DECLINED;<br />
491 if (r->finfo.st_mode == 0) return NOT_FOUND;<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 />
498 f = fopen (r->filename, "r");<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 />
507 register_timeout ("send", r);<br />
508 ap_send_http_header (r);<br />
510 if (!r->header_only) send_fd (f, r);<br />
511 ap_pfclose (r->pool, f);<br />
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>
526 <p>(Invoking <code>ap_internal_redirect</code> from handlers which are
527 <em>not</em> response handlers will lead to serious confusion).</p>
530 <h3><a name="auth_handlers" id="auth_handlers">Special considerations for authentication
532 <p>Stuff that should be discussed here in detail:</p>
535 <li>Authentication-phase handlers not invoked unless auth is
536 configured for the directory.</li>
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>
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>
552 <h3><a name="log_handlers" id="log_handlers">Special considerations for logging
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
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>
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>
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>
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>
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>
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>
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>
620 <div class="example"><p><code>
621 int my_handler(request_rec *r)<br />
623 <span class="indent">
624 struct my_structure *foo;<br />
627 foo = (foo *)ap_palloc (r->pool, sizeof(my_structure));<br />
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>
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>
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>
663 <div class="example"><p><code>
664 ap_pstrcat (r->pool, "foo", "/", "bar", NULL);
667 <p>returns a pointer to 8 bytes worth of memory, initialized to
668 <code>"foo/bar"</code>.</p>
671 <h3><a name="pools-used" id="pools-used">Commonly-used pools in the Apache Web
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
679 <dt><code>permanent_pool</code></dt>
680 <dd>never passed to anything else, this is the ancestor of all pools</dd>
682 <dt><code>pconf</code></dt>
685 <li>subpool of permanent_pool</li>
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>
692 <li>passed to the module init() functions</li>
696 <dt><code>ptemp</code></dt>
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
705 <li>subpool of permanent_pool</li>
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
717 <dt><code>pchild</code></dt>
720 <li>subpool of permanent_pool</li>
722 <li>created when a child is spawned (or a thread is
723 created); lives until that child (thread) is
726 <li>passed to the module child_init functions</li>
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>
734 <dt><code>ptrans</code></dt>
737 <li>should be a subpool of pchild, but currently is a
738 subpool of permanent_pool, see above</li>
740 <li>cleared by the child before going into the accept()
741 loop to receive a connection</li>
743 <li>used as connection->pool</li>
747 <dt><code>r->pool</code></dt>
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>
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>
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>
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>
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>
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>
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>
799 <div class="example"><p><code>
801 FILE *f = ap_pfopen (r->pool, r->filename, "r");<br />
803 if (f == NULL) { ... } else { ... }<br />
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
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>
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>
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>
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>
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>
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>
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>
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>
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>
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#forcetype">ForceType</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>
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
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
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
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>
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>
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
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>
961 <div class="example"><p><code>
962 void *create_mime_dir_config (pool *p, char *dummy)<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 />
970 new->forced_types = ap_make_table (p, 4);<br />
971 new->encoding_types = ap_make_table (p, 4);<br />
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
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
994 <div class="example"><p><code>
995 void *merge_mime_dir_configs (pool *p, void *parent_dirv, void *subdirv)<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 />
1005 new->forced_types = ap_overlay_tables (p, subdir->forced_types,<br />
1006 <span class="indent">
1007 parent_dir->forced_types);<br />
1009 new->encoding_types = ap_overlay_tables (p, subdir->encoding_types,<br />
1010 <span class="indent">
1011 parent_dir->encoding_types);<br />
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>
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
1039 <div class="example"><p><code>
1040 char *add_type(cmd_parms *cmd, mime_dir_config *m, char *ct, char *ext)<br />
1042 <span class="indent">
1043 if (*ext == '.') ++ext;<br />
1044 ap_table_set (m->forced_types, ext, ct);<br />
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>
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,
1070 <p>The MIME module's command table has entries for these commands, which
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 />
1087 <p>The entries in these tables are:</p>
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>
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>
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>
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>
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>
1134 <div class="example"><p><code>
1135 int find_ct(request_rec *r)<br />
1137 <span class="indent">
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 />
1146 if (S_ISDIR(r->finfo.st_mode)) {<br />
1147 <span class="indent">
1148 r->content_type = DIR_MAGIC_TYPE;<br />
1153 if((i=ap_rind(fn,'.')) < 0) return DECLINED;<br />
1156 if ((type = ap_table_get (conf->encoding_types, &fn[i])))<br />
1158 <span class="indent">
1159 r->content_encoding = type;<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 />
1168 if ((type = ap_table_get (conf->forced_types, &fn[i])))<br />
1170 <span class="indent">
1171 r->content_type = type;<br />
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
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>
1200 <div class="example"><p><code>
1201 char *add_redirect(cmd_parms *cmd, void *dummy, char *f, char *url)<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 />
1209 alias_entry *new = ap_push_array (conf->redirects);<br />
1211 if (!ap_is_url (url)) return "Redirect to non-URL";<br />
1213 new->fake = f; new->real = url;<br />
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>This section is experimental!</strong><br />Comments placed here should not be expected
1223 to last beyond the testing phase of this system, nor do we in any way guarantee that we'll read them.</div>
1224 <div id="disqus_thread">
1225 </div><script type="text/javascript"><!--//--><![CDATA[//><!--
1227 var disqus_shortname = 'httpd';
1228 var disqus_identifier = window.location.href.replace(/(current|trunk)/, "2.4").replace(/\/[a-z]{2}\//, "/").replace(window.location.protocol, "http:") + '.' + lang;
1229 if (disqus_identifier.indexOf("httpd.apache.org") != -1) {
1231 var dsq = document.createElement('script'); dsq.type = 'text/javascript'; dsq.async = true;
1232 dsq.src = window.location.protocol + '//' + disqus_shortname + '.disqus.com/embed.js';
1233 (document.getElementsByTagName('head')[0] || document.getElementsByTagName('body')[0]).appendChild(dsq);
1236 var text = document.createTextNode("Comments have been disabled for offline viewing.");
1237 document.getElementById('disqus_thread').appendChild(text);
1239 //--><!]]></script></div><div id="footer">
1240 <p class="apache">Copyright 2012 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[//><!--
1242 if (typeof(prettyPrint) !== undefined) {