request.c

Apache HTTP Server 是一个功能强大的灵活的与HTTP/1.1相兼容的web服务器.这里给出的是Apache HTTP服务器的源码。

         *     ordered by number of segments. See core_reorder_directories
         * startseg tells us how many segments describe the root path
         *     e.g. the complete path "//host/foo/" to a UNC share (4)
         */
        startseg = seg = ap_count_dirs(r->filename);
        sec_idx = 0;

        /*
         * Go down the directory hierarchy.  Where we have to check for
         * symlinks, do so.  Where a .htaccess file has permission to
         * override anything, try to find one.
         */
        do {
            int res;
            char *seg_name;
            char *delim;
            int temp_slash=0;

            /* We have no trailing slash, but we sure would appreciate one.
             * However, we don't want to append a / our first time through.
             */
            if ((seg > startseg) && r->filename[filename_len-1] != '/') {
                r->filename[filename_len++] = '/';
                r->filename[filename_len] = 0;
                temp_slash=1;
            }

            /* Begin *this* level by looking for matching <Directory> sections
             * from the server config.
             */
            for (; sec_idx < num_sec; ++sec_idx) {

                ap_conf_vector_t *entry_config = sec_ent[sec_idx];
                core_dir_config *entry_core;
                entry_core = ap_get_module_config(entry_config, &core_module);

                /* No more possible matches for this many segments?
                 * We are done when we find relative/regex/longer components.
                 */
                if (entry_core->r || entry_core->d_components > seg) {
                    break;
                }

                /* We will never skip '0' element components, e.g. plain old
                 * <Directory >, and <Directory "/"> are classified as zero
                 * so that Win32/Netware/OS2 etc all pick them up.
                 * Otherwise, skip over the mismatches.
                 */
                if (entry_core->d_components
                    && ((entry_core->d_components < seg)
                     || (entry_core->d_is_fnmatch
                         ? (apr_fnmatch(entry_core->d, r->filename,
                                        FNM_PATHNAME) != APR_SUCCESS)
                         : (strcmp(r->filename, entry_core->d) != 0)))) {
                    continue;
                }

                /* If we haven't continue'd above, we have a match.
                 *
                 * Calculate our full-context core opts & override.
                 */
                core_opts_merge(sec_ent[sec_idx], &opts);

                /* If we merged this same section last time, reuse it
                 */
                if (matches) {
                    if (last_walk->matched == sec_ent[sec_idx]) {
                        now_merged = last_walk->merged;
                        ++last_walk;
                        --matches;
                        continue;
                    }

                    /* We fell out of sync.  This is our own copy of walked,
                     * so truncate the remaining matches and reset remaining.
                     */
                    cache->walked->nelts -= matches;
                    matches = 0;
                }

                if (now_merged) {
                    now_merged = ap_merge_per_dir_configs(r->pool,
                                                          now_merged,
                                                          sec_ent[sec_idx]);
                }
                else {
                    now_merged = sec_ent[sec_idx];
                }

                last_walk = (walk_walked_t*)apr_array_push(cache->walked);
                last_walk->matched = sec_ent[sec_idx];
                last_walk->merged = now_merged;
            }

            /* If .htaccess files are enabled, check for one, provided we
             * have reached a real path.
             */
            do {  /* Not really a loop, just a break'able code block */

                ap_conf_vector_t *htaccess_conf = NULL;

                /* No htaccess in an incomplete root path, 
                 * nor if it's disabled 
                 */
                if (seg < startseg || !opts.override) {
                    break;
                }

                res = ap_parse_htaccess(&htaccess_conf, r, opts.override,
                                        apr_pstrdup(r->pool, r->filename),
                                        sconf->access_name);
                if (res) {
                    return res;
                }

                if (!htaccess_conf) {
                    break;
                }

                /* If we are still here, we found our htaccess.
                 *
                 * Calculate our full-context core opts & override.
                 */
                core_opts_merge(htaccess_conf, &opts);

                /* If we merged this same htaccess last time, reuse it...
                 * this wouldn't work except that we cache the htaccess
                 * sections for the lifetime of the request, so we match
                 * the same conf.  Good planning (no, pure luck ;)
                 */
                if (matches) {
                    if (last_walk->matched == htaccess_conf) {
                        now_merged = last_walk->merged;
                        ++last_walk;
                        --matches;
                        break;
                    }

                    /* We fell out of sync.  This is our own copy of walked,
                     * so truncate the remaining matches and reset
                     * remaining.
                     */
                    cache->walked->nelts -= matches;
                    matches = 0;
                }

                if (now_merged) {
                    now_merged = ap_merge_per_dir_configs(r->pool,
                                                          now_merged,
                                                          htaccess_conf);
                }
                else {
                    now_merged = htaccess_conf;
                }

                last_walk = (walk_walked_t*)apr_array_push(cache->walked);
                last_walk->matched = htaccess_conf;
                last_walk->merged = now_merged;

            } while (0); /* Only one htaccess, not a real loop */

            /* That temporary trailing slash was useful, now drop it.
             */
            if (temp_slash) {
                r->filename[--filename_len] = '\0';
            }

            /* Time for all good things to come to an end?
             */
            if (!r->path_info || !*r->path_info) {
                break;
            }

            /* Now it's time for the next segment...
             * We will assume the next element is an end node, and fix it up
             * below as necessary...
             */

            seg_name = r->filename + filename_len;
            delim = strchr(r->path_info + (*r->path_info == '/' ? 1 : 0), '/');
            if (delim) {
                size_t path_info_len = delim - r->path_info;
                *delim = '\0';
                memcpy(seg_name, r->path_info, path_info_len + 1);
                filename_len += path_info_len;
                r->path_info = delim;
                *delim = '/';
            }
            else {
                size_t path_info_len = strlen(r->path_info);
                memcpy(seg_name, r->path_info, path_info_len + 1);
                filename_len += path_info_len;
                r->path_info += path_info_len;
            }
            if (*seg_name == '/')
                ++seg_name;

            /* If nothing remained but a '/' string, we are finished
             */
            if (!*seg_name) {
                break;
            }

            /* First optimization;
             * If...we knew r->filename was a file, and
             * if...we have strict (case-sensitive) filenames, or
             *      we know the canonical_filename matches to _this_ name, and
             * if...we have allowed symlinks
             * skip the lstat and dummy up an APR_DIR value for thisinfo.
             */
            if (r->finfo.filetype
#ifdef CASE_BLIND_FILESYSTEM
                && (filename_len <= canonical_len)
#endif
                && ((opts.opts & (OPT_SYM_OWNER | OPT_SYM_LINKS)) == OPT_SYM_LINKS))
            {

                thisinfo.filetype = APR_DIR;
                ++seg;
                continue;
            }

            /* We choose apr_lstat here, rather that apr_stat, so that we
             * capture this path object rather than its target.  We will
             * replace the info with our target's info below.  We especially
             * want the name of this 'link' object, not the name of its
             * target, if we are fixing the filename case/resolving aliases.
             */
            rv = apr_lstat(&thisinfo, r->filename,
                           APR_FINFO_MIN | APR_FINFO_NAME, r->pool);

            if (APR_STATUS_IS_ENOENT(rv)) {
                /* Nothing?  That could be nice.  But our directory
                 * walk is done.
                 */
                thisinfo.filetype = APR_NOFILE;
                break;
            }
            else if (APR_STATUS_IS_EACCES(rv)) {
                ap_log_rerror(APLOG_MARK, APLOG_ERR, rv, r,
                              "access to %s denied", r->uri);
                return r->status = HTTP_FORBIDDEN;
            }
            else if ((rv != APR_SUCCESS && rv != APR_INCOMPLETE)
                     || !(thisinfo.valid & APR_FINFO_TYPE)) {
                /* If we hit ENOTDIR, we must have over-optimized, deny
                 * rather than assume not found.
                 */
                ap_log_rerror(APLOG_MARK, APLOG_ERR, rv, r,
                              "access to %s failed", r->uri);
                return r->status = HTTP_FORBIDDEN;
            }

            /* Fix up the path now if we have a name, and they don't agree
             */
            if ((thisinfo.valid & APR_FINFO_NAME)
                && strcmp(seg_name, thisinfo.name)) {
                /* TODO: provide users an option that an internal/external
                 * redirect is required here?  We need to walk the URI and
                 * filename in tandem to properly correlate these.
                 */
                strcpy(seg_name, thisinfo.name);
                filename_len = strlen(r->filename);
            }

            if (thisinfo.filetype == APR_LNK) {
                /* Is this a possibly acceptable symlink?
                 */
                if ((res = resolve_symlink(r->filename, &thisinfo,
                                           opts.opts, r->pool)) != OK) {
                    ap_log_rerror(APLOG_MARK, APLOG_ERR, 0, r,
                                  "Symbolic link not allowed: %s",
                                  r->filename);
                    return r->status = res;
                }
            }

            /* Ok, we are done with the link's info, test the real target
             */
            if (thisinfo.filetype == APR_REG || 
                thisinfo.filetype == APR_NOFILE) {
                /* That was fun, nothing left for us here
                 */
                break;
            }
            else if (thisinfo.filetype != APR_DIR) {
                ap_log_rerror(APLOG_MARK, APLOG_ERR, 0, r,
                              "Forbidden: %s doesn't point to "
                              "a file or directory",
                              r->filename);
                return r->status = HTTP_FORBIDDEN;
            }

            ++seg;
        } while (thisinfo.filetype == APR_DIR);

        /* If we have _not_ optimized, this is the time to recover
         * the final stat result.
         */
        if (!r->finfo.filetype || r->finfo.filetype == APR_LNK) {
            r->finfo = thisinfo;
        }

        /* Now splice the saved path_info back onto any new path_info
         */
        if (save_path_info) {
            if (r->path_info && *r->path_info) {
                r->path_info = ap_make_full_path(r->pool, r->path_info,
                                                 save_path_info);
            }
            else {
                r->path_info = save_path_info;
            }
        }

        /*
         * Now we'll deal with the regexes, note we pick up sec_idx
         * where we left off (we gave up after we hit entry_core->r)
         */
        for (; sec_idx < num_sec; ++sec_idx) {

            core_dir_config *entry_core;
            entry_core = ap_get_module_config(sec_ent[sec_idx], &core_module);

            if (!entry_core->r) {
                continue;
            }

            if (ap_regexec(entry_core->r, r->filename, 0, NULL, REG_NOTEOL)) {
                continue;
            }

            /* If we haven't already continue'd above, we have a match.
             *
             * Calculate our full-context core opts & override.
             */
            core_opts_merge(sec_ent[sec_idx], &opts);

            /* If we merged this same section last time, reuse it
             */
            if (matches) {
                if (last_walk->matched == sec_ent[sec_idx]) {
                    now_merged = last_walk->merged;
                    ++last_walk;
                    --matches;
                    continue;
                }

                /* We fell out of sync.  This is our own copy of walked,
                 * so truncate the remaining matches and reset remaining.
                 */
                cache->walked->nelts -= matches;
                matches = 0;
            }

            if (now_merged) {
                now_merged = ap_merge_per_dir_configs(r->pool,