1/* Licensed to the Apache Software Foundation (ASF) under one or more
2 * contributor license agreements.  See the NOTICE file distributed with
3 * this work for additional information regarding copyright ownership.
4 * The ASF licenses this file to You under the Apache License, Version 2.0
5 * (the "License"); you may not use this file except in compliance with
6 * the License.  You may obtain a copy of the License at
7 *
8 *     http://www.apache.org/licenses/LICENSE-2.0
9 *
10 * Unless required by applicable law or agreed to in writing, software
11 * distributed under the License is distributed on an "AS IS" BASIS,
12 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
13 * See the License for the specific language governing permissions and
14 * limitations under the License.
15 */
16
17/*
18 * http_alias.c: Stuff for dealing with directory aliases
19 *
20 * Original by Rob McCool, rewritten in succession by David Robinson
21 * and rst.
22 *
23 */
24
25#include "apr_strings.h"
26#include "apr_lib.h"
27
28#define APR_WANT_STRFUNC
29#include "apr_want.h"
30
31#include "ap_config.h"
32#include "httpd.h"
33#include "http_core.h"
34#include "http_config.h"
35#include "http_request.h"
36#include "http_log.h"
37
38
39typedef struct {
40    const char *real;
41    const char *fake;
42    char *handler;
43    ap_regex_t *regexp;
44    int redir_status;                /* 301, 302, 303, 410, etc */
45} alias_entry;
46
47typedef struct {
48    apr_array_header_t *aliases;
49    apr_array_header_t *redirects;
50} alias_server_conf;
51
52typedef struct {
53    apr_array_header_t *redirects;
54} alias_dir_conf;
55
56module AP_MODULE_DECLARE_DATA alias_module;
57
58static void *create_alias_config(apr_pool_t *p, server_rec *s)
59{
60    alias_server_conf *a =
61    (alias_server_conf *) apr_pcalloc(p, sizeof(alias_server_conf));
62
63    a->aliases = apr_array_make(p, 20, sizeof(alias_entry));
64    a->redirects = apr_array_make(p, 20, sizeof(alias_entry));
65    return a;
66}
67
68static void *create_alias_dir_config(apr_pool_t *p, char *d)
69{
70    alias_dir_conf *a =
71    (alias_dir_conf *) apr_pcalloc(p, sizeof(alias_dir_conf));
72    a->redirects = apr_array_make(p, 2, sizeof(alias_entry));
73    return a;
74}
75
76static void *merge_alias_config(apr_pool_t *p, void *basev, void *overridesv)
77{
78    alias_server_conf *a =
79    (alias_server_conf *) apr_pcalloc(p, sizeof(alias_server_conf));
80    alias_server_conf *base = (alias_server_conf *) basev;
81    alias_server_conf *overrides = (alias_server_conf *) overridesv;
82
83    a->aliases = apr_array_append(p, overrides->aliases, base->aliases);
84    a->redirects = apr_array_append(p, overrides->redirects, base->redirects);
85    return a;
86}
87
88static void *merge_alias_dir_config(apr_pool_t *p, void *basev, void *overridesv)
89{
90    alias_dir_conf *a =
91    (alias_dir_conf *) apr_pcalloc(p, sizeof(alias_dir_conf));
92    alias_dir_conf *base = (alias_dir_conf *) basev;
93    alias_dir_conf *overrides = (alias_dir_conf *) overridesv;
94    a->redirects = apr_array_append(p, overrides->redirects, base->redirects);
95    return a;
96}
97
98/* need prototype for overlap check */
99static int alias_matches(const char *uri, const char *alias_fakename);
100
101static const char *add_alias_internal(cmd_parms *cmd, void *dummy,
102                                      const char *f, const char *r,
103                                      int use_regex)
104{
105    server_rec *s = cmd->server;
106    alias_server_conf *conf = ap_get_module_config(s->module_config,
107                                                   &alias_module);
108    alias_entry *new = apr_array_push(conf->aliases);
109    alias_entry *entries = (alias_entry *)conf->aliases->elts;
110    int i;
111
112    /* XX r can NOT be relative to DocumentRoot here... compat bug. */
113
114    if (use_regex) {
115        new->regexp = ap_pregcomp(cmd->pool, f, AP_REG_EXTENDED);
116        if (new->regexp == NULL)
117            return "Regular expression could not be compiled.";
118        new->real = r;
119    }
120    else {
121        /* XXX This may be optimized, but we must know that new->real
122         * exists.  If so, we can dir merge later, trusing new->real
123         * and just canonicalizing the remainder.  Not till I finish
124         * cleaning out the old ap_canonical stuff first.
125         */
126        new->real = r;
127    }
128    new->fake = f;
129    new->handler = cmd->info;
130
131    /* check for overlapping (Script)Alias directives
132     * and throw a warning if found one
133     */
134    if (!use_regex) {
135        for (i = 0; i < conf->aliases->nelts - 1; ++i) {
136            alias_entry *p = &entries[i];
137
138            if (  (!p->regexp &&  alias_matches(f, p->fake) > 0)
139                || (p->regexp && !ap_regexec(p->regexp, f, 0, NULL, 0))) {
140                ap_log_error(APLOG_MARK, APLOG_WARNING, 0, cmd->server,
141                             "The %s directive in %s at line %d will probably "
142                             "never match because it overlaps an earlier "
143                             "%sAlias%s.",
144                             cmd->cmd->name, cmd->directive->filename,
145                             cmd->directive->line_num,
146                             p->handler ? "Script" : "",
147                             p->regexp ? "Match" : "");
148
149                break; /* one warning per alias should be sufficient */
150            }
151        }
152    }
153
154    return NULL;
155}
156
157static const char *add_alias(cmd_parms *cmd, void *dummy, const char *f,
158                             const char *r)
159{
160    return add_alias_internal(cmd, dummy, f, r, 0);
161}
162
163static const char *add_alias_regex(cmd_parms *cmd, void *dummy, const char *f,
164                                   const char *r)
165{
166    return add_alias_internal(cmd, dummy, f, r, 1);
167}
168
169static const char *add_redirect_internal(cmd_parms *cmd,
170                                         alias_dir_conf *dirconf,
171                                         const char *arg1, const char *arg2,
172                                         const char *arg3, int use_regex)
173{
174    alias_entry *new;
175    server_rec *s = cmd->server;
176    alias_server_conf *serverconf = ap_get_module_config(s->module_config,
177                                                         &alias_module);
178    int status = (int) (long) cmd->info;
179    int grokarg1 = 1;
180    ap_regex_t *r = NULL;
181    const char *f = arg2;
182    const char *url = arg3;
183
184    /*
185     * Logic flow:
186     *   Go ahead and try to grok the 1st arg, in case it is a
187     *   Redirect status. Now if we have 3 args, we expect that
188     *   we were able to understand that 1st argument (it's something
189     *   we expected, so if not, then we bail
190     */
191    if (!strcasecmp(arg1, "permanent"))
192        status = HTTP_MOVED_PERMANENTLY;
193    else if (!strcasecmp(arg1, "temp"))
194        status = HTTP_MOVED_TEMPORARILY;
195    else if (!strcasecmp(arg1, "seeother"))
196        status = HTTP_SEE_OTHER;
197    else if (!strcasecmp(arg1, "gone"))
198        status = HTTP_GONE;
199    else if (apr_isdigit(*arg1))
200        status = atoi(arg1);
201    else
202        grokarg1 = 0;
203
204    if (arg3 && !grokarg1)
205        return "Redirect: invalid first argument (of three)";
206
207    /*
208     * if we don't have the 3rd arg and we didn't understand the 1st
209     * one, then assume URL-path URL. This also handles case, eg, GONE
210     * we even though we don't have a 3rd arg, we did understand the 1st
211     * one, so we don't want to re-arrange
212     */
213    if (!arg3 && !grokarg1) {
214        f = arg1;
215        url = arg2;
216    }
217
218    if (use_regex) {
219        r = ap_pregcomp(cmd->pool, f, AP_REG_EXTENDED);
220        if (r == NULL)
221            return "Regular expression could not be compiled.";
222    }
223
224    if (ap_is_HTTP_REDIRECT(status)) {
225        if (!url)
226            return "URL to redirect to is missing";
227        /* PR#35314: we can allow path components here;
228         * they get correctly resolved to full URLs.
229         */
230        if (!use_regex && !ap_is_url(url) && (url[0] != '/'))
231            return "Redirect to non-URL";
232    }
233    else {
234        if (url)
235            return "Redirect URL not valid for this status";
236    }
237
238    if (cmd->path)
239        new = apr_array_push(dirconf->redirects);
240    else
241        new = apr_array_push(serverconf->redirects);
242
243    new->fake = f;
244    new->real = url;
245    new->regexp = r;
246    new->redir_status = status;
247    return NULL;
248}
249
250static const char *add_redirect(cmd_parms *cmd, void *dirconf,
251                                const char *arg1, const char *arg2,
252                                const char *arg3)
253{
254    return add_redirect_internal(cmd, dirconf, arg1, arg2, arg3, 0);
255}
256
257static const char *add_redirect2(cmd_parms *cmd, void *dirconf,
258                                 const char *arg1, const char *arg2)
259{
260    return add_redirect_internal(cmd, dirconf, arg1, arg2, NULL, 0);
261}
262
263static const char *add_redirect_regex(cmd_parms *cmd, void *dirconf,
264                                      const char *arg1, const char *arg2,
265                                      const char *arg3)
266{
267    return add_redirect_internal(cmd, dirconf, arg1, arg2, arg3, 1);
268}
269
270static const command_rec alias_cmds[] =
271{
272    AP_INIT_TAKE2("Alias", add_alias, NULL, RSRC_CONF,
273                  "a fakename and a realname"),
274    AP_INIT_TAKE2("ScriptAlias", add_alias, "cgi-script", RSRC_CONF,
275                  "a fakename and a realname"),
276    AP_INIT_TAKE23("Redirect", add_redirect, (void *) HTTP_MOVED_TEMPORARILY,
277                   OR_FILEINFO,
278                   "an optional status, then document to be redirected and "
279                   "destination URL"),
280    AP_INIT_TAKE2("AliasMatch", add_alias_regex, NULL, RSRC_CONF,
281                  "a regular expression and a filename"),
282    AP_INIT_TAKE2("ScriptAliasMatch", add_alias_regex, "cgi-script", RSRC_CONF,
283                  "a regular expression and a filename"),
284    AP_INIT_TAKE23("RedirectMatch", add_redirect_regex,
285                   (void *) HTTP_MOVED_TEMPORARILY, OR_FILEINFO,
286                   "an optional status, then a regular expression and "
287                   "destination URL"),
288    AP_INIT_TAKE2("RedirectTemp", add_redirect2,
289                  (void *) HTTP_MOVED_TEMPORARILY, OR_FILEINFO,
290                  "a document to be redirected, then the destination URL"),
291    AP_INIT_TAKE2("RedirectPermanent", add_redirect2,
292                  (void *) HTTP_MOVED_PERMANENTLY, OR_FILEINFO,
293                  "a document to be redirected, then the destination URL"),
294    {NULL}
295};
296
297static int alias_matches(const char *uri, const char *alias_fakename)
298{
299    const char *aliasp = alias_fakename, *urip = uri;
300
301    while (*aliasp) {
302        if (*aliasp == '/') {
303            /* any number of '/' in the alias matches any number in
304             * the supplied URI, but there must be at least one...
305             */
306            if (*urip != '/')
307                return 0;
308
309            do {
310                ++aliasp;
311            } while (*aliasp == '/');
312            do {
313                ++urip;
314            } while (*urip == '/');
315        }
316        else {
317            /* Other characters are compared literally */
318            if (*urip++ != *aliasp++)
319                return 0;
320        }
321    }
322
323    /* Check last alias path component matched all the way */
324
325    if (aliasp[-1] != '/' && *urip != '\0' && *urip != '/')
326        return 0;
327
328    /* Return number of characters from URI which matched (may be
329     * greater than length of alias, since we may have matched
330     * doubled slashes)
331     */
332
333    return urip - uri;
334}
335
336static char *try_alias_list(request_rec *r, apr_array_header_t *aliases,
337                            int doesc, int *status)
338{
339    alias_entry *entries = (alias_entry *) aliases->elts;
340    ap_regmatch_t regm[AP_MAX_REG_MATCH];
341    char *found = NULL;
342    int i;
343
344    for (i = 0; i < aliases->nelts; ++i) {
345        alias_entry *p = &entries[i];
346        int l;
347
348        if (p->regexp) {
349            if (!ap_regexec(p->regexp, r->uri, AP_MAX_REG_MATCH, regm, 0)) {
350                if (p->real) {
351                    found = ap_pregsub(r->pool, p->real, r->uri,
352                                       AP_MAX_REG_MATCH, regm);
353                    if (found && doesc) {
354                        apr_uri_t uri;
355                        apr_uri_parse(r->pool, found, &uri);
356                        /* Do not escape the query string or fragment. */
357                        found = apr_uri_unparse(r->pool, &uri,
358                                                APR_URI_UNP_OMITQUERY);
359                        found = ap_escape_uri(r->pool, found);
360                        if (uri.query) {
361                            found = apr_pstrcat(r->pool, found, "?",
362                                                uri.query, NULL);
363                        }
364                        if (uri.fragment) {
365                            found = apr_pstrcat(r->pool, found, "#",
366                                                uri.fragment, NULL);
367                        }
368                    }
369                }
370                else {
371                    /* need something non-null */
372                    found = apr_pstrdup(r->pool, "");
373                }
374            }
375        }
376        else {
377            l = alias_matches(r->uri, p->fake);
378
379            if (l > 0) {
380                if (doesc) {
381                    char *escurl;
382                    escurl = ap_os_escape_path(r->pool, r->uri + l, 1);
383
384                    found = apr_pstrcat(r->pool, p->real, escurl, NULL);
385                }
386                else
387                    found = apr_pstrcat(r->pool, p->real, r->uri + l, NULL);
388            }
389        }
390
391        if (found) {
392            if (p->handler) {    /* Set handler, and leave a note for mod_cgi */
393                r->handler = p->handler;
394                apr_table_setn(r->notes, "alias-forced-type", r->handler);
395            }
396            /* XXX This is as SLOW as can be, next step, we optimize
397             * and merge to whatever part of the found path was already
398             * canonicalized.  After I finish eliminating os canonical.
399             * Better fail test for ap_server_root_relative needed here.
400             */
401            if (!doesc) {
402                found = ap_server_root_relative(r->pool, found);
403            }
404            if (found) {
405                *status = p->redir_status;
406            }
407            return found;
408        }
409
410    }
411
412    return NULL;
413}
414
415static int translate_alias_redir(request_rec *r)
416{
417    ap_conf_vector_t *sconf = r->server->module_config;
418    alias_server_conf *serverconf = ap_get_module_config(sconf, &alias_module);
419    char *ret;
420    int status;
421
422    if (r->uri[0] != '/' && r->uri[0] != '\0') {
423        return DECLINED;
424    }
425
426    if ((ret = try_alias_list(r, serverconf->redirects, 1, &status)) != NULL) {
427        if (ap_is_HTTP_REDIRECT(status)) {
428            char *orig_target = ret;
429            if (ret[0] == '/') {
430
431                ret = ap_construct_url(r->pool, ret, r);
432                ap_log_rerror(APLOG_MARK, APLOG_DEBUG, 0, r,
433                              "incomplete redirection target of '%s' for "
434                              "URI '%s' modified to '%s'",
435                              orig_target, r->uri, ret);
436            }
437            if (!ap_is_url(ret)) {
438                ap_log_rerror(APLOG_MARK, APLOG_ERR, 0, r,
439                              "cannot redirect '%s' to '%s'; "
440                              "target is not a valid absoluteURI or abs_path",
441                              r->uri, ret);
442                /* restore the config value, so as not to get a
443                 * "regression" on existing "working" configs.
444                 */
445                ret = orig_target;
446            }
447            /* append requested query only, if the config didn't
448             * supply its own.
449             */
450            if (r->args && !ap_strchr(ret, '?')) {
451                ret = apr_pstrcat(r->pool, ret, "?", r->args, NULL);
452            }
453            apr_table_setn(r->headers_out, "Location", ret);
454        }
455        return status;
456    }
457
458    if ((ret = try_alias_list(r, serverconf->aliases, 0, &status)) != NULL) {
459        r->filename = ret;
460        return OK;
461    }
462
463    return DECLINED;
464}
465
466static int fixup_redir(request_rec *r)
467{
468    void *dconf = r->per_dir_config;
469    alias_dir_conf *dirconf =
470    (alias_dir_conf *) ap_get_module_config(dconf, &alias_module);
471    char *ret;
472    int status;
473
474    /* It may have changed since last time, so try again */
475
476    if ((ret = try_alias_list(r, dirconf->redirects, 1, &status)) != NULL) {
477        if (ap_is_HTTP_REDIRECT(status)) {
478            if (ret[0] == '/') {
479                char *orig_target = ret;
480
481                ret = ap_construct_url(r->pool, ret, r);
482                ap_log_rerror(APLOG_MARK, APLOG_DEBUG, 0, r,
483                              "incomplete redirection target of '%s' for "
484                              "URI '%s' modified to '%s'",
485                              orig_target, r->uri, ret);
486            }
487            if (!ap_is_url(ret)) {
488                status = HTTP_INTERNAL_SERVER_ERROR;
489                ap_log_rerror(APLOG_MARK, APLOG_ERR, 0, r,
490                              "cannot redirect '%s' to '%s'; "
491                              "target is not a valid absoluteURI or abs_path",
492                              r->uri, ret);
493            }
494            else {
495                /* append requested query only, if the config didn't
496                 * supply its own.
497                 */
498                if (r->args && !ap_strchr(ret, '?')) {
499                    ret = apr_pstrcat(r->pool, ret, "?", r->args, NULL);
500                }
501                apr_table_setn(r->headers_out, "Location", ret);
502            }
503        }
504        return status;
505    }
506
507    return DECLINED;
508}
509
510static void register_hooks(apr_pool_t *p)
511{
512    static const char * const aszSucc[]={ "mod_userdir.c",
513                                          "mod_vhost_alias.c",NULL };
514
515    ap_hook_translate_name(translate_alias_redir,NULL,aszSucc,APR_HOOK_MIDDLE);
516    ap_hook_fixups(fixup_redir,NULL,NULL,APR_HOOK_MIDDLE);
517}
518
519module AP_MODULE_DECLARE_DATA alias_module =
520{
521    STANDARD20_MODULE_STUFF,
522    create_alias_dir_config,       /* dir config creater */
523    merge_alias_dir_config,        /* dir merger --- default is to override */
524    create_alias_config,           /* server config */
525    merge_alias_config,            /* merge server configs */
526    alias_cmds,                    /* command apr_table_t */
527    register_hooks                 /* register hooks */
528};
529