mod_speling.c

apache简化版

    /* good = /correct-file */
    good = ap_pstrndup(r->pool, r->filename, filoc);
    /* bad = mispelling */
    bad = ap_pstrdup(r->pool, r->filename + filoc + 1);
    /* postgood = mispelling/more */
    postgood = ap_pstrcat(r->pool, bad, r->path_info, NULL);

    urlen = strlen(r->uri);
    pglen = strlen(postgood);

    /* Check to see if the URL pieces add up */
    if (strcmp(postgood, r->uri + (urlen - pglen))) {
        return DECLINED;
    }

    /* url = /correct-url */
    url = ap_pstrndup(r->pool, r->uri, (urlen - pglen));

    /* Now open the directory and do ourselves a check... */
    dirp = ap_popendir(r->pool, good);
    if (dirp == NULL) {          /* Oops, not a directory... */
        return DECLINED;
    }

    candidates = ap_make_array(r->pool, 2, sizeof(misspelled_file));

    dotloc = ap_ind(bad, '.');
    if (dotloc == -1) {
        dotloc = strlen(bad);
    }

    while ((dir_entry = readdir(dirp)) != NULL) {
        sp_reason q;

        /*
         * If we end up with a "fixed" URL which is identical to the
         * requested one, we must have found a broken symlink or some such.
         * Do _not_ try to redirect this, it causes a loop!
         */
        if (strcmp(bad, dir_entry->d_name) == 0) {
            ap_pclosedir(r->pool, dirp);
            return OK;
        }
        /*
         * miscapitalization errors are checked first (like, e.g., lower case
         * file, upper case request)
         */
        else if (strcasecmp(bad, dir_entry->d_name) == 0) {
            misspelled_file *sp_new;

	    sp_new = (misspelled_file *) ap_push_array(candidates);
            sp_new->name = ap_pstrdup(r->pool, dir_entry->d_name);
            sp_new->quality = SP_MISCAPITALIZED;
        }
        /*
         * simple typing errors are checked next (like, e.g.,
         * missing/extra/transposed char)
         */
        else if ((q = spdist(bad, dir_entry->d_name)) != SP_VERYDIFFERENT) {
            misspelled_file *sp_new;

	    sp_new = (misspelled_file *) ap_push_array(candidates);
            sp_new->name = ap_pstrdup(r->pool, dir_entry->d_name);
            sp_new->quality = q;
        }
        /*
	 * The spdist() should have found the majority of the misspelled
	 * requests.  It is of questionable use to continue looking for
	 * files with the same base name, but potentially of totally wrong
	 * type (index.html <-> index.db).
	 * I would propose to not set the WANT_BASENAME_MATCH define.
         *      08-Aug-1997 <[e-mail]>
         *
         * However, Alexei replied giving some reasons to add it anyway:
         * > Oh, by the way, I remembered why having the
         * > extension-stripping-and-matching stuff is a good idea:
         * >
         * > If you're using MultiViews, and have a file named foobar.html,
	 * > which you refer to as "foobar", and someone tried to access
	 * > "Foobar", mod_speling won't find it, because it won't find
	 * > anything matching that spelling. With the extension-munging,
	 * > it would locate "foobar.html". Not perfect, but I ran into
	 * > that problem when I first wrote the module.
	 */
        else {
#ifdef WANT_BASENAME_MATCH
            /*
             * Okay... we didn't find anything. Now we take out the hard-core
             * power tools. There are several cases here. Someone might have
             * entered a wrong extension (.htm instead of .html or vice
             * versa) or the document could be negotiated. At any rate, now
             * we just compare stuff before the first dot. If it matches, we
             * figure we got us a match. This can result in wrong things if
             * there are files of different content types but the same prefix
             * (e.g. foo.gif and foo.html) This code will pick the first one
             * it finds. Better than a Not Found, though.
             */
            int entloc = ap_ind(dir_entry->d_name, '.');
            if (entloc == -1) {
                entloc = strlen(dir_entry->d_name);
	    }

            if ((dotloc == entloc)
                && !strncasecmp(bad, dir_entry->d_name, dotloc)) {
                misspelled_file *sp_new;

		sp_new = (misspelled_file *) ap_push_array(candidates);
                sp_new->name = ap_pstrdup(r->pool, dir_entry->d_name);
                sp_new->quality = SP_VERYDIFFERENT;
            }
#endif
        }
    }
    ap_pclosedir(r->pool, dirp);

    if (candidates->nelts != 0) {
        /* Wow... we found us a mispelling. Construct a fixed url */
        char *nuri;
	const char *ref;
        misspelled_file *variant = (misspelled_file *) candidates->elts;
        int i;

        ref = ap_table_get(r->headers_in, "Referer");

        qsort((void *) candidates->elts, candidates->nelts,
              sizeof(misspelled_file), sort_by_quality);

        /*
         * Conditions for immediate redirection: 
         *     a) the first candidate was not found by stripping the suffix 
         * AND b) there exists only one candidate OR the best match is not
	 *        ambiguous
         * then return a redirection right away.
         */
        if (variant[0].quality != SP_VERYDIFFERENT
	    && (candidates->nelts == 1
		|| variant[0].quality != variant[1].quality)) {

            nuri = ap_pstrcat(r->pool, url, variant[0].name, r->path_info,
			      r->parsed_uri.query ? "?" : "",
			      r->parsed_uri.query ? r->parsed_uri.query : "",
			      NULL);

            ap_table_setn(r->headers_out, "Location",
			  ap_construct_url(r->pool, nuri, r));

            ap_log_rerror(APLOG_MARK, APLOG_NOERRNO | APLOG_INFO, r,
			 ref ? "Fixed spelling: %s to %s from %s"
			     : "Fixed spelling: %s to %s",
			 r->uri, nuri, ref);

            return HTTP_MOVED_PERMANENTLY;
        }
        /*
         * Otherwise, a "[300] Multiple Choices" list with the variants is
         * returned.
         */
        else {
            char *t;
            pool *p;
            table *notes;

            if (r->main == NULL) {
                p = r->pool;
                notes = r->notes;
            }
            else {
                p = r->main->pool;
                notes = r->main->notes;
            }

            /* Generate the response text. */
            /*
	     * Since the text is expanded by repeated calls of
             * t = pstrcat(p, t, ".."), we can avoid a little waste
             * of memory by adding the header AFTER building the list.
             * XXX: FIXME: find a way to build a string concatenation
             *             without repeatedly requesting new memory
             * XXX: FIXME: Limit the list to a maximum number of entries
             */
            t = "";

            for (i = 0; i < candidates->nelts; ++i) {
		char *vuri;
		const char *reason;

		reason = sp_reason_str[(int) (variant[i].quality)];
                /* The format isn't very neat... */
		vuri = ap_pstrcat(p, url, variant[i].name, r->path_info,
				  (r->parsed_uri.query != NULL) ? "?" : "",
				  (r->parsed_uri.query != NULL)
				      ? r->parsed_uri.query : "",
				  NULL);
		ap_table_mergen(r->subprocess_env, "VARIANTS",
				ap_pstrcat(p, "\"", vuri, "\";\"",
					   reason, "\"", NULL));
                t = ap_pstrcat(p, t, "<li><a href=\"", vuri,
			       "\">", vuri, "</a> (", reason, ")\n", NULL);

                /*
                 * when we have printed the "close matches" and there are
                 * more "distant matches" (matched by stripping the suffix),
                 * then we insert an additional separator text to suggest
                 * that the user LOOK CLOSELY whether these are really the
                 * files she wanted.
                 */
                if (i > 0 && i < candidates->nelts - 1
                    && variant[i].quality != SP_VERYDIFFERENT
                    && variant[i + 1].quality == SP_VERYDIFFERENT) {
                    t = ap_pstrcat(p, t, 
				   "</ul>\nFurthermore, the following related "
				   "documents were found:\n<ul>\n", NULL);
                }
            }
            t = ap_pstrcat(p, "The document name you requested (<code>",
			   r->uri,
			   "</code>) could not be found on this server.\n"
			   "However, we found documents with names similar "
			   "to the one you requested.<p>"
			   "Available documents:\n<ul>\n", t, "</ul>\n", NULL);

            /* If we know there was a referring page, add a note: */
            if (ref != NULL) {
                t = ap_pstrcat(p, t,
			       "Please consider informing the owner of the "
			       "<a href=\"", ref, 
			       "\">referring page</a> "
			       "about the broken link.\n",
			       NULL);
	    }

            /* Pass our table to http_protocol.c (see mod_negotiation): */
            ap_table_setn(notes, "variant-list", t);

            ap_log_rerror(APLOG_MARK, APLOG_NOERRNO | APLOG_INFO, r,
			 ref ? "Spelling fix: %s: %d candidates from %s"
			     : "Spelling fix: %s: %d candidates",
			 r->uri, candidates->nelts, ref);

            return HTTP_MULTIPLE_CHOICES;
        }
    }

    return OK;
}

module MODULE_VAR_EXPORT speling_module =
{
    STANDARD_MODULE_STUFF,
    NULL,                       /* initializer */
    create_mconfig_for_directory,  /* create per-dir config */
    NULL,                       /* merge per-dir config */
    create_mconfig_for_server,  /* server config */
    NULL,                       /* merge server config */
    speling_cmds,               /* command table */
    NULL,                       /* handlers */
    NULL,                       /* filename translation */
    NULL,                       /* check_user_id */
    NULL,                       /* check auth */
    NULL,                       /* check access */
    NULL,                       /* type_checker */
    check_speling,              /* fixups */
    NULL,                       /* logger */
    NULL,                       /* header parser */
    NULL,                       /* child_init */
    NULL,                       /* child_exit */
    NULL                        /* post read-request */
};