mod_rewrite.c

Apache官方在今天放出产品系列2.2的最新版本2.2.11的源码包 最流行的HTTP服务器软件之一

            return 0;
        }
        conf->rewritelogfp = ap_piped_log_write_fd(pl);
    }
    else {
        apr_status_t rc;

        fname = ap_server_root_relative(p, conf->rewritelogfile);
        if (!fname) {
            ap_log_error(APLOG_MARK, APLOG_ERR, APR_EBADPATH, s,
                         "mod_rewrite: Invalid RewriteLog "
                         "path %s", conf->rewritelogfile);
            return 0;
        }

        if ((rc = apr_file_open(&conf->rewritelogfp, fname,
                                REWRITELOG_FLAGS, REWRITELOG_MODE, p))
                != APR_SUCCESS) {
            ap_log_error(APLOG_MARK, APLOG_ERR, rc, s,
                         "mod_rewrite: could not open RewriteLog "
                         "file %s", fname);
            return 0;
        }
    }

    return 1;
}

static void do_rewritelog(request_rec *r, int level, char *perdir,
                          const char *fmt, ...)
{
    rewrite_server_conf *conf;
    char *logline, *text;
    const char *rhost, *rname;
    apr_size_t nbytes;
    int redir;
    apr_status_t rv;
    request_rec *req;
    va_list ap;

    conf = ap_get_module_config(r->server->module_config, &rewrite_module);

    if (!conf->rewritelogfp || level > conf->rewriteloglevel) {
        return;
    }

    rhost = ap_get_remote_host(r->connection, r->per_dir_config,
                               REMOTE_NOLOOKUP, NULL);
    rname = ap_get_remote_logname(r);

    for (redir=0, req=r; req->prev; req = req->prev) {
        ++redir;
    }

    va_start(ap, fmt);
    text = apr_pvsprintf(r->pool, fmt, ap);
    va_end(ap);

    logline = apr_psprintf(r->pool, "%s %s %s %s [%s/sid#%pp][rid#%pp/%s%s%s] "
                                    "(%d) %s%s%s%s" APR_EOL_STR,
                           rhost ? rhost : "UNKNOWN-HOST",
                           rname ? rname : "-",
                           r->user ? (*r->user ? r->user : "\"\"") : "-",
                           current_logtime(r),
                           ap_get_server_name(r),
                           (void *)(r->server),
                           (void *)r,
                           r->main ? "subreq" : "initial",
                           redir ? "/redir#" : "",
                           redir ? apr_itoa(r->pool, redir) : "",
                           level,
                           perdir ? "[perdir " : "",
                           perdir ? perdir : "",
                           perdir ? "] ": "",
                           text);

    rv = apr_global_mutex_lock(rewrite_log_lock);
    if (rv != APR_SUCCESS) {
        ap_log_rerror(APLOG_MARK, APLOG_ERR, rv, r,
                      "apr_global_mutex_lock(rewrite_log_lock) failed");
        /* XXX: Maybe this should be fatal? */
    }

    nbytes = strlen(logline);
    apr_file_write(conf->rewritelogfp, logline, &nbytes);

    rv = apr_global_mutex_unlock(rewrite_log_lock);
    if (rv != APR_SUCCESS) {
        ap_log_rerror(APLOG_MARK, APLOG_ERR, rv, r,
                      "apr_global_mutex_unlock(rewrite_log_lock) failed");
        /* XXX: Maybe this should be fatal? */
    }

    return;
}
#endif /* !REWRITELOG_DISABLED */


/*
 * +-------------------------------------------------------+
 * |                                                       |
 * |                URI and path functions
 * |                                                       |
 * +-------------------------------------------------------+
 */

/* return number of chars of the scheme (incl. '://')
 * if the URI is absolute (includes a scheme etc.)
 * otherwise 0.
 *
 * NOTE: If you add new schemes here, please have a
 *       look at escape_absolute_uri and splitout_queryargs.
 *       Not every scheme takes query strings and some schemes
 *       may be handled in a special way.
 *
 * XXX: we may consider a scheme registry, perhaps with
 *      appropriate escape callbacks to allow other modules
 *      to extend mod_rewrite at runtime.
 */
static unsigned is_absolute_uri(char *uri)
{
    /* fast exit */
    if (*uri == '/' || strlen(uri) <= 5) {
        return 0;
    }

    switch (*uri++) {
    case 'a':
    case 'A':
        if (!strncasecmp(uri, "jp://", 5)) {        /* ajp://    */
          return 6;
        }

    case 'b':
    case 'B':
        if (!strncasecmp(uri, "alancer://", 10)) {   /* balancer:// */
          return 11;
        }
        break;

    case 'f':
    case 'F':
        if (!strncasecmp(uri, "tp://", 5)) {        /* ftp://    */
            return 6;
        }
        break;

    case 'g':
    case 'G':
        if (!strncasecmp(uri, "opher://", 8)) {     /* gopher:// */
            return 9;
        }
        break;

    case 'h':
    case 'H':
        if (!strncasecmp(uri, "ttp://", 6)) {       /* http://   */
            return 7;
        }
        else if (!strncasecmp(uri, "ttps://", 7)) { /* https://  */
            return 8;
        }
        break;

    case 'l':
    case 'L':
        if (!strncasecmp(uri, "dap://", 6)) {       /* ldap://   */
            return 7;
        }
        break;

    case 'm':
    case 'M':
        if (!strncasecmp(uri, "ailto:", 6)) {       /* mailto:   */
            return 7;
        }
        break;

    case 'n':
    case 'N':
        if (!strncasecmp(uri, "ews:", 4)) {         /* news:     */
            return 5;
        }
        else if (!strncasecmp(uri, "ntp://", 6)) {  /* nntp://   */
            return 7;
        }
        break;
    }

    return 0;
}

/*
 * escape absolute uri, which may or may not be path oriented.
 * So let's handle them differently.
 */
static char *escape_absolute_uri(apr_pool_t *p, char *uri, unsigned scheme)
{
    char *cp;

    /* be safe.
     * NULL should indicate elsewhere, that something's wrong
     */
    if (!scheme || strlen(uri) < scheme) {
        return NULL;
    }

    cp = uri + scheme;

    /* scheme with authority part? */
    if (cp[-1] == '/') {
        /* skip host part */
        while (*cp && *cp != '/') {
            ++cp;
        }

        /* nothing after the hostpart. ready! */
        if (!*cp || !*++cp) {
            return apr_pstrdup(p, uri);
        }

        /* remember the hostname stuff */
        scheme = cp - uri;

        /* special thing for ldap.
         * The parts are separated by question marks. From RFC 2255:
         *     ldapurl = scheme "://" [hostport] ["/"
         *               [dn ["?" [attributes] ["?" [scope]
         *               ["?" [filter] ["?" extensions]]]]]]
         */
        if (!strncasecmp(uri, "ldap", 4)) {
            char *token[5];
            int c = 0;

            token[0] = cp = apr_pstrdup(p, cp);
            while (*cp && c < 4) {
                if (*cp == '?') {
                    token[++c] = cp + 1;
                    *cp = '\0';
                }
                ++cp;
            }

            return apr_pstrcat(p, apr_pstrndup(p, uri, scheme),
                                          ap_escape_uri(p, token[0]),
                               (c >= 1) ? "?" : NULL,
                               (c >= 1) ? ap_escape_uri(p, token[1]) : NULL,
                               (c >= 2) ? "?" : NULL,
                               (c >= 2) ? ap_escape_uri(p, token[2]) : NULL,
                               (c >= 3) ? "?" : NULL,
                               (c >= 3) ? ap_escape_uri(p, token[3]) : NULL,
                               (c >= 4) ? "?" : NULL,
                               (c >= 4) ? ap_escape_uri(p, token[4]) : NULL,
                               NULL);
        }
    }

    /* Nothing special here. Apply normal escaping. */
    return apr_pstrcat(p, apr_pstrndup(p, uri, scheme),
                       ap_escape_uri(p, cp), NULL);
}

/*
 * split out a QUERY_STRING part from
 * the current URI string
 */
static void splitout_queryargs(request_rec *r, int qsappend)
{
    char *q;

    /* don't touch, unless it's an http or mailto URL.
     * See RFC 1738 and RFC 2368.
     */
    if (is_absolute_uri(r->filename)
        && strncasecmp(r->filename, "ajp", 3)
        && strncasecmp(r->filename, "balancer", 8)
        && strncasecmp(r->filename, "http", 4)
        && strncasecmp(r->filename, "mailto", 6)) {
        r->args = NULL; /* forget the query that's still flying around */
        return;
    }

    q = ap_strchr(r->filename, '?');
    if (q != NULL) {
        char *olduri;
        apr_size_t len;

        olduri = apr_pstrdup(r->pool, r->filename);
        *q++ = '\0';
        if (qsappend) {
            r->args = apr_pstrcat(r->pool, q, "&", r->args, NULL);
        }
        else {
            r->args = apr_pstrdup(r->pool, q);
        }

        len = strlen(r->args);
        if (!len) {
            r->args = NULL;
        }
        else if (r->args[len-1] == '&') {
            r->args[len-1] = '\0';
        }

        rewritelog((r, 3, NULL, "split uri=%s -> uri=%s, args=%s", olduri,
                    r->filename, r->args ? r->args : "<none>"));
    }

    return;
}

/*
 * strip 'http[s]://ourhost/' from URI
 */
static void reduce_uri(request_rec *r)
{
    char *cp;
    apr_size_t l;

    cp = (char *)ap_http_scheme(r);
    l  = strlen(cp);
    if (   strlen(r->filename) > l+3
        && strncasecmp(r->filename, cp, l) == 0
        && r->filename[l]   == ':'
        && r->filename[l+1] == '/'
        && r->filename[l+2] == '/' ) {

        unsigned short port;
        char *portp, *host, *url, *scratch;

        scratch = apr_pstrdup(r->pool, r->filename); /* our scratchpad */

        /* cut the hostname and port out of the URI */
        cp = host = scratch + l + 3;    /* 3 == strlen("://") */
        while (*cp && *cp != '/' && *cp != ':') {
            ++cp;
        }

        if (*cp == ':') {      /* additional port given */
            *cp++ = '\0';
            portp = cp;
            while (*cp && *cp != '/') {
                ++cp;
            }
            *cp = '\0';

            port = atoi(portp);
            url = r->filename + (cp - scratch);
            if (!*url) {
                url = "/";
            }
        }
        else if (*cp == '/') { /* default port */
            *cp = '\0';

            port = ap_default_port(r);
            url = r->filename + (cp - scratch);
        }
        else {
            port = ap_default_port(r);
            url = "/";
        }

        /* now check whether we could reduce it to a local path... */
        if (ap_matches_request_vhost(r, host, port)) {
            rewritelog((r, 3, NULL, "reduce %s -> %s", r->filename, url));
            r->filename = apr_pstrdup(r->pool, url);
        }
    }

    return;
}

/*
 * add 'http[s]://ourhost[:ourport]/' to URI
 * if URI is still not fully qualified
 */
static void fully_qualify_uri(request_rec *r)
{
    if (!is_absolute_uri(r->filename)) {
        const char *thisserver;
        char *thisport;
        int port;

        thisserver = ap_get_server_name(r);
        port = ap_get_server_port(r);
        thisport = ap_is_default_port(port, r)
                   ? ""
                   : apr_psprintf(r->pool, ":%u", port);

        r->filename = apr_psprintf(r->pool, "%s://%s%s%s%s",
                                   ap_http_scheme(r), thisserver, thisport,
                                   (*r->filename == '/') ? "" : "/",
                                   r->filename);
    }

    return;
}

/*
 * stat() only the first segment of a path
 */
static int prefix_stat(const char *path, apr_pool_t *pool)
{
    const char *curpath = path;
    const char *root;
    const char *slash;
    char *statpath;
    apr_status_t rv;

    rv = apr_filepath_root(&root, &curpath, APR_FILEPATH_TRUENAME, pool);

    if (rv != APR_SUCCESS) {
        return 0;
    }

    /* let's recognize slashes only, the mod_rewrite semantics are opaque
     * enough.
     */
    if ((slash = ap_strchr_c(curpath, '/')) != NULL) {
        rv = apr_filepath_merge(&statpath, root,
                                apr_pstrndup(pool, curpath,
                                             (apr_size_t)(slash - curpath)),
                                APR_FILEPATH_NOTABOVEROOT |
                                APR_FILEPATH_NOTRELATIVE, pool);
    }
    else {
        rv = apr_filepath_merge(&statpath, root, curpath,
                                APR_FILEPATH_NOTABOVEROOT |
                                APR_FILEPATH_NOTRELATIVE, pool);
    }

    if (rv == APR_SUCCESS) {
        apr_finfo_t sb;

        if (apr_stat(&sb, statpath, APR_FINFO_MIN, pool) == APR_SUCCESS) {
            return 1;
        }