mod_negotiation.c

apache 安装教程 apache 安装教程

/* Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

/*
 * mod_negotiation.c: keeps track of MIME types the client is willing to
 * accept, and contains code to handle type arbitration.
 *
 * rst
 */

#include "httpd.h"
#include "http_config.h"
#include "http_request.h"
#include "http_protocol.h"
#include "http_core.h"
#include "http_log.h"
#include "util_script.h"

/* Commands --- configuring document caching on a per (virtual?)
 * server basis... 
 */

typedef struct {
    array_header *language_priority;
} neg_dir_config;

module MODULE_VAR_EXPORT negotiation_module;

static void *create_neg_dir_config(pool *p, char *dummy)
{
    neg_dir_config *new = (neg_dir_config *) ap_palloc(p, sizeof(neg_dir_config));

    new->language_priority = ap_make_array(p, 4, sizeof(char *));
    return new;
}

static void *merge_neg_dir_configs(pool *p, void *basev, void *addv)
{
    neg_dir_config *base = (neg_dir_config *) basev;
    neg_dir_config *add = (neg_dir_config *) addv;
    neg_dir_config *new = (neg_dir_config *) ap_palloc(p, sizeof(neg_dir_config));

    /* give priority to the config in the subdirectory */
    new->language_priority = ap_append_arrays(p, add->language_priority,
                                           base->language_priority);
    return new;
}

static const char *set_language_priority(cmd_parms *cmd, void *n, char *lang)
{
    array_header *arr = ((neg_dir_config *) n)->language_priority;
    char **langp = (char **) ap_push_array(arr);

    *langp = lang;
    return NULL;
}

static const char *cache_negotiated_docs(cmd_parms *cmd, void *dummy,
                                         char *dummy2)
{
    void *server_conf = cmd->server->module_config;

    ap_set_module_config(server_conf, &negotiation_module, "Cache");
    return NULL;
}

static int do_cache_negotiated_docs(server_rec *s)
{
    return (ap_get_module_config(s->module_config, &negotiation_module) != NULL);
}

static const command_rec negotiation_cmds[] =
{
    {"CacheNegotiatedDocs", cache_negotiated_docs, NULL, RSRC_CONF, NO_ARGS,
     "no arguments (either present or absent)"},
    {"LanguagePriority", set_language_priority, NULL, OR_FILEINFO, ITERATE,
     "space-delimited list of MIME language abbreviations"},
    {NULL}
};

/*
 * Record of available info on a media type specified by the client
 * (we also use 'em for encodings and languages)
 */

typedef struct accept_rec {
    char *name;                 /* MUST be lowercase */
    float quality;
    float level;
    char *charset;              /* for content-type only */
} accept_rec;

/*
 * Record of available info on a particular variant
 *
 * Note that a few of these fields are updated by the actual negotiation
 * code.  These are:
 *
 * level_matched --- initialized to zero.  Set to the value of level
 *             if the client actually accepts this media type at that
 *             level (and *not* if it got in on a wildcard).  See level_cmp
 *             below.
 * mime_stars -- initialized to zero.  Set to the number of stars
 *               present in the best matching Accept header element.
 *               1 for star/star, 2 for type/star and 3 for
 *               type/subtype.
 *
 * definite -- initialized to 1.  Set to 0 if there is a match which
 *             makes the variant non-definite according to the rules
 *             in rfc2296.
 */

typedef struct var_rec {
    request_rec *sub_req;       /* May be NULL (is, for map files) */
    char *mime_type;            /* MUST be lowercase */
    char *file_name;
    const char *content_encoding;
    array_header *content_languages;   /* list of languages for this variant */
    char *content_charset;
    char *description;

    /* The next five items give the quality values for the dimensions
     * of negotiation for this variant. They are obtained from the
     * appropriate header lines, except for source_quality, which
     * is obtained from the variant itself (the 'qs' parameter value
     * from the variant's mime-type). Apart from source_quality,
     * these values are set when we find the quality for each variant
     * (see best_match()). source_quality is set from the 'qs' parameter
     * of the variant description or mime type: see set_mime_fields().
     */
    float lang_quality;         /* quality of this variant's language */
    float encoding_quality;     /* ditto encoding */
    float charset_quality;      /* ditto charset */
    float mime_type_quality;    /* ditto media type */
    float source_quality;       /* source quality for this variant */

    /* Now some special values */
    float level;                /* Auxiliary to content-type... */
    float bytes;                /* content length, if known */
    int lang_index;             /* pre HTTP/1.1 language priority stuff */
    int is_pseudo_html;         /* text/html, *or* the INCLUDES_MAGIC_TYPEs */

    /* Above are all written-once properties of the variant.  The
     * three fields below are changed during negotiation:
     */

    float level_matched;
    int mime_stars;
    int definite;
} var_rec;

/* Something to carry around the state of negotiation (and to keep
 * all of this thread-safe)...
 */

typedef struct {
    pool *pool;
    request_rec *r;
    char *dir_name;
    int accept_q;               /* 1 if an Accept item has a q= param */
    float default_lang_quality; /* fiddle lang q for variants with no lang */

    /* the array pointers below are NULL if the corresponding accept
     * headers are not present
     */
    array_header *accepts;            /* accept_recs */
    array_header *accept_encodings;   /* accept_recs */
    array_header *accept_charsets;    /* accept_recs */
    array_header *accept_langs;       /* accept_recs */

    array_header *avail_vars;         /* available variants */

    int count_multiviews_variants;    /* number of variants found on disk */

    int is_transparent;       /* 1 if this resource is trans. negotiable */

    int dont_fiddle_headers;  /* 1 if we may not fiddle with accept hdrs */
    int ua_supports_trans;    /* 1 if ua supports trans negotiation */
    int send_alternates;      /* 1 if we want to send an Alternates header */
    int may_choose;           /* 1 if we may choose a variant for the client */
    int use_rvsa;             /* 1 if we must use RVSA/1.0 negotiation algo */
} negotiation_state;

/* A few functions to manipulate var_recs.
 * Cleaning out the fields...
 */

static void clean_var_rec(var_rec *mime_info)
{
    mime_info->sub_req = NULL;
    mime_info->mime_type = "";
    mime_info->file_name = "";
    mime_info->content_encoding = NULL;
    mime_info->content_languages = NULL;
    mime_info->content_charset = "";
    mime_info->description = "";

    mime_info->is_pseudo_html = 0;
    mime_info->level = 0.0f;
    mime_info->level_matched = 0.0f;
    mime_info->bytes = 0.0f;
    mime_info->lang_index = -1;
    mime_info->mime_stars = 0;
    mime_info->definite = 1;

    mime_info->charset_quality = 1.0f;
    mime_info->encoding_quality = 1.0f;
    mime_info->lang_quality = 1.0f;
    mime_info->mime_type_quality = 1.0f;
    mime_info->source_quality = 0.0f;
}

/* Initializing the relevant fields of a variant record from the
 * accept_info read out of its content-type, one way or another.
 */

static void set_mime_fields(var_rec *var, accept_rec *mime_info)
{
    var->mime_type = mime_info->name;
    var->source_quality = mime_info->quality;
    var->level = mime_info->level;
    var->content_charset = mime_info->charset;

    var->is_pseudo_html = (!strcmp(var->mime_type, "text/html")
                           || !strcmp(var->mime_type, INCLUDES_MAGIC_TYPE)
                           || !strcmp(var->mime_type, INCLUDES_MAGIC_TYPE3));
}

/* Create a variant list validator in r using info from vlistr. */

static void set_vlist_validator(request_rec *r, request_rec *vlistr)
{
    /* Calculating the variant list validator is similar to
     * calculating an etag for the source of the variant list
     * information, so we use ap_make_etag().  Note that this
     * validator can be 'weak' in extreme case.
     */

    ap_update_mtime (vlistr, vlistr->finfo.st_mtime);
    r->vlist_validator = ap_make_etag(vlistr, 0);

    /* ap_set_etag will later take r->vlist_validator into account
     * when creating the etag header
     */
}


/*****************************************************************
 *
 * Parsing (lists of) media types and their parameters, as seen in
 * HTTPD header lines and elsewhere.
 */

/*
 * Get a single mime type entry --- one media type and parameters;
 * enter the values we recognize into the argument accept_rec
 */

static const char *get_entry(pool *p, accept_rec *result,
                             const char *accept_line)
{
    result->quality = 1.0f;
    result->level = 0.0f;
    result->charset = "";

    /*
     * Note that this handles what I gather is the "old format",
     *
     *    Accept: text/html text/plain moo/zot
     *
     * without any compatibility kludges --- if the token after the
     * MIME type begins with a semicolon, we know we're looking at parms,
     * otherwise, we know we aren't.  (So why all the pissing and moaning
     * in the CERN server code?  I must be missing something).
     */

    result->name = ap_get_token(p, &accept_line, 0);
    ap_str_tolower(result->name);     /* You want case-insensitive,
                                       * you'll *get* case-insensitive.
                                       */

    /* KLUDGE!!! Default HTML to level 2.0 unless the browser
     * *explicitly* says something else.
     */

    if (!strcmp(result->name, "text/html") && (result->level == 0.0)) {
        result->level = 2.0f;
    }
    else if (!strcmp(result->name, INCLUDES_MAGIC_TYPE)) {
        result->level = 2.0f;
    }
    else if (!strcmp(result->name, INCLUDES_MAGIC_TYPE3)) {
        result->level = 3.0f;
    }

    while (*accept_line == ';') {
        /* Parameters ... */

        char *parm;
        char *cp;
        char *end;

        ++accept_line;
        parm = ap_get_token(p, &accept_line, 1);

        /* Look for 'var = value' --- and make sure the var is in lcase. */

        for (cp = parm; (*cp && !ap_isspace(*cp) && *cp != '='); ++cp) {
            *cp = ap_tolower(*cp);
        }

        if (!*cp) {
            continue;           /* No '='; just ignore it. */
        }

        *cp++ = '\0';           /* Delimit var */
        while (*cp && (ap_isspace(*cp) || *cp == '=')) {
            ++cp;
        }

        if (*cp == '"') {
            ++cp;
            for (end = cp;
                 (*end && *end != '\n' && *end != '\r' && *end != '\"');
                 end++);
        }
        else {
            for (end = cp; (*end && !ap_isspace(*end)); end++);
        }
        if (*end) {
            *end = '\0';        /* strip ending quote or return */
        }
        ap_str_tolower(cp);

        if (parm[0] == 'q'
            && (parm[1] == '\0' || (parm[1] == 's' && parm[2] == '\0'))) {
            result->quality = (float)atof(cp);
        }
        else if (parm[0] == 'l' && !strcmp(&parm[1], "evel")) {
            result->level = (float)atof(cp);
        }
        else if (!strcmp(parm, "charset")) {
            result->charset = cp;
        }
    }

    if (*accept_line == ',') {
        ++accept_line;
    }

    return accept_line;
}

/*****************************************************************
 *
 * Dealing with header lines ...
 *
 * Accept, Accept-Charset, Accept-Language and Accept-Encoding
 * are handled by do_header_line() - they all have the same
 * basic structure of a list of items of the format
 *    name; q=N; charset=TEXT
 *
 * where charset is only valid in Accept.
 */

static array_header *do_header_line(pool *p, const char *accept_line)
{
    array_header *accept_recs;

    if (!accept_line) {
        return NULL;
    }

    accept_recs = ap_make_array(p, 40, sizeof(accept_rec));

    while (*accept_line) {
        accept_rec *new = (accept_rec *) ap_push_array(accept_recs);
        accept_line = get_entry(p, new, accept_line);
    }

    return accept_recs;
}