Message-ID: <384E91DF.D92FABC2@Golux.Com>
Date: Wed, 08 Dec 1999 12:14:07 -0500
From: Rodent of Unusual Size <Ken.Coar@Golux.Com>
Organization: The Apache Software Foundation
MIME-Version: 1.0
To: Apache Developers <new-httpd@Apache.Org>
Subject: [PATCH] enhance mod_mime with AddCharset
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: new-httpd-owner@apache.org
Precedence: bulk
Reply-To: new-httpd@apache.org
Status: O

This is something that's been requested forever, and was
submitted by Youichirou Koga a while ago.  Martin Duerst
has re-opened the request to add it for 1.3.

I'm +1 after preliminary testing.  There's docco to go
with it, but it's not in this patch.
-- 
#ken    P-)}

Ken Coar                    <http://Web.Golux.Com/coar/>
Apache Software Foundation  <http://www.apache.org/>
"Apache Server for Dummies" <http://ASFD.MeepZor.Com/>

Index: src/modules/standard/mod_mime.c
===================================================================
RCS file: /home/cvs/apache-1.3/src/modules/standard/mod_mime.c,v
retrieving revision 1.49
diff -u -r1.49 mod_mime.c
--- mod_mime.c  1999/04/20 17:27:51     1.49
+++ mod_mime.c  1999/12/08 17:10:24
@@ -75,6 +75,7 @@
 typedef struct {
     table *forced_types;        /* Additional AddTyped stuff */
     table *encoding_types;      /* Added with AddEncoding... */
+    table *charset_types;      /* Added with AddCharset... */
     table *language_types;      /* Added with AddLanguage... */
     table *handlers;            /* Added with AddHandler...  */
     array_header *handlers_remove;     /* List of handlers to remove */
@@ -84,6 +85,24 @@
     char *default_language;     /* Language if no AddLanguage ext found */
 } mime_dir_config;
 
+typedef struct param_s {
+    char *attr;
+    char *val;
+    struct param_s *next;
+} param;
+
+typedef struct {
+    char *type;
+    char *subtype;
+    param *param;
+} content_type;
+
+static char tspecial[] = {
+    '(', ')', '<', '>', '@', ',', ';', ':',
+    '\\', '"', '/', '[', ']', '?', '=',
+    '\0'
+};
+
 module MODULE_VAR_EXPORT mime_module;
 
 static void *create_mime_dir_config(pool *p, char *dummy)
@@ -93,6 +112,7 @@
 
     new->forced_types = ap_make_table(p, 4);
     new->encoding_types = ap_make_table(p, 4);
+    new->charset_types = ap_make_table(p, 4);
     new->language_types = ap_make_table(p, 4);
     new->handlers = ap_make_table(p, 4);
     new->handlers_remove = ap_make_array(p, 4, sizeof(handlers_info));
@@ -119,9 +139,11 @@
     }
 
     new->forced_types = ap_overlay_tables(p, add->forced_types,
-                                       base->forced_types);
+                                        base->forced_types);
     new->encoding_types = ap_overlay_tables(p, add->encoding_types,
                                          base->encoding_types);
+    new->charset_types = ap_overlay_tables(p, add->charset_types,
+                                          base->charset_types);
     new->language_types = ap_overlay_tables(p, add->language_types,
                                          base->language_types);
     new->handlers = ap_overlay_tables(p, add->handlers,
@@ -155,11 +177,23 @@
     return NULL;
 }
 
-static const char *add_language(cmd_parms *cmd, mime_dir_config * m, char *lang,
+static const char *add_charset(cmd_parms *cmd, mime_dir_config *m,
+                              char *charset, char *ext)
+{
+    if (*ext == '.') {
+       ++ext;
+    }
+    ap_str_tolower(charset);
+    ap_table_setn(m->charset_types, ext, charset);
+    return NULL;
+}
+
+static const char *add_language(cmd_parms *cmd, mime_dir_config *m, char *lang,
                                 char *ext)
 {
-    if (*ext == '.')
-        ++ext;
+    if (*ext == '.') {
+       ++ext;
+    }
     ap_str_tolower(lang);
     ap_table_setn(m->language_types, ext, lang);
     return NULL;
@@ -209,6 +243,8 @@
      "a mime type followed by one or more file extensions"},
     {"AddEncoding", add_encoding, NULL, OR_FILEINFO, ITERATE2,
      "an encoding (e.g., gzip), followed by one or more file extensions"},
+    {"AddCharset", add_charset, NULL, OR_FILEINFO, ITERATE2,
+     "a charset (e.g., iso-2022-jp), followed by one or more file extensions"},
     {"AddLanguage", add_language, NULL, OR_FILEINFO, ITERATE2,
      "a language (e.g., fr), followed by one or more file extensions"},
     {"AddHandler", add_handler, NULL, OR_FILEINFO, ITERATE2,
@@ -275,6 +311,257 @@
     ap_cfg_closefile(f);
 }
 
+char *zap_sp(char *s)
+{
+    char *tp;
+
+    if (s == NULL) {
+       return (NULL);
+    }
+    if (*s == '\0') {
+       return (s);
+    }
+
+    /* delete prefixed white space */
+    for (; *s == ' ' || *s == '\t' || *s == '\n'; s++);
+
+    /* delete postfixed white space */
+    for (tp = s; *tp != '\0'; tp++);
+    for (tp--; tp != s && (*tp == ' ' || *tp == '\t' || *tp == '\n'); tp--) {
+       *tp = '\0';
+    }
+    return (s);
+}
+
+int is_token(int c)
+{
+    int res = -1;
+    int i;
+
+    if (c >= 33 && c <= 127) {
+       res = 1;
+       for (i = 0; tspecial[i] != '\0'; i++) {
+           if (c == tspecial[i]) {
+               res = -1;
+               break;
+           }
+       }
+    }
+    return (res);
+}
+
+int is_qtext(int c)
+{
+    int res = -1;
+
+    if ((c >= 0 && c <= 127) && !(c == '"' || c == '\\' || c == '\n')) {
+       res = 1;
+    }
+    return (res);
+}
+
+int is_quoted_pair(char *s)
+{
+    int res = -1;
+    int c;
+
+    if (((s + 1) != NULL) && (*s == '\\')) {
+       c = (int) *(s + 1);
+       if (c >= 0 && c <= 127) {
+           res = 1;
+       }
+    }
+    return (res);
+}
+
+content_type *analyze_ct(pool * p, char *s)
+{
+    char *tp, *mp, *cp;
+    char *attribute, *value;
+    int quoted = 0;
+
+    content_type *ctp;
+    param *pp, *npp;
+
+    /* initialize ctp */
+    ctp = (content_type *) ap_palloc(p, sizeof(content_type));
+    ctp->type = NULL;
+    ctp->subtype = NULL;
+    ctp->param = NULL;
+
+    tp = ap_pstrdup(p, s);
+
+    mp = tp;
+    cp = mp;
+
+    /* getting a type */
+    if (!(cp = strchr(mp, '/'))) {
+       ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                    "Cannot get media type.");
+       return (NULL);
+    }
+    ctp->type = ap_pstrndup(p, mp, cp - mp);
+    ctp->type = zap_sp(ctp->type);
+    if (ctp->type == NULL || *(ctp->type) == '\0' ||
+       strchr(ctp->type, ';') || strchr(ctp->type, ' ') ||
+       strchr(ctp->type, '\t')) {
+       ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                    "Cannot get media subtype.");
+       return (NULL);
+    }
+
+    /* getting a subtype */
+    cp++;
+    mp = cp;
+
+    for (; *cp != ';' && *cp != '\0'; cp++);
+    ctp->subtype = ap_pstrndup(p, mp, cp - mp);
+    ctp->subtype = zap_sp(ctp->subtype);
+    if ((ctp->subtype == NULL) || (*(ctp->subtype) == '\0') ||
+       strchr(ctp->subtype, ' ') || strchr(ctp->subtype, '\t')) {
+       ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                    "Cannot get media subtype.");
+       return (NULL);
+    }
+    cp = zap_sp(cp);
+    if (cp == NULL || *cp == '\0') {
+       return (ctp);
+    }
+
+    /* getting parameters */
+    cp++;
+    cp = zap_sp(cp);
+    if (cp == NULL || *cp == '\0') {
+       ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                    "Cannot get media parameter.");
+       return (NULL);
+    }
+    mp = cp;
+    attribute = NULL;
+    value = NULL;
+
+    while (cp != NULL && *cp != '\0') {
+       if (attribute == NULL) {
+           if (is_token((int) *cp) > 0) {
+               cp++;
+               continue;
+           }
+           else if (*cp == ' ' || *cp == '\t' || *cp == '\n') {
+               cp++;
+               continue;
+           }
+           else if (*cp == '=') {
+               attribute = ap_pstrndup(p, mp, cp - mp);
+               attribute = zap_sp(attribute);
+               if (attribute == NULL || *attribute == '\0') {
+                   ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                                "Cannot get media parameter.");
+                   return (NULL);
+               }
+               cp++;
+               cp = zap_sp(cp);
+               if (cp == NULL || *cp == '\0') {
+                   ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                                "Cannot get media parameter.");
+                   return (NULL);
+               }
+               mp = cp;
+               continue;
+           }
+           else {
+               ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                            "Cannot get media parameter.");
+               return (NULL);
+           }
+       }
+       else {
+           if (mp == cp) {
+               if (*cp == '"') {
+                   quoted = 1;
+                   cp++;
+               }
+               else {
+                   quoted = 0;
+               }
+           }
+           if (quoted > 0) {
+               while (quoted && *cp != '\0') {
+                   if (is_qtext((int) *cp) > 0) {
+                       cp++;
+                   }
+                   else if (is_quoted_pair(cp) > 0) {
+                       cp += 2;
+                   }
+                   else if (*cp == '"') {
+                       cp++;
+                       while (*cp == ' ' || *cp == '\t' || *cp == '\n') {
+                           cp++;
+                       }
+                       if (*cp != ';' && *cp != '\0') {
+                           ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                                        "Cannot get media parameter.");
+                           return(NULL);
+                       }
+                       quoted = 0;
+                   }
+                   else {
+                       ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                                    "Cannot get media parameter.");
+                       return (NULL);
+                   }
+               }
+           }
+           else {
+               while (1) {
+                   if (is_token((int) *cp) > 0) {
+                       cp++;
+                   }
+                   else if (*cp == '\0' || *cp == ';') {
+                       break;
+                   }
+                   else {
+                       ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                                    "Cannot get media parameter.");
+                       return (NULL);
+                   }
+               }
+           }
+           value = ap_pstrndup(p, mp, cp - mp);
+           value = zap_sp(value);
+           if (value == NULL || *value == '\0') {
+               ap_log_error(APLOG_MARK, APLOG_WARNING, NULL,
+                            "Cannot get media parameter.");
+               return (NULL);
+           }
+
+           pp = ap_palloc(p, sizeof(param));
+           pp->attr = attribute;
+           pp->val = value;
+           pp->next = NULL;
+
+           if (ctp->param == NULL) {
+               ctp->param = pp;
+           }
+           else {
+               npp = ctp->param;
+               while (npp->next) {
+                   npp = npp->next;
+               }
+               npp->next = pp;
+           }
+           quoted = 0;
+           attribute = NULL;
+           value = NULL;
+           if (*cp == '\0') {
+               break;
+           }
+           cp++;
+           mp = cp;
+       }
+    }
+    return (ctp);
+}
+
 static int find_ct(request_rec *r)
 {
     const char *fn = strrchr(r->filename, '/');
@@ -283,6 +570,7 @@
     char *ext;
     const char *orighandler = r->handler;
     const char *type;
+    const char *charset;
 
     if (S_ISDIR(r->finfo.st_mode)) {
         r->content_type = DIR_MAGIC_TYPE;
@@ -308,6 +596,12 @@
             found = 1;
         }
 
+       /* Add charset to Content-Type */
+       if ((type = ap_table_get(conf->charset_types, ext))) {
+           charset = type;
+           found = 1;
+       }
+
         /* Check for Content-Language */
         if ((type = ap_table_get(conf->language_types, ext))) {
             const char **new;
@@ -347,8 +641,46 @@
             r->content_languages = NULL;
             r->content_encoding = NULL;
             r->handler = orighandler;
-        }
+           charset = NULL;
+       }
+    }
 
+    if (r->content_type) {
+       content_type *ctp;
+       char *ct;
+       int override = 0;
+
+       ct = (char *) ap_palloc(r->pool,
+                               sizeof(char) * (strlen(r->content_type) + 1));
+       strcpy(ct, r->content_type);
+
+       if ((ctp = analyze_ct(r->pool, ct))) {
+           param *pp = ctp->param;
+           r->content_type = ap_pstrcat(r->pool, ctp->type, "/",
+                                        ctp->subtype, NULL);
+           while (pp != NULL) {
+               if (charset && !strcmp(pp->attr, "charset")) {
+                   if (!override) {
+                       r->content_type = ap_pstrcat(r->pool, r->content_type,
+                                                    "; charset=", charset,
+                                                    NULL);
+                       override = 1;
+                   }
+               }
+               else {
+                   r->content_type = ap_pstrcat(r->pool, r->content_type,
+                                                "; ", pp->attr,
+                                                "=", pp->val,
+                                                NULL);
+               }
+               pp = pp->next;
+           }
+           if (charset && !override) {
+               r->content_type = ap_pstrcat(r->pool, r->content_type,
+                                            "; charset=", charset,
+                                            NULL);
+           }
+       }
     }
 
     /* Set default language, if none was specified by the extensions