removed a bunch of blank comment lines
[citadel.git] / webcit / html2html.c
index ee9f705d55515e2d144e3dfa71989541f33fc241..965dc55d69bc57ea9458068f8d19602897b261aa 100644 (file)
@@ -1,20 +1,24 @@
 /*
- * $Id$
- */
-/**
- * \defgroup HTML2HTML Output an HTML message, modifying it slightly to make sure it plays nice
+ * Output an HTML message, modifying it slightly to make sure it plays nice
  * with the rest of our web framework.
- * \ingroup WebcitHttpServer
+ *
+ * Copyright (c) 2005-2012 by the citadel.org team
+ *
+ * This program is open source software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License, version 3.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
  */
-/*@{*/
+
 #include "webcit.h"
 #include "webserver.h"
 
 
-/**
- * \brief      Strip surrounding single or double quotes from a string.
- *
- * \param s    String to be stripped.
+/*
+ * Strip surrounding single or double quotes from a string.
  */
 void stripquotes(char *s)
 {
@@ -32,12 +36,12 @@ void stripquotes(char *s)
 }
 
 
-/**
- * \brief Check to see if a META tag has overridden the declared MIME character set.
+/*
+ * Check to see if a META tag has overridden the declared MIME character set.
  *
- * \param charset              Character set name (left unchanged if we don't do anything)
- * \param meta_http_equiv      Content of the "http-equiv" portion of the META tag
- * \param meta_content         Content of the "content" portion of the META tag
+ * charset             Character set name (left unchanged if we don't do anything)
+ * meta_http_equiv     Content of the "http-equiv" portion of the META tag
+ * meta_content                Content of the "content" portion of the META tag
  */
 void extract_charset_from_meta(char *charset, char *meta_http_equiv, char *meta_content)
 {
@@ -70,17 +74,19 @@ void extract_charset_from_meta(char *charset, char *meta_http_equiv, char *meta_
                        strcpy(charset, "UTF-8");
                }
 
+               /* Remove wandering punctuation */
+               if ((ptr=strchr(charset, '\"'))) *ptr = 0;
+               striplt(charset);
        }
 }
 
 
 
-/**
- * \brief Sanitize and enhance an HTML message for display.
- *        Also convert weird character sets to UTF-8 if necessary.
- *        Also fixup img src="cid:..." type inline images to fetch the image
+/*
+ * Sanitize and enhance an HTML message for display.
+ * Also convert weird character sets to UTF-8 if necessary.
+ * Also fixup img src="cid:..." type inline images to fetch the image
  *
- * \param supplied_charset the input charset as declared in the MIME headers
  */
 void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, StrBuf *Source, StrBuf *Target) {
        char buf[SIZ];
@@ -103,11 +109,11 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
        StrBuf *BodyArea = NULL;
 #ifdef HAVE_ICONV
        iconv_t ic = (iconv_t)(-1) ;
-       char *ibuf;                   /**< Buffer of characters to be converted */
-       char *obuf;                   /**< Buffer for converted characters      */
-       size_t ibuflen;               /**< Length of input buffer               */
-       size_t obuflen;               /**< Length of output buffer              */
-       char *osav;                   /**< Saved pointer to output buffer       */
+       char *ibuf;                   /* Buffer of characters to be converted */
+       char *obuf;                   /* Buffer for converted characters      */
+       size_t ibuflen;               /* Length of input buffer               */
+       size_t obuflen;               /* Length of output buffer              */
+       char *osav;                   /* Saved pointer to output buffer       */
 #endif
        if (Target == NULL)
                Target = WC->WBuf;
@@ -123,9 +129,9 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                if (ptr == NULL) {
                        StrBufAppendPrintf(Target, "<b>");
                        StrBufAppendPrintf(Target, _("realloc() error! couldn't get %d bytes: %s"),
-                               buffer_length + 1,
-                               strerror(errno));
-                       StrBufAppendPrintf(Target, "</b><br /><br />\n");
+                                       buffer_length + 1,
+                                       strerror(errno));
+                       StrBufAppendPrintf(Target, "</b><br><br>\n");
                        while (serv_getln(buf, sizeof buf), strcmp(buf, "000")) {
                                /** flush */
                        }
@@ -158,7 +164,7 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                ++ptr;
                if ((ptr == NULL) || (ptr >= msgend)) break;
 
-               /**
+               /*
                 *  Look for META tags.  Some messages (particularly in
                 *  Asian locales) illegally declare a message's character
                 *  set in the HTML instead of in the MIME headers.  This
@@ -193,7 +199,7 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                                                        stripquotes(meta_http_equiv);
                                                        stripquotes(meta_content);
                                                        extract_charset_from_meta(charset,
-                                                               meta_http_equiv, meta_content);
+                                                                       meta_http_equiv, meta_content);
                                                }
                                                free(meta_content);
                                        }
@@ -203,16 +209,16 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                        }
                }
 
-               /**
+               /*
                 * Any of these tags cause everything up to and including
                 * the tag to be removed.
                 */     
                if ( (!strncasecmp(ptr, "HTML", 4))
-                  ||(!strncasecmp(ptr, "HEAD", 4))
-                  ||(!strncasecmp(ptr, "/HEAD", 5))
-                  ||(!strncasecmp(ptr, "BODY", 4)) ) {
+                               ||(!strncasecmp(ptr, "HEAD", 4))
+                               ||(!strncasecmp(ptr, "/HEAD", 5))
+                               ||(!strncasecmp(ptr, "BODY", 4)) ) {
                        char *pBody = NULL;
-                       
+
                        if (!strncasecmp(ptr, "BODY", 4)) {
                                pBody = ptr;
                        }
@@ -234,8 +240,8 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                                                cid_start = src + 4;
                                                cid_end = cid_start;
                                                while ((*cid_end != '"') && 
-                                                      !isspace(*cid_end) &&
-                                                      (cid_end < ptr))
+                                                               !isspace(*cid_end) &&
+                                                               (cid_end < ptr))
                                                        cid_end ++;
 
                                                /* copy tag and attributes up to src="cid: */
@@ -244,13 +250,13 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                                                /* add in /webcit/mimepart/<msgno>/CID/ 
                                                   trailing / stops dumb URL filters getting excited */
                                                StrBufAppendPrintf(BodyArea,
-                                                                  "/webcit/mimepart/%d/",msgnum);
+                                                               "/webcit/mimepart/%d/",msgnum);
                                                StrBufAppendBufPlain(BodyArea, cid_start, cid_end - cid_start, 0);
 
                                                if (ptr - cid_end > 0)
                                                        StrBufAppendBufPlain(BodyArea, 
-                                                                            cid_end + 1, 
-                                                                            ptr - cid_end, 0);
+                                                                       cid_end + 1, 
+                                                                       ptr - cid_end, 0);
                                        }
                                        else 
                                                StrBufAppendBufPlain(BodyArea, pBody, ptr - pBody, 0);
@@ -262,16 +268,16 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                        msgstart = ptr;
                }
 
-               /**
+               /*
                 * Any of these tags cause everything including and following
                 * the tag to be removed.
                 */
                if ( (!strncasecmp(ptr, "/HTML", 5))
-                  ||(!strncasecmp(ptr, "/BODY", 5)) ) {
+                               ||(!strncasecmp(ptr, "/BODY", 5)) ) {
                        --ptr;
                        msgend = ptr;
                        strcpy(ptr, "");
-                       
+
                }
 
                ++ptr;
@@ -280,21 +286,21 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                strcpy(msg, msgstart);
        }
 
-       /** Now go through the message, parsing tags as necessary. */
+       /* Now go through the message, parsing tags as necessary. */
        converted_msg = NewStrBufPlain(NULL, content_length + 8192);
 
 
        /** Convert foreign character sets to UTF-8 if necessary. */
 #ifdef HAVE_ICONV
        if ( (strcasecmp(charset, "us-ascii"))
-          && (strcasecmp(charset, "UTF-8"))
-          && (strcasecmp(charset, ""))
-       ) {
-               lprintf(9, "Converting %s to UTF-8\n", charset);
+                       && (strcasecmp(charset, "UTF-8"))
+                       && (strcasecmp(charset, ""))
+          ) {
+               syslog(9, "Converting %s to UTF-8\n", charset);
                ctdl_iconv_open("UTF-8", charset, &ic);
                if (ic == (iconv_t)(-1) ) {
-                       lprintf(5, "%s:%d iconv_open() failed: %s\n",
-                               __FILE__, __LINE__, strerror(errno));
+                       syslog(5, "%s:%d iconv_open() failed: %s\n",
+                                       __FILE__, __LINE__, strerror(errno));
                }
        }
        if  (Source == NULL) {
@@ -321,10 +327,10 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                        msg = (char*)ChrPtr(Source); /* TODO: get rid of this. */
                }
        }
-               
+
 #endif
 
-       /**
+       /*
         *      At this point, the message has been stripped down to
         *      only the content inside the <BODY></BODY> tags, and has
         *      been converted to UTF-8 if it was originally in a foreign
@@ -333,7 +339,7 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
         */
 
        if (converted_msg == NULL) {
-               StrBufAppendPrintf(Target, "Error %d: %s<br />%s:%d", errno, strerror(errno), __FILE__, __LINE__);
+               StrBufAppendPrintf(Target, "Error %d: %s<br>%s:%d", errno, strerror(errno), __FILE__, __LINE__);
                goto BAIL;
        }
 
@@ -366,7 +372,7 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                if (!strncasecmp(ptr, "<a href=\"mailto:", 16)) {
                        content_length += 64;
                        StrBufAppendPrintf(converted_msg,
-                               "<a href=\"display_enter?force_room=_MAIL_&recp=");
+                                       "<a href=\"display_enter?force_room=_MAIL_?recp=");
                        ptr = &ptr[16];
                        ++alevel;
                        ++brak;
@@ -376,15 +382,22 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                        ++alevel;
                        ++brak;
                        if ( ((strchr(ptr, ':') < strchr(ptr, '/')))
-                            &&  ((strchr(ptr, '/') < strchr(ptr, '>'))) 
-                            ) {
+                                       &&  ((strchr(ptr, '/') < strchr(ptr, '>'))) 
+                          ) {
                                /* open external links to new window */
                                StrBufAppendPrintf(converted_msg, new_window);
                                ptr = &ptr[8];
                        }
-                       else if ( (treat_as_wiki) && (strncasecmp(ptr, "<a href=\"wiki?", 14)) ) {
+                       else if (
+                               (treat_as_wiki)
+                               && (strncasecmp(ptr, "<a href=\"wiki?", 14))
+                               && (strncasecmp(ptr, "<a href=\"dotgoto?", 17))
+                               && (strncasecmp(ptr, "<a href=\"knrooms?", 17))
+                       ) {
                                content_length += 64;
-                               StrBufAppendPrintf(converted_msg, "<a href=\"wiki?page=");
+                               StrBufAppendPrintf(converted_msg, "<a href=\"wiki?go=");
+                               StrBufUrlescAppend(converted_msg, WC->CurRoom.name, NULL);
+                               StrBufAppendPrintf(converted_msg, "?page=");
                                ptr = &ptr[9];
                        }
                        else {
@@ -394,27 +407,39 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                }
                /** Fixup <img src="cid:... ...> to fetch the mime part */
                else if (!strncasecmp(ptr, "<img ", 5)) {
-                       char* tag_end=strchr(ptr,'>');
-                       char* src=strstr(ptr, " src=\"cid:");
                        char *cid_start, *cid_end;
-                       ++brak;
+                       char* tag_end=strchr(ptr,'>');
+                       char* src;
+                       /* FIXME - handle this situation (maybe someone opened an <img cid... 
+                        * and then ended the message)
+                        */
+                       if (!tag_end) {
+                               syslog(9, "tag_end is null and ptr is:\n");
+                               syslog(9, "%s\n", ptr);
+                               syslog(9, "Theoretical bytes remaining: %d\n", (int)(msgend - ptr));
+                       }
 
-                       if (src && 
-                               (cid_start=strchr(src,':')) && 
-                               (cid_end=strchr(cid_start,'"')) &&
-                               (cid_end < tag_end)) {
+                       src=strstr(ptr, "src=\"cid:");
+                       ++brak;
 
+                       if (src
+                           && isspace(*(src-1))
+                               && tag_end
+                               && (cid_start=strchr(src,':'))
+                               && (cid_end=strchr(cid_start,'"'))
+                               && (cid_end < tag_end)
+                       ) {
                                /* copy tag and attributes up to src="cid: */
                                StrBufAppendBufPlain(converted_msg, ptr, src - ptr, 0);
                                cid_start++;
 
                                /* add in /webcit/mimepart/<msgno>/CID/ 
-                                  trailing / stops dumb URL filters getting excited */
+                                  trailing / stops dumb URL filters getting excited */
                                StrBufAppendPrintf(converted_msg,
-                                       "src=\"/webcit/mimepart/%d/",msgnum);
+                                               " src=\"/webcit/mimepart/%d/",msgnum);
                                StrBufAppendBufPlain(converted_msg, cid_start, cid_end - cid_start, 0);
                                StrBufAppendBufPlain(converted_msg, "/\"", -1, 0);
-                               
+
                                ptr = cid_end+1;
                        }
                        StrBufAppendBufPlain(converted_msg, ptr, tag_end - ptr, 0);
@@ -461,7 +486,7 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                                        char *ltreviewptr;
                                        char *nbspreviewptr;
                                        char linkedchar;
-                                       int len = linklen;
+                                       int len;
                                        
                                        len = linklen;
                                        linkedchar = ptr[len];
@@ -497,23 +522,27 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
                        ptr++;
                }
 
-               /**
-                * We need to know when we're inside a tag,
-                * so we don't turn things that look like URL's into
-                * links, when they're already links - or image sources.
-                */
-               if (*(ptr-1) == '<') {
-                       ++brak;
-               }
-               if (*(ptr-1) == '>') {
-                       --brak;
-                       if ((scriptlevel == 0) && (script_start_pos >= 0)) {
-                               StrBufCutRight(converted_msg, StrLength(converted_msg) - script_start_pos);
-                               script_start_pos = (-1);
+
+               if ((ptr >= msg) && (ptr <= msgend)) {
+                       /*
+                        * We need to know when we're inside a tag,
+                        * so we don't turn things that look like URL's into
+                        * links, when they're already links - or image sources.
+                        */
+                       if ((ptr > msg) && (*(ptr-1) == '<')) {
+                               ++brak;
+                       }
+                       if ((ptr > msg) && (*(ptr-1) == '>')) {
+                               --brak;
+                               if ((scriptlevel == 0) && (script_start_pos >= 0)) {
+                                       StrBufCutRight(converted_msg, StrLength(converted_msg) - script_start_pos);
+                                       script_start_pos = (-1);
+                               }
                        }
+                       if (!strncasecmp(ptr, "</A>", 3)) --alevel;
                }
-               if (!strncasecmp(ptr, "</A>", 3)) --alevel;
        }
+
        if (BodyArea != NULL) {
                StrBufAppendBufPlain(converted_msg, HKEY("</td></tr></table>"), 0);  
                FreeStrBuf(&BodyArea);
@@ -527,7 +556,7 @@ void output_html(const char *supplied_charset, int treat_as_wiki, int msgnum, St
        StrBufAppendBuf(Target, converted_msg, 0);
 
 BAIL:  /** A little trailing vertical whitespace... */
-       StrBufAppendPrintf(Target, "<br /><br />\n");
+       StrBufAppendPrintf(Target, "<br><br>\n");
 
        /** Now give back the memory */
        FreeStrBuf(&converted_msg);
@@ -600,6 +629,8 @@ void UrlizeText(StrBuf* Target, StrBuf *Source, StrBuf *WrkBuf)
        if (TrailerLen > 0)
                StrBufAppendBufPlain(Target, end, TrailerLen, 0);
 }
+
+
 void url(char *buf, size_t bufsize)
 {
        int len, UrlLen, Offset, TrailerLen, outpos;
@@ -610,7 +641,7 @@ void url(char *buf, size_t bufsize)
        start = NULL;
        len = strlen(buf);
        if (len > bufsize) {
-               lprintf(1, "URL: content longer than buffer!");
+               syslog(1, "URL: content longer than buffer!");
                return;
        }
        end = buf + len;
@@ -646,7 +677,7 @@ void url(char *buf, size_t bufsize)
        
        UrlLen = end - start;
        if (UrlLen > sizeof(urlbuf)){
-               lprintf(1, "URL: content longer than buffer!");
+               syslog(1, "URL: content longer than buffer!");
                return;
        }
        memcpy(urlbuf, start, UrlLen);
@@ -659,7 +690,7 @@ void url(char *buf, size_t bufsize)
                          "%ca href=%c%s%c TARGET=%c%s%c%c%s%c/A%c",
                          LB, QU, urlbuf, QU, QU, TARGET, QU, RB, urlbuf, LB, RB);
        if (outpos >= sizeof(outbuf) - Offset) {
-               lprintf(1, "URL: content longer than buffer!");
+               syslog(1, "URL: content longer than buffer!");
                return;
        }
 
@@ -667,14 +698,10 @@ void url(char *buf, size_t bufsize)
        if (TrailerLen > 0)
                memcpy(outbuf + Offset + outpos, end, TrailerLen);
        if (Offset + outpos + TrailerLen > bufsize) {
-               lprintf(1, "URL: content longer than buffer!");
+               syslog(1, "URL: content longer than buffer!");
                return;
        }
        memcpy (buf, outbuf, Offset + outpos + TrailerLen);
        *(buf + Offset + outpos + TrailerLen) = '\0';
 }
 
-
-
-
-/*@}*/