1 files changed, 330 insertions, 322 deletions

diff --git a/src/plugins/html_extractor.c b/src/plugins/html_extractor.c
index 8cd4aba..5ebf97b 100644
--- a/src/plugins/html_extractor.c
+++ b/src/plugins/html_extractor.c
@@ -87,9 +87,9 @@ tag_to_type (const char *tag)
 {
   unsigned int i;
 
-  for (i=0; NULL != tagmap[i].name; i++)
+  for (i = 0; NULL != tagmap[i].name; i++)
     if (0 == strcasecmp (tag,
-                         tagmap[i].name))
+                         tagmap[i].name))
       return tagmap[i].type;
   return EXTRACTOR_METATYPE_RESERVED;
 }
@@ -107,10 +107,10 @@ tag_to_type (const char *tag)
  */
 static Bool TIDY_CALL
 report_cb (TidyDoc doc,
-           TidyReportLevel lvl,
-           uint line,
-           uint col,
-           ctmbstr mssg)
+           TidyReportLevel lvl,
+           uint line,
+           uint col,
+           ctmbstr mssg)
 {
   return 0;
 }
@@ -130,7 +130,7 @@ get_byte_cb (void *sourceData)
 
   if (1 !=
       ec->read (ec->cls,
-                &data, 1))
+                &data, 1))
     return EndOfStream;
   return *(unsigned char*) data;
 }
@@ -188,130 +188,129 @@ EXTRACTOR_html_extract_method (struct EXTRACTOR_ExtractContext *ec)
   const char *mime;
 
   if (-1 == (iret = ec->read (ec->cls,
-                              &data,
-                              16 * 1024)))
+                              &data,
+                              16 * 1024)))
     return;
   if (NULL == (mime = magic_buffer (magic, data, iret)))
     return;
   if (0 != strncmp (mime,
-                    "text/html",
-                    strlen ("text/html")))
+                    "text/html",
+                    strlen ("text/html")))
     return; /* not HTML */
 
   if (0 != ec->seek (ec->cls, 0, SEEK_SET))
     return; /* seek failed !? */
 
   tidyInitSource (&src, ec,
-                  &get_byte_cb,
-                  &unget_byte_cb,
-                  &eof_cb);
+                  &get_byte_cb,
+                  &unget_byte_cb,
+                  &eof_cb);
   if (NULL == (doc = tidyCreate ()))
     return;
   tidySetReportFilter (doc, &report_cb);
   tidySetAppData (doc, ec);
   if (0 > tidyParseSource (doc, &src))
-    {
-      tidyRelease (doc);
-      return;
-    }
+  {
+    tidyRelease (doc);
+    return;
+  }
   if (1 != tidyStatus (doc))
-    {
-      tidyRelease (doc);
-      return;
-    }
+  {
+    tidyRelease (doc);
+    return;
+  }
   if (NULL == (head = tidyGetHead (doc)))
-    {
-      fprintf (stderr, "no head\n");
-      tidyRelease (doc);
-      return;
-    }
+  {
+    fprintf (stderr, "no head\n");
+    tidyRelease (doc);
+    return;
+  }
   for (child = tidyGetChild (head); NULL != child; child = tidyGetNext (child))
+  {
+    switch (tidyNodeGetType (child))
     {
-      switch (tidyNodeGetType(child))
-        {
-        case TidyNode_Root:
-          break;
-        case TidyNode_DocType:
-          break;
-        case TidyNode_Comment:
-          break;
-        case TidyNode_ProcIns:
-          break;
-        case TidyNode_Text:
-          break;
-        case TidyNode_CDATA:
-          break;
-        case TidyNode_Section:
-          break;
-        case TidyNode_Asp:
-          break;
-        case TidyNode_Jste:
-          break;
-        case TidyNode_Php:
-          break;
-        case TidyNode_XmlDecl:
-          break;
-        case TidyNode_Start:
-        case TidyNode_StartEnd:
-          name = tidyNodeGetName (child);
-          if ( (0 == strcasecmp (name, "title")) &&
-               (NULL != (title = tidyGetChild (child))) )
-            {
-              tidyBufInit (&tbuf);
-              tidyNodeGetValue (doc, title, &tbuf);
-              /* add 0-termination */
-              tidyBufPutByte (&tbuf, 0);
-              if (0 !=
-                  ec->proc (ec->cls,
-                            "html",
-                            EXTRACTOR_METATYPE_TITLE,
-                            EXTRACTOR_METAFORMAT_UTF8,
-                            "text/plain",
-                            (const char *) tbuf.bp,
-                            tbuf.size))
-                {
-                  tidyBufFree (&tbuf);
-                  goto CLEANUP;
-                }
-              tidyBufFree (&tbuf);
-              break;
-            }
-          if (0 == strcasecmp (name, "meta"))
-            {
-              if (NULL == (attr = tidyAttrGetById (child,
-                                                   TidyAttr_NAME)))
-                break;
-              if (EXTRACTOR_METATYPE_RESERVED ==
-                  (type = tag_to_type (tidyAttrValue (attr))))
-                break;
-              if (NULL == (attr = tidyAttrGetById (child,
-                                                   TidyAttr_CONTENT)))
-                break;
-              name = tidyAttrValue (attr);
-              if (0 !=
-                  ec->proc (ec->cls,
-                            "html",
-                            type,
-                            EXTRACTOR_METAFORMAT_UTF8,
-                            "text/plain",
-                            name,
-                            strlen (name) + 1))
-                goto CLEANUP;
-              break;
-            }
-          break;
-        case TidyNode_End:
-          break;
-        default:
-          break;
-        }
+    case TidyNode_Root:
+      break;
+    case TidyNode_DocType:
+      break;
+    case TidyNode_Comment:
+      break;
+    case TidyNode_ProcIns:
+      break;
+    case TidyNode_Text:
+      break;
+    case TidyNode_CDATA:
+      break;
+    case TidyNode_Section:
+      break;
+    case TidyNode_Asp:
+      break;
+    case TidyNode_Jste:
+      break;
+    case TidyNode_Php:
+      break;
+    case TidyNode_XmlDecl:
+      break;
+    case TidyNode_Start:
+    case TidyNode_StartEnd:
+      name = tidyNodeGetName (child);
+      if ( (0 == strcasecmp (name, "title")) &&
+           (NULL != (title = tidyGetChild (child))) )
+      {
+        tidyBufInit (&tbuf);
+        tidyNodeGetValue (doc, title, &tbuf);
+        /* add 0-termination */
+        tidyBufPutByte (&tbuf, 0);
+        if (0 !=
+            ec->proc (ec->cls,
+                      "html",
+                      EXTRACTOR_METATYPE_TITLE,
+                      EXTRACTOR_METAFORMAT_UTF8,
+                      "text/plain",
+                      (const char *) tbuf.bp,
+                      tbuf.size))
+        {
+          tidyBufFree (&tbuf);
+          goto CLEANUP;
+        }
+        tidyBufFree (&tbuf);
+        break;
+      }
+      if (0 == strcasecmp (name, "meta"))
+      {
+        if (NULL == (attr = tidyAttrGetById (child,
+                                             TidyAttr_NAME)))
+          break;
+        if (EXTRACTOR_METATYPE_RESERVED ==
+            (type = tag_to_type (tidyAttrValue (attr))))
+          break;
+        if (NULL == (attr = tidyAttrGetById (child,
+                                             TidyAttr_CONTENT)))
+          break;
+        name = tidyAttrValue (attr);
+        if (0 !=
+            ec->proc (ec->cls,
+                      "html",
+                      type,
+                      EXTRACTOR_METAFORMAT_UTF8,
+                      "text/plain",
+                      name,
+                      strlen (name) + 1))
+          goto CLEANUP;
+        break;
+      }
+      break;
+    case TidyNode_End:
+      break;
+    default:
+      break;
     }
- CLEANUP:
+  }
+CLEANUP:
   tidyRelease (doc);
 }
 
 
-
 #if OLD
 
 
@@ -323,66 +322,71 @@ tagMatch (const char *tag, const char *s, const char *e)
   return (((e - s) == strlen (tag)) && (0 == strncasecmp (tag, s, e - s)));
 }
 
+
 static int
-lookFor (char c, size_t * pos, const char *data, size_t size)
+lookFor (char c, size_t *pos, const char *data, size_t size)
 {
   size_t p = *pos;
 
   while ((p < size) && (data[p] != c))
-    {
-      if (data[p] == '\0')
-        return 0;
-      p++;
-    }
+  {
+    if (data[p] == '\0')
+      return 0;
+    p++;
+  }
   *pos = p;
   return p < size;
 }
 
+
 static int
-skipWhitespace (size_t * pos, const char *data, size_t size)
+skipWhitespace (size_t *pos, const char *data, size_t size)
 {
   size_t p = *pos;
 
   while ((p < size) && (isspace ( (unsigned char) data[p])))
-    {
-      if (data[p] == '\0')
-        return 0;
-      p++;
-    }
+  {
+    if (data[p] == '\0')
+      return 0;
+    p++;
+  }
   *pos = p;
   return p < size;
 }
 
+
 static int
-skipLetters (size_t * pos, const char *data, size_t size)
+skipLetters (size_t *pos, const char *data, size_t size)
 {
   size_t p = *pos;
 
   while ((p < size) && (isalpha ( (unsigned char) data[p])))
-    {
-      if (data[p] == '\0')
-        return 0;
-      p++;
-    }
+  {
+    if (data[p] == '\0')
+      return 0;
+    p++;
+  }
   *pos = p;
   return p < size;
 }
 
+
 static int
-lookForMultiple (const char *c, size_t * pos, const char *data, size_t size)
+lookForMultiple (const char *c, size_t *pos, const char *data, size_t size)
 {
   size_t p = *pos;
 
   while ((p < size) && (strchr (c, data[p]) == NULL))
-    {
-      if (data[p] == '\0')
-        return 0;
-      p++;
-    }
+  {
+    if (data[p] == '\0')
+      return 0;
+    p++;
+  }
   *pos = p;
   return p < size;
 }
 
+
 static void
 findEntry (const char *key,
            const char *start,
@@ -394,32 +398,33 @@ findEntry (const char *key,
   *mend = NULL;
   len = strlen (key);
   while (start < end - len - 1)
+  {
+    start++;
+    if (start[len] != '=')
+      continue;
+    if (0 == strncasecmp (start, key, len))
     {
-      start++;
-      if (start[len] != '=')
-        continue;
-      if (0 == strncasecmp (start, key, len))
-        {
-          start += len + 1;
-          *mstart = start;
-          if ((*start == '\"') || (*start == '\''))
-            {
-              start++;
-              while ((start < end) && (*start != **mstart))
-                start++;
-              (*mstart)++;      /* skip quote */
-            }
-          else
-            {
-              while ((start < end) && (!isspace ( (unsigned char) *start)))
-                start++;
-            }
-          *mend = start;
-          return;
-        }
+      start += len + 1;
+      *mstart = start;
+      if ((*start == '\"') || (*start == '\''))
+      {
+        start++;
+        while ((start < end) && (*start != **mstart))
+          start++;
+        (*mstart)++;            /* skip quote */
+      }
+      else
+      {
+        while ((start < end) && (! isspace ( (unsigned char) *start)))
+          start++;
+      }
+      *mend = start;
+      return;
     }
+  }
 }
 
+
 /**
  * Search all tags that correspond to "tagname".  Example:
  * If the tag is <meta name="foo" desc="bar">, and
@@ -430,7 +435,7 @@ findEntry (const char *key,
  * @return NULL if nothing is found
  */
 static char *
-findInTags (struct TagInfo * t,
+findInTags (struct TagInfo *t,
             const char *tagname,
             const char *keyname, const char *keyvalue, const char *searchname)
 {
@@ -438,26 +443,26 @@ findInTags (struct TagInfo * t,
   const char *pend;
 
   while (t != NULL)
+  {
+    if (tagMatch (tagname, t->tagStart, t->tagEnd))
     {
-      if (tagMatch (tagname, t->tagStart, t->tagEnd))
+      findEntry (keyname, t->tagEnd, t->dataStart, &pstart, &pend);
+      if ((pstart != NULL) && (tagMatch (keyvalue, pstart, pend)))
+      {
+        findEntry (searchname, t->tagEnd, t->dataStart, &pstart, &pend);
+        if (pstart != NULL)
         {
-          findEntry (keyname, t->tagEnd, t->dataStart, &pstart, &pend);
-          if ((pstart != NULL) && (tagMatch (keyvalue, pstart, pend)))
-            {
-              findEntry (searchname, t->tagEnd, t->dataStart, &pstart, &pend);
-              if (pstart != NULL)
-                {
-                  char *ret = malloc (pend - pstart + 1);
-                  if (ret == NULL)
-                    return NULL;
-                  memcpy (ret, pstart, pend - pstart);
-                  ret[pend - pstart] = '\0';
-                  return ret;
-                }
-            }
+          char *ret = malloc (pend - pstart + 1);
+          if (ret == NULL)
+            return NULL;
+          memcpy (ret, pstart, pend - pstart);
+          ret[pend - pstart] = '\0';
+          return ret;
         }
-      t = t->next;
+      }
     }
+    t = t->next;
+  }
   return NULL;
 }
 
@@ -465,10 +470,10 @@ findInTags (struct TagInfo * t,
 /* mimetype = text/html */
 int
 EXTRACTOR_html_extract (const char *data,
-                        size_t size,
-                        EXTRACTOR_MetaDataProcessor proc,
-                        void *proc_cls,
-                        const char *options)
+                        size_t size,
+                        EXTRACTOR_MetaDataProcessor proc,
+                        void *proc_cls,
+                        const char *options)
 {
   size_t xsize;
   struct TagInfo *tags;
@@ -494,60 +499,60 @@ EXTRACTOR_html_extract (const char *data,
   tag.next = NULL;
   pos = 0;
   while (pos < xsize)
+  {
+    if (! lookFor ('<', &pos, data, size))
+      break;
+    tag.tagStart = &data[++pos];
+    if (! skipLetters (&pos, data, size))
+      break;
+    tag.tagEnd = &data[pos];
+    if (! skipWhitespace (&pos, data, size))
+      break;
+STEP3:
+    if (! lookForMultiple (">\"\'", &pos, data, size))
+      break;
+    if (data[pos] != '>')
     {
-      if (!lookFor ('<', &pos, data, size))
-        break;
-      tag.tagStart = &data[++pos];
-      if (!skipLetters (&pos, data, size))
-        break;
-      tag.tagEnd = &data[pos];
-      if (!skipWhitespace (&pos, data, size))
-        break;
-    STEP3:
-      if (!lookForMultiple (">\"\'", &pos, data, size))
-        break;
-      if (data[pos] != '>')
-        {
-          /* find end-quote, ignore escaped quotes (\') */
-          do
-            {
-              tpos = pos;
-              pos++;
-              if (!lookFor (data[tpos], &pos, data, size))
-                break;
-            }
-          while (data[pos - 1] == '\\');
-          pos++;
-          goto STEP3;
-        }
+      /* find end-quote, ignore escaped quotes (\') */
+      do
+      {
+        tpos = pos;
+        pos++;
+        if (! lookFor (data[tpos], &pos, data, size))
+          break;
+      }
+      while (data[pos - 1] == '\\');
       pos++;
-      if (!skipWhitespace (&pos, data, size))
-        break;
-      tag.dataStart = &data[pos];
-      if (!lookFor ('<', &pos, data, size))
-        break;
-      tag.dataEnd = &data[pos];
-      i = 0;
-      while (relevantTags[i] != NULL)
-        {
-          if ((strlen (relevantTags[i]) == tag.tagEnd - tag.tagStart) &&
-              (0 == strncasecmp (relevantTags[i],
-                                 tag.tagStart, tag.tagEnd - tag.tagStart)))
-            {
-              t = malloc (sizeof (struct TagInfo));
-              if (t == NULL)
-                return 0;
-              *t = tag;
-              t->next = tags;
-              tags = t;
-              break;
-            }
-          i++;
-        }
-      /* abort early if we hit the body tag */
-      if (tagMatch ("body", tag.tagStart, tag.tagEnd))
+      goto STEP3;
+    }
+    pos++;
+    if (! skipWhitespace (&pos, data, size))
+      break;
+    tag.dataStart = &data[pos];
+    if (! lookFor ('<', &pos, data, size))
+      break;
+    tag.dataEnd = &data[pos];
+    i = 0;
+    while (relevantTags[i] != NULL)
+    {
+      if ((strlen (relevantTags[i]) == tag.tagEnd - tag.tagStart) &&
+          (0 == strncasecmp (relevantTags[i],
+                             tag.tagStart, tag.tagEnd - tag.tagStart)))
+      {
+        t = malloc (sizeof (struct TagInfo));
+        if (t == NULL)
+          return 0;
+        *t = tag;
+        t->next = tags;
+        tags = t;
         break;
+      }
+      i++;
     }
+    /* abort early if we hit the body tag */
+    if (tagMatch ("body", tag.tagStart, tag.tagEnd))
+      break;
+  }
 
   /* fast exit */
   if (tags == NULL)
@@ -557,110 +562,112 @@ EXTRACTOR_html_extract (const char *data,
   /* first, try to determine mime type and/or character set */
   tmp = findInTags (tags, "meta", "http-equiv", "content-type", "content");
   if (tmp != NULL)
-    {
-      /* ideally, tmp == "test/html; charset=ISO-XXXX-Y" or something like that;
-         if text/html is present, we take that as the mime-type; if charset=
-         is present, we try to use that for character set conversion. */
-      if (0 == strncasecmp (tmp, "text/html", strlen ("text/html")))
-        ret = proc (proc_cls,
-                    "html",
-                    EXTRACTOR_METATYPE_MIMETYPE,
-                    EXTRACTOR_METAFORMAT_UTF8,
-                    "text/plain",
-                    "text/html",
-                    strlen ("text/html")+1);
-      charset = strcasestr (tmp, "charset=");
-      if (charset != NULL)
-        charset = strdup (&charset[strlen ("charset=")]);
-      free (tmp);
-    }
+  {
+    /* ideally, tmp == "test/html; charset=ISO-XXXX-Y" or something like that;
+       if text/html is present, we take that as the mime-type; if charset=
+       is present, we try to use that for character set conversion. */
+    if (0 == strncasecmp (tmp, "text/html", strlen ("text/html")))
+      ret = proc (proc_cls,
+                  "html",
+                  EXTRACTOR_METATYPE_MIMETYPE,
+                  EXTRACTOR_METAFORMAT_UTF8,
+                  "text/plain",
+                  "text/html",
+                  strlen ("text/html") + 1);
+    charset = strcasestr (tmp, "charset=");
+    if (charset != NULL)
+      charset = strdup (&charset[strlen ("charset=")]);
+    free (tmp);
+  }
   i = 0;
   while (tagmap[i].name != NULL)
+  {
+    tmp = findInTags (tags, "meta", "name", tagmap[i].name, "content");
+    if ( (tmp != NULL) &&
+         (ret == 0) )
     {
-      tmp = findInTags (tags, "meta", "name", tagmap[i].name, "content");
-      if ( (tmp != NULL) &&
-           (ret == 0) )
+      if (charset == NULL)
+      {
+        ret = proc (proc_cls,
+                    "html",
+                    tagmap[i].type,
+                    EXTRACTOR_METAFORMAT_C_STRING,
+                    "text/plain",
+                    tmp,
+                    strlen (tmp) + 1);
+      }
+      else
+      {
+        xtmp = EXTRACTOR_common_convert_to_utf8 (tmp,
+                                                 strlen (tmp),
+                                                 charset);
+        if (xtmp != NULL)
         {
-          if (charset == NULL)
-            {
-              ret = proc (proc_cls,
-                          "html",
-                          tagmap[i].type,
-                          EXTRACTOR_METAFORMAT_C_STRING,
-                          "text/plain",
-                          tmp,
-                          strlen (tmp) + 1);
-            }
-          else
-            {
-              xtmp = EXTRACTOR_common_convert_to_utf8 (tmp,
-                                                       strlen (tmp),
-                                                       charset);
-              if (xtmp != NULL)
-                {
-                  ret = proc (proc_cls,
-                              "html",
-                              tagmap[i].type,
-                              EXTRACTOR_METAFORMAT_UTF8,
-                              "text/plain",
-                              xtmp,
-                              strlen (xtmp) + 1);
-                  free (xtmp);
-                }
-            }
+          ret = proc (proc_cls,
+                      "html",
+                      tagmap[i].type,
+                      EXTRACTOR_METAFORMAT_UTF8,
+                      "text/plain",
+                      xtmp,
+                      strlen (xtmp) + 1);
+          free (xtmp);
         }
-      if (tmp != NULL)
-        free (tmp);
-      i++;
+      }
     }
+    if (tmp != NULL)
+      free (tmp);
+    i++;
+  }
   while (tags != NULL)
+  {
+    t = tags;
+    if ( (tagMatch ("title", t->tagStart, t->tagEnd)) &&
+         (ret == 0) )
     {
-      t = tags;
-      if ( (tagMatch ("title", t->tagStart, t->tagEnd)) &&
-           (ret == 0) )
-        {
-          if (charset == NULL)
-            {
-              xtmp = malloc (t->dataEnd - t->dataStart + 1);
-              if (xtmp != NULL)
-                {
-                  memcpy (xtmp, t->dataStart, t->dataEnd - t->dataStart);
-                  xtmp[t->dataEnd - t->dataStart] = '\0';
-                  ret = proc (proc_cls,
-                              "html",
-                              EXTRACTOR_METATYPE_TITLE,
-                              EXTRACTOR_METAFORMAT_C_STRING,
-                              "text/plain",
-                              xtmp,
-                              strlen (xtmp) + 1);
-                  free (xtmp);
-                }
-            }
-          else
-            {
-              xtmp = EXTRACTOR_common_convert_to_utf8 (t->dataStart,
-                                                       t->dataEnd - t->dataStart,
-                                                       charset);
-              if (xtmp != NULL)
-                {
-                  ret = proc (proc_cls,
-                              "html",
-                              EXTRACTOR_METATYPE_TITLE,
-                              EXTRACTOR_METAFORMAT_UTF8,
-                              "text/plain",
-                              xtmp,
-                              strlen (xtmp) + 1);
-                  free (xtmp);
-                }
-            }
-        }
-      tags = t->next;
-      free (t);
+      if (charset == NULL)
+      {
+        xtmp = malloc (t->dataEnd - t->dataStart + 1);
+        if (xtmp != NULL)
+        {
+          memcpy (xtmp, t->dataStart, t->dataEnd - t->dataStart);
+          xtmp[t->dataEnd - t->dataStart] = '\0';
+          ret = proc (proc_cls,
+                      "html",
+                      EXTRACTOR_METATYPE_TITLE,
+                      EXTRACTOR_METAFORMAT_C_STRING,
+                      "text/plain",
+                      xtmp,
+                      strlen (xtmp) + 1);
+          free (xtmp);
+        }
+      }
+      else
+      {
+        xtmp = EXTRACTOR_common_convert_to_utf8 (t->dataStart,
+                                                 t->dataEnd - t->dataStart,
+                                                 charset);
+        if (xtmp != NULL)
+        {
+          ret = proc (proc_cls,
+                      "html",
+                      EXTRACTOR_METATYPE_TITLE,
+                      EXTRACTOR_METAFORMAT_UTF8,
+                      "text/plain",
+                      xtmp,
+                      strlen (xtmp) + 1);
+          free (xtmp);
+        }
+      }
     }
+    tags = t->next;
+    free (t);
+  }
   if (charset != NULL)
     free (charset);
   return ret;
 }
+
+
 #endif
 
 
@@ -672,9 +679,9 @@ html_gobject_init ()
 {
   magic = magic_open (MAGIC_MIME_TYPE);
   if (0 != magic_load (magic, NULL))
-    {
-      /* FIXME: how to deal with errors? */
-    }
+  {
+    /* FIXME: how to deal with errors? */
+  }
 }
 
 
@@ -685,10 +692,11 @@ void __attribute__ ((destructor))
 html_ltdl_fini ()
 {
   if (NULL != magic)
-    {
-      magic_close (magic);
-      magic = NULL;
-    }
+  {
+    magic_close (magic);
+    magic = NULL;
+  }
 }
 
+
 /* end of html_extractor.c */