More fixes to html parsing.

diff --git a/src/libserver/html.c b/src/libserver/html.c
index b979b3f8c26c127f6ded591225609a5b9c509516..0f7f987585efe4437ed9b3fc7e81bc91b8fe6813 100644 (file)
--- a/src/libserver/html.c
+++ b/src/libserver/html.c
@@ -638,8 +638,16 @@ tag_cmp (const void *m1, const void *m2)
 {
        const struct html_tag_def *p1 = m1;
        const struct html_tag_def *p2 = m2;
+       gsize l1, l2;
 
-       return g_ascii_strcasecmp (p1->name, p2->name);
+       l1 = strlen (p1->name);
+       l2 = strlen (p2->name);
+
+       if (l1 == l2) {
+               return g_ascii_strcasecmp (p1->name, p2->name);
+       }
+
+       return l1 - l2;
 }
 
 static gint
@@ -647,8 +655,15 @@ tag_find (const void *skey, const void *elt)
 {
        const struct html_tag *tag = skey;
        const struct html_tag_def *d = elt;
+       gsize tlen;
 
-       return g_ascii_strncasecmp (tag->name.start, d->name, tag->name.len);
+       tlen = strlen (d->name);
+
+       if (tlen == tag->name.len) {
+               return g_ascii_strncasecmp (tag->name.start, d->name, tag->name.len);
+       }
+
+       return tag->name.len - tlen;
 }
 
 static gint
@@ -1058,7 +1073,7 @@ rspamd_html_process_tag (rspamd_mempool_t *pool, struct html_content *hc,
                return TRUE;
        }
 
-       return FALSE;
+       return TRUE;
 }
 
 static gboolean
@@ -1371,12 +1386,12 @@ rspamd_html_process_part (rspamd_mempool_t *pool, struct html_content *hc,
 {
        const guchar *p, *c, *end, *tag_start = NULL, *savep = NULL;
        guchar t;
-       gboolean closing = FALSE, need_decode = FALSE;
+       gboolean closing = FALSE, need_decode = FALSE, save_space = FALSE;
        GByteArray *dest;
        guint obrace = 0, ebrace = 0;
        GNode *cur_level = NULL;
        gint substate, len;
-       struct html_tag *cur_tag;
+       struct html_tag *cur_tag = NULL;
        enum {
                parse_start = 0,
                tag_begin,
@@ -1558,11 +1573,13 @@ rspamd_html_process_part (rspamd_mempool_t *pool, struct html_content *hc,
                        break;
 
                case content_write:
+
                        if (t != '<') {
                                if (t == '&') {
                                        need_decode = TRUE;
                                }
                                else if (g_ascii_isspace (t)) {
+                                       save_space = TRUE;
 
                                        if (c != p) {
                                                if (need_decode) {
@@ -1579,6 +1596,16 @@ rspamd_html_process_part (rspamd_mempool_t *pool, struct html_content *hc,
                                        c = p;
                                        state = content_ignore_sp;
                                }
+                               else {
+                                       if (save_space) {
+                                               /* Append one space if needed */
+                                               if (dest->len > 0 &&
+                                                               !g_ascii_isspace (dest->data[dest->len - 1])) {
+                                                       g_byte_array_append (dest, " ", 1);
+                                               }
+                                               save_space = FALSE;
+                                       }
+                               }
                        }
                        else {
                                if (c != p) {
@@ -1605,14 +1632,6 @@ rspamd_html_process_part (rspamd_mempool_t *pool, struct html_content *hc,
                        if (!g_ascii_isspace (t)) {
                                c = p;
                                state = content_write;
-
-                               if (t != '<') {
-                                       /* Append one space if needed */
-                                       if (dest->len > 0 &&
-                                                       !g_ascii_isspace (dest->data[dest->len - 1])) {
-                                               g_byte_array_append (dest, " ", 1);
-                                       }
-                               }
                                continue;
                        }
 

diff --git a/test/lua/unit/html.lua b/test/lua/unit/html.lua
index 5c58e209acdc34f991ca78943893e86c5bd76e34..f9788c34919b2935accb23fb2dac7e7cec773dec 100644 (file)
--- a/test/lua/unit/html.lua
+++ b/test/lua/unit/html.lua
@@ -15,10 +15,31 @@ context("HTML processing", function()
   </head>
   <body>
     <!-- page content -->
-    Hello, world!
+    Hello, world! <b>test</b>
+    <p>data<>
+    </P>
+    <b>stuff</p>?
   </body>
 </html>
-      ]], 'Hello, world!'},
+      ]], 'Hello, world! test data stuff?'},
+      {[[
+<?xml version="1.0" encoding="iso-8859-1"?>
+ <!DOCTYPE html 
+   PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
+   "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
+ <html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
+   <head>
+     <title>
+       Wikibooks
+     </title>
+   </head>
+   <body>
+     <p>
+       Hello,          world! 
+       
+     </p>
+   </body>
+ </html>]], 'Hello, world!'},
     }
     
     for _,c in ipairs(cases) do