X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=src%2Fhtml_parser.cpp;h=95db7c7fd41993645193ba617e1d5404993a8307;hb=586d78659d671683f33ec55f4a7d32b28e345ccd;hp=22d4a31db94f9b4607b0e18e34348851052a9f25;hpb=dd17e564b39fa38a3d611853a3b13dacdff3d070;p=metaproxy-moved-to-github.git

diff --git a/src/html_parser.cpp b/src/html_parser.cpp
index 22d4a31..95db7c7 100644
--- a/src/html_parser.cpp
+++ b/src/html_parser.cpp
@@ -1,5 +1,5 @@
 /* This file is part of Metaproxy.
-   Copyright (C) 2005-2013 Index Data
+   Copyright (C) Index Data
 
 Metaproxy is free software; you can redistribute it and/or modify it under
 the terms of the GNU General Public License as published by the Free
@@ -24,9 +24,11 @@ Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 #include <stdlib.h>
 #include <ctype.h>
 #include <stdio.h>
+#include <yaz/matchstr.h>
 
 #define SPACECHR " \t\r\n\f"
 
+// http://www.whatwg.org/specs/web-apps/current-work/multipage/parsing.html
 
 namespace metaproxy_1 {
     class HTMLParser::Rep {
@@ -46,6 +48,7 @@ namespace metaproxy_1 {
         Rep();
         ~Rep();
         int m_verbose;
+        bool nest;
     };
 }
 
@@ -54,6 +57,7 @@ namespace mp = metaproxy_1;
 mp::HTMLParser::Rep::Rep()
 {
     m_verbose = 0;
+    nest = true;
 }
 
 mp::HTMLParser::Rep::~Rep()
@@ -95,7 +99,7 @@ static int skipSpace(const char *cp)
 static int skipName(const char *cp)
 {
     int i;
-    for (i = 0; cp[i] && !strchr(SPACECHR "/>=", cp[i]); i++)
+    for (i = 0; cp[i] && !strchr(SPACECHR "/><=", cp[i]); i++)
         ;
     return i;
 }
@@ -146,7 +150,7 @@ int mp::HTMLParser::Rep::tagAttrs(HTMLParserEvent &event,
                                   const char *cp)
 {
     int i = skipSpace(cp);
-    while (cp[i] && cp[i] != '>' && cp[i] != '/')
+    while (cp[i] && !strchr("/><", cp[i]))
     {
         const char *attr_name = cp + i;
         int attr_len;
@@ -162,8 +166,12 @@ int mp::HTMLParser::Rep::tagAttrs(HTMLParserEvent &event,
         x[0] = tr;
         x[1] = 0;
         if (m_verbose)
-            printf ("------ attr %.*s=%.*s\n", attr_len, attr_name,
-                    val_len, value);
+        {
+            printf("------ attr %.*s", attr_len, attr_name);
+            if (value)
+                printf("=%.*s", val_len, value);
+            printf("\n");
+        }
         event.attribute(name, len, attr_name, attr_len, value, val_len, x);
     }
     return i;
@@ -174,7 +182,7 @@ int mp::HTMLParser::Rep::tagEnd(HTMLParserEvent &event,
 {
     int i = 0;
     int close_it = 0;
-    for (; cp[i] && cp[i] != '/' && cp[i] != '>'; i++)
+    for (; cp[i] && !strchr("/><", cp[i]); i++)
         ;
     if (i > 0)
     {
@@ -191,7 +199,7 @@ int mp::HTMLParser::Rep::tagEnd(HTMLParserEvent &event,
     {
         if (m_verbose)
             printf("------ any tag %s %.*s\n",
-                   close_it ? " close" : "end", tag_len, tag);
+                   close_it ? "close" : "end", tag_len, tag);
         event.anyTagEnd(tag, tag_len, close_it);
         i++;
     }
@@ -218,20 +226,33 @@ void mp::HTMLParser::Rep::parse_str(HTMLParserEvent &event, const char *cp)
         if (*cp++ != '<')
             continue;
 
-        if (*cp == '!')
+        if (nest && *cp == '!')
         {
             int i;
             tagText(event, text_start, cp - 1);
-            for (i = 1; cp[i] && cp[i] != '>'; i++)
-                ;
-            event.openTagStart(cp, i);
+            if (cp[1] == '-' && cp[2] == '-')
+            {
+                for (i = 3; cp[i]; i++)
+                    if (cp[i] == '-' && cp[i+1] == '-' && cp[i+2] == '>')
+                    {
+                        i+= 2;
+                        event.openTagStart(cp, i);
+                        break;
+                    }
+            }
+            else
+            {
+                for (i = 1; cp[i] && cp[i] != '>'; i++)
+                    ;
+                event.openTagStart(cp, i);
+            }
             if (m_verbose)
                 printf("------ dtd %.*s\n", i, cp);
             i += tagEnd(event, cp, i, cp + i);
             cp += i;
             text_start = cp;
         }
-        else if (*cp == '?')
+        else if (nest && *cp == '?')
         {
             int i;
             tagText(event, text_start, cp - 1);
@@ -247,9 +268,21 @@ void mp::HTMLParser::Rep::parse_str(HTMLParserEvent &event, const char *cp)
         else if (*cp == '/' && isAlpha(cp[1]))
         {
             int i;
-            tagText(event, text_start, cp - 1);
-
+ 
             i = skipName(++cp);
+
+            if (!nest)
+            {
+                if (i == 6 && !yaz_strncasecmp(cp, "script", i))
+                {
+                    int ws = skipSpace(cp + 6);
+                    if (cp[ws + 6] == '>')
+                        nest = true; /* really terminated */
+                }
+                if (!nest)
+                    continue;
+            }
+            tagText(event, text_start, cp - 2);
             event.closeTag(cp, i);
             if (m_verbose)
                 printf("------ tag close %.*s\n", i, cp);
@@ -257,7 +290,7 @@ void mp::HTMLParser::Rep::parse_str(HTMLParserEvent &event, const char *cp)
             cp += i;
             text_start = cp;
         }
-        else if (isAlpha(*cp))
+        else if (nest && isAlpha(*cp))
         {
             int i, j;
             tagText(event, text_start, cp - 1);
@@ -267,6 +300,10 @@ void mp::HTMLParser::Rep::parse_str(HTMLParserEvent &event, const char *cp)
                 printf("------ tag open %.*s\n", i, cp);
             j = tagAttrs(event, cp, i, cp + i);
             j += tagEnd(event, cp, i, cp + i + j);
+
+            if (i == 6 && !yaz_strncasecmp(cp, "script", i))
+                nest = false;
+
             cp += i + j;
             text_start = cp;
         }
@@ -274,6 +311,10 @@ void mp::HTMLParser::Rep::parse_str(HTMLParserEvent &event, const char *cp)
     tagText(event, text_start, cp);
 }
 
+mp::HTMLParserEvent::~HTMLParserEvent()
+{
+}
+
 /*
  * Local variables:
  * c-basic-offset: 4