Removes HTML from quoted messages and previews. Closes #4664 Fixes #4404

src/client/wscript_build

@@ -22,8 +22,8 @@ client_src = [
 'ui/message-viewer.vala',
 'ui/preferences-dialog.vala',
 
+'util/util-email.vala',
 'util/util-keyring.vala',
-'util/util-email.vala'
 ]
 
 gsettings_schemas = [

src/engine/rfc822/rfc822-message-data.vala

@@ -240,6 +240,7 @@ public class Geary.RFC822.PreviewText : Geary.RFC822.Text {
         if (preview_header != null) {
             string? charset = null;
             string? encoding = null;
+            bool is_html = false;
             
             // Parse the header.
             GMime.Stream header_stream = new GMime.StreamMem.with_buffer(
@@ -247,6 +248,8 @@ public class Geary.RFC822.PreviewText : Geary.RFC822.Text {
             GMime.Parser parser = new GMime.Parser.with_stream(header_stream);
             GMime.Part? part = parser.construct_part() as GMime.Part;
             if (part != null) {
+                is_html = (part.get_content_type().to_string() == "text/html");
+                
                 charset = part.get_content_type_parameter("charset");
                 encoding = part.get_header("Content-Transfer-Encoding");
             }
@@ -268,7 +271,8 @@ public class Geary.RFC822.PreviewText : Geary.RFC822.Text {
             input_stream.write_to_stream(filter);
             uint8[] data = output.data;
             data += (uint8) '\0';
-            buffer = new Geary.Memory.StringBuffer((string) data);
+            buffer = new Geary.Memory.StringBuffer(is_html ? Geary.HTML.remove_html_tags(
+                (string) data) : (string) data);
         }
         
         base (buffer);

src/engine/rfc822/rfc822-utils.vala

@@ -60,18 +60,25 @@ public string quote_email_for_forward(Geary.Email email) {
 }
 
 private string quote_body(Geary.Email email, bool line_start_char = true) {
-    string ret = "";
+    string body_text = "";
     try {
-        string[] lines = email.get_message().get_first_mime_part_of_content_type("text/plain")
-            .to_utf8().split("\n");
-        for (int i = 0; i < lines.length; i++) {
-            if (line_start_char)
-                ret += "> ";
-            
-            ret += lines[i];
-        }
+        body_text = email.get_message().get_first_mime_part_of_content_type("text/plain").to_utf8();
     } catch (Error err) {
-        debug("Could not get message text. %s", err.message);
+        try {
+            body_text = Geary.HTML.remove_html_tags(email.get_message().
+                get_first_mime_part_of_content_type("text/html").to_utf8());
+        } catch (Error err2) {
+            debug("Could not get message text. %s", err2.message);
+        }
+    }
+    
+    string ret = "";
+    string[] lines = body_text.split("\n");
+    for (int i = 0; i < lines.length; i++) {
+        if (line_start_char)
+            ret += "> ";
+        
+        ret += lines[i];
     }
     
     return ret;

src/engine/util/util-html.vala

@@ -0,0 +1,56 @@
+/* Copyright 2011-2012 Yorba Foundation
+ *
+ * This software is licensed under the GNU Lesser General Public License
+ * (version 2.1 or later).  See the COPYING file in this distribution. 
+ */
+
+namespace Geary.HTML {
+
+// Removes any text between < and >.  Additionally, if input terminates in the middle of a tag, 
+// the tag will be removed.
+// If the HTML is invalid, the original string will be returned.
+public string remove_html_tags(string input) {
+    try {
+        string output = input;
+        
+        // Count the number of < and > characters.
+        unichar c;
+        uint64 less_than = 0;
+        uint64 greater_than = 0;
+        for (int i = 0; output.get_next_char (ref i, out c);) {
+            if (c == '<')
+                less_than++;
+            else if (c == '>')
+                greater_than++;
+        }
+        
+        if (less_than == greater_than + 1) {
+            output += ">"; // Append an extra > so our regex works.
+            greater_than++;
+        }
+        
+        if (less_than != greater_than)
+            return input; // Invalid HTML.
+        
+        // Removes script tags and everything between them.
+        // Based on regex here: http://stackoverflow.com/questions/116403/im-looking-for-a-regular-expression-to-remove-a-given-xhtml-tag-from-a-string
+        Regex script = new Regex("<script[^>]*?>[\\s\\S]*?<\\/script>", RegexCompileFlags.CASELESS);
+        output = script.replace(output, -1, 0, "");
+        
+        // Removes style tags and everything between them.
+        // Based on regex above.
+        Regex style = new Regex("<style[^>]*?>[\\s\\S]*?<\\/style>", RegexCompileFlags.CASELESS);
+        output = style.replace(output, -1, 0, "");
+        
+        // Removes remaining tags. Based on this regex:
+        // http://osherove.com/blog/2003/5/13/strip-html-tags-from-a-string-using-regular-expressions.html
+        Regex tags = new Regex("<(.|\n)*?>", RegexCompileFlags.CASELESS);
+        return tags.replace(output, -1, 0, "");
+    } catch (Error e) {
+        debug("Error stripping HTML tags: %s", e.message);
+    }
+    
+    return input;
+}
+
+}

src/wscript

@@ -147,6 +147,7 @@ def build(bld):
 	'../engine/state/state-machine.vala',
 	'../engine/state/state-mapping.vala',
 	
+	'../engine/util/util-html.vala',
 	'../engine/util/util-inet.vala',
 	'../engine/util/util-interfaces.vala',
 	'../engine/util/util-memory.vala',