Fix bug in diff view for filenames with non-ASCII characters (issue 128)

James Moger

2012-09-10 b34048803ad6cf0a0a0c998696a41de118715452

Fix bug in diff view for filenames with non-ASCII characters (issue-128)

 docs/03_faq.mkd

@@ -71,6 +71,13 @@
2. *web.mountParameters = false* and use non-pretty, parameterized urls

3. *web.forwardSlashCharacter = !* which tells Gitblit to use **!** instead of **/**



#### UTF-8 Filenames



Tomcat also dislikes urls with non-ASCII characters. If your repositories have non-ASCII filenames you will have to modify your connector properties to allow UTF-8 encoded urls.  


[Tomcat Character Encoding](http://wiki.apache.org/tomcat/FAQ/CharacterEncoding)  
[Tomcat Connector Properties](http://tomcat.apache.org/tomcat-6.0-doc/config/http.html)



## General Interest Questions



### Gitblit?  What kind of name is that?


 docs/04_releases.mkd

@@ -9,6 +9,14 @@


**%VERSION%** ([go](http://code.google.com/p/gitblit/downloads/detail?name=%GO%) | [war](http://code.google.com/p/gitblit/downloads/detail?name=%WAR%) | [express](http://code.google.com/p/gitblit/downloads/detail?name=%EXPRESS%) | [fedclient](http://code.google.com/p/gitblit/downloads/detail?name=%FEDCLIENT%) | [manager](http://code.google.com/p/gitblit/downloads/detail?name=%MANAGER%) | [api](http://code.google.com/p/gitblit/downloads/detail?name=%API%)) based on [%JGIT%][jgit]   *released %BUILDDATE%*



#### fixes



- Fixed bug in the diff view for filenames that have non-ASCII characters (issue 128)



#### additions



- added RedmineUserService (github/mallowlabs)



#### changes



- Emit a warning in the log file if running on a Tomcat-based servlet container which is unfriendly to %2F forward-slash url encoding AND Gitblit is configured to mount parameters with %2F forward-slash url encoding (Github/jpyeron, issue 126)


 src/com/gitblit/utils/GitBlitDiffFormatter.java

@@ -127,15 +127,24 @@
            } else if (line.startsWith("---") || line.startsWith("+++")) {

                // skip --- +++ lines

            } else if (line.startsWith("diff")) {

                line = StringUtils.convertOctal(line);

                if (line.indexOf(oldnull) > -1) {

                    // a is null, use b

                    line = line.substring(("diff --git " + oldnull).length()).trim();

                    // trim b/

                    line = line.substring(2);

                    line = line.substring(2).trim();

                } else {

                    // use a

                    line = line.substring("diff --git a/".length()).trim();

                    line = line.substring(0, line.indexOf(" b/")).trim();

                    line = line.substring("diff --git ".length()).trim();

                    line = line.substring(line.startsWith("\"a/") ? 3 : 2);					
                    line = line.substring(0, line.indexOf(" b/") > -1 ? line.indexOf(" b/") : line.indexOf("\"b/")).trim();

                }

				
                if (line.charAt(0) == '"') {

                    line = line.substring(1);

                }

                if (line.charAt(line.length() - 1) == '"') {

                    line = line.substring(0, line.length() - 1);

                }

                if (inFile) {

                    sb.append("</tbody></table></div>\n");


 src/com/gitblit/utils/GitWebDiffFormatter.java

@@ -106,10 +106,10 @@
            throws IOException {

        switch (prefix) {

        case '+':

            os.write("<span class=\"diff add\">".getBytes());

            os.write("<span style=\"color:#008000;\">".getBytes());

            break;

        case '-':

            os.write("<span class=\"diff remove\">".getBytes());

            os.write("<span style=\"color:#800000;\">".getBytes());

            break;

        }

        os.write(prefix);

@@ -140,11 +140,11 @@
        sb.append("<div class=\"diff\">");

        for (String line : lines) {

            if (line.startsWith("diff")) {

                sb.append("<div class=\"diff header\">").append(line).append("</div>");

                sb.append("<div class=\"diff header\">").append(StringUtils.convertOctal(line)).append("</div>");

            } else if (line.startsWith("---")) {

                sb.append("<span class=\"diff remove\">").append(line).append("</span><br/>");

                sb.append("<span style=\"color:#800000;\">").append(StringUtils.convertOctal(line)).append("</span><br/>");

            } else if (line.startsWith("+++")) {

                sb.append("<span class=\"diff add\">").append(line).append("</span><br/>");

                sb.append("<span style=\"color:#008000;\">").append(StringUtils.convertOctal(line)).append("</span><br/>");

            } else {

                sb.append(line).append('\n');

            }


 src/com/gitblit/utils/StringUtils.java

@@ -15,6 +15,7 @@
 */

package com.gitblit.utils;



import java.io.ByteArrayOutputStream;

import java.io.UnsupportedEncodingException;

import java.nio.ByteBuffer;

import java.nio.CharBuffer;

@@ -624,4 +625,39 @@
        }

        return url;

    }

	
    /**

     * Converts a string with \nnn sequences into a UTF-8 encoded string.

     * @param input

     * @return

     */

    public static String convertOctal(String input) {

        try {

            ByteArrayOutputStream bytes = new ByteArrayOutputStream();

            Pattern p = Pattern.compile("(\\\\\\d{3})");

            Matcher m = p.matcher(input);

            int i = 0;

            while (m.find()) {

                bytes.write(input.substring(i, m.start()).getBytes("UTF-8"));

                // replace octal encoded value

                // strip leading \ character

                String oct = m.group().substring(1);

                bytes.write(Integer.parseInt(oct, 8));

                i = m.end();			
            }

            if (bytes.size() == 0) {

                // no octal matches

                return input;

            } else {

                if (i < input.length()) {

                    // add remainder of string

                    bytes.write(input.substring(i).getBytes("UTF-8"));

                }

            }

            return bytes.toString("UTF-8");

        } catch (Exception e) {

            e.printStackTrace();

        }

        return input;

    }

}