[tor-commits] [onionoo/master] Provide utf-characters instead of escape strings.

[karsten at torproject.org]

commit 947ce706bb46c5d7e5c4cd2b0716a2458b4af86b
Author: iwakeh <iwakeh at torproject.org>
Date:   Fri Apr 6 09:53:35 2018 +0000

    Provide utf-characters instead of escape strings.
    
    Also adapt tests to check the new functionality.
    Implements task-25740.
---
 CHANGELOG.md                                       |  6 +++++
 .../torproject/onionoo/util/FormattingUtils.java   |  6 +++--
 src/test/resources/lines-for-escape-tests.txt      | 26 +++++++++++++---------
 3 files changed, 26 insertions(+), 12 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index b903acc..692a417 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -1,3 +1,9 @@
+# Changes in version 6.0-1.13.0 - 2018-04-??
+
+ * Minor changes
+   - Turn valid utf-8 escape sequences into utf-8 characters.
+
+
 # Changes in version 5.2-1.12.0 - 2018-04-06
 
  * Medium changes
diff --git a/src/main/java/org/torproject/onionoo/util/FormattingUtils.java b/src/main/java/org/torproject/onionoo/util/FormattingUtils.java
index 3d16f5a..558477d 100644
--- a/src/main/java/org/torproject/onionoo/util/FormattingUtils.java
+++ b/src/main/java/org/torproject/onionoo/util/FormattingUtils.java
@@ -3,6 +3,8 @@
 
 package org.torproject.onionoo.util;
 
+import static org.apache.commons.lang3.StringEscapeUtils.unescapeJava;
+
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
@@ -58,8 +60,8 @@ public class FormattingUtils {
       StringBuffer sb = new StringBuffer();
       Matcher mat = escapePattern.matcher(text);
       while (mat.find()) {
-        String unescaped = mat.group(1);
-        mat.appendReplacement(sb, unescaped);
+        String unescaped = mat.group(1).substring(1);
+        mat.appendReplacement(sb, unescapeJava(unescaped));
       }
       mat.appendTail(sb);
       return sb.toString();
diff --git a/src/test/resources/lines-for-escape-tests.txt b/src/test/resources/lines-for-escape-tests.txt
index 4fb5895..b08b723 100644
--- a/src/test/resources/lines-for-escape-tests.txt
+++ b/src/test/resources/lines-for-escape-tests.txt
@@ -2,15 +2,21 @@
 
 abc
 abc
-\\\\u
-\\\\u
+\\\\u00bb \\\\u \\\\u00ab
+» \\\\u «
 Haha/\\\\@/\\\\live/\\\\./\\\\co/\\\\./\\\uk
 Haha/\\\\@/\\\\live/\\\\./\\\\co/\\\\./\\\uk
-\\\\u20ac
-\\u20ac
-\\\\u0024
-\\u0024
-some \\\\u20ac other string \\\\u0024 to unescape
-some \\u20ac other string \\u0024 to unescape
-abcd efg\\\\u0024xyz\\\\uxxxx
-abcd efg\\u0024xyz\\\\uxxxx
+Euro \\\\u20ac
+Euro €
+Dollar \\\\u0024
+Dollar $
+Pound \\\\u00a3
+Pound £
+Yen \\\\u00a5
+Yen ¥
+\\\\u00a1some \\\\u00dE other string \\\\u00e7 to unescape!
+¡some Þ other string ç to unescape!
+abcd efg cent \\\\u00a2xyz\\\\uxxxx
+abcd efg cent ¢xyz\\\\uxxxx
+Telef\\\\u00F4nica
+Telefônica