From e1742e66a6b0f09eab51356821ca6308e5ebc68b Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Julian=20B=C3=A4ume?= <julian@svg4all.de>
Date: Mon, 23 Jun 2014 18:24:02 +0200
Subject: [PATCH] properly escape all unprintable characters

writing messages should no be in line with gettext tools. I tested
using msgcat, it provides the same results.

For some common use-cases I wrote explicit tests, for uncommon and
even unwanted use-cases I wrote one test to make sure pofile works
like msgcat for those messages
---
 lib/po.js                  | 24 +++++++++++++++++++++--
 test/fixtures/c-strings.po | 40 ++++++++++++++++++++++++++++++++++++++
 test/parse.js              | 18 +++++++++++++----
 3 files changed, 76 insertions(+), 6 deletions(-)

diff --git a/lib/po.js b/lib/po.js
index a337063..b7a6866 100644
--- a/lib/po.js
+++ b/lib/po.js
@@ -236,8 +236,27 @@ PO.Item.prototype.toString = function () {
 
     // reverse what extract(string) method during PO.parse does
     var _escape = function (string) {
-        string = string.replace(/\\/g, '\\\\');
-        return string.replace(/"/g, '\\"');
+        // don't unescape \n, since string can never contain it
+        // since split('\n') is called on it
+        string = string.replace(/[\x07\b\t\v\f\r"\\]/g, function (match) {
+            switch (match) {
+                case '\x07':
+                    return '\\a';
+                case '\b':
+                    return '\\b';
+                case '\t':
+                    return '\\t';
+                case '\v':
+                    return '\\v';
+                case '\f':
+                    return '\\f';
+                case '\r':
+                    return '\\r';
+                default:
+                    return '\\' + match;
+            }
+        });
+        return string;
     };
 
     var _process = function (keyword, text, i) {
@@ -286,6 +305,7 @@ PO.Item.prototype.toString = function () {
             } else {
                 text = isArray(text) ? text.join() : text;
                 var processed = _process(keyword, text);
+                //handle \n in single-line texts (can not be handled in _escape)
                 for (var i = 1; i < processed.length - 1; i++) {
                     processed[i] = processed[i].slice(0, -1) + '\\n"';
                 }
diff --git a/test/fixtures/c-strings.po b/test/fixtures/c-strings.po
index 597951e..d05b8e3 100644
--- a/test/fixtures/c-strings.po
+++ b/test/fixtures/c-strings.po
@@ -17,3 +17,43 @@ msgstr ""
 
 msgid "The name field must not contain characters like \" or \\"
 msgstr ""
+
+# possibility to reorder items depending on locale
+#. Format of addresses
+#. %1$s is the street
+#. %2$s is the postal code
+#. %3$s is the city
+#. %4$s is the state
+#. %5$s is the country
+msgid ""
+"%1$s\n"
+"%2$s %3$s\n"
+"%4$s\n"
+"%5$s"
+msgstr ""
+
+# "i18"ned code
+#. used in <pre> environment, so don't remove any control sequences
+msgid ""
+"define('some/test/module', function () {\n"
+"\t'use strict';\n"
+"\treturn {};\n"
+"});\n"
+""
+msgstr ""
+"define('random/test/file', function () {\n"
+"\t'use strict';\n"
+"\treturn {};\n"
+"});\n"
+""
+
+# all one-letter escape characters
+# be aware, that \a, \b, \v, \f and \r should not be used
+# in i18ned messages (according to gettext tools)
+# however, they should be properly parsed, anyway
+msgid ""
+"\a\b\t\n"
+"\v\f\r"
+msgstr ""
+"\a\b\t\n"
+"\v\f\r"
diff --git a/test/parse.js b/test/parse.js
index d12ee10..f17c7af 100644
--- a/test/parse.js
+++ b/test/parse.js
@@ -133,19 +133,29 @@ describe('Parse', function () {
     });
 
     describe('C-Strings', function () {
+        var po = PO.parse(fs.readFileSync(__dirname + '/fixtures/c-strings.po', 'utf8'));
         it('should parse the c-strings.po file', function () {
-            var po = PO.parse(fs.readFileSync(__dirname + '/fixtures/c-strings.po', 'utf8'));
-
             assert.notEqual(po, null);
         });
 
         it('should extract strings containing " and \\ characters', function () {
-            var po = PO.parse(fs.readFileSync(__dirname + '/fixtures/c-strings.po', 'utf8'));
-
             var items = po.items.filter(function (item) {
                 return (/^The name field must not contain/).test(item.msgid);
             });
             assert.equal(items[0].msgid, 'The name field must not contain characters like " or \\');
         });
+
+        it('should handle \n characters', function () {
+            var item = po.items[1];
+            assert.equal(item.msgid, '%1$s\n%2$s %3$s\n%4$s\n%5$s');
+        });
+
+        it('should handle \t characters', function () {
+            var item = po.items[2];
+            assert.equal(item.msgid, 'define(\'some/test/module\', function () {\n' +
+                '\t\'use strict\';\n' +
+                '\treturn {};\n' +
+                '});\n');
+        });
     });
 });