Issue 5477 - Import everything from imported locales

packagerChrome.py

 # This Source Code Form is subject to the terms of the Mozilla Public
 # License, v. 2.0. If a copy of the MPL was not distributed with this
 # file, You can obtain one at http://mozilla.org/MPL/2.0/.
 
 import errno
 import io
 import json
 import os
 import re
 from StringIO import StringIO
@@ skipping 172 matching lines @@
         ).encode('utf-8')
 
 
 def toJson(data):
     return json.dumps(
         data, ensure_ascii=False, sort_keys=True,
         indent=2, separators=(',', ': ')
     ).encode('utf-8') + '\n'
 
 
-def importGeckoLocales(params, files):
+def set_translation_and_additional_info(data, key, source, stringID):
+    """Sets data[key] to the desired {'message':value}, adds all additional
+    info from the source dict"""

[Sebastian Noack, 2017/08/18 14:53:59]

As per PEP-257, the trailing """ goes onto a new line in multi-line docstrings.

[tlucas, 2017/08/18 15:10:35]

Done.

+    value = source[stringID]['message']

[Sebastian Noack, 2017/08/18 14:53:59]

It seems all usage of "source" and "stringID" here and in the other function are
in the form of "source[stringID]". So how about passing "source[stringID]"
directly to these functions rather than passing these arguments seperately and
duplicating the lookup?

[tlucas, 2017/08/18 15:10:34]

Done.

+
+    data[key] = {'messages': value}
+    for k, v in source[stringID].items():
+        data[key].setdefault(k, v)
+
+
+def set_translation_without_access_keys(data, key, source, stringID):

[Sebastian Noack, 2017/08/18 14:53:59]

I think better names for these functions would be import_string_webext() and
import_string_gecko().

[tlucas, 2017/08/18 15:10:35]

Done.

+    """only set {'message': value} in data-dictionary, after stripping
+    undesired gecko-style access keys"""
+    value = source[stringID]
+
+    # Remove access keys from possible gecko-style
+    # translations
+    match = re.search(r'^(.*?)\s*\(&.\)$', value)
+    if match:
+        value = match.group(1)
+    else:
+        index = value.find('&')
+        if index >= 0:
+            value = value[0:index] + value[index + 1:]
+
+    data[key] = {'message': value}
+
+
+def import_locales(params, files):
     import localeTools
 
     # FIXME: localeTools doesn't use real Chrome locales, it uses dash as
     # separator instead.
     convert_locale_code = lambda code: code.replace('-', '_')
 
     # We need to map Chrome locales to Gecko locales. Start by mapping Chrome
     # locales to themselves, merely with the dash as separator.
     locale_mapping = {convert_locale_code(l): l for l in localeTools.chromeLocales}
 
@@ skipping 16 matching lines @@
             fileName, keys = item
             parts = map(lambda n: source if n == '*' else n, fileName.split('/'))
             sourceFile = os.path.join(os.path.dirname(item.source), *parts)
             incompleteMarker = os.path.join(os.path.dirname(sourceFile), '.incomplete')
             if not os.path.exists(sourceFile) or os.path.exists(incompleteMarker):
                 continue
 
             data = json.loads(files[targetFile].decode('utf-8'))
 
             try:
+                # .json and other formats provide translations differently and

[Sebastian Noack, 2017/08/18 14:53:59]

The other format is the Gecko format.

  "The WebExtensions (.json) and Gecko format provide ..."

[tlucas, 2017/08/18 15:10:35]

Done.

+                # / or provide additional information like e.g. "placeholders".

[Sebastian Noack, 2017/08/18 14:53:59]

It is supposed to be "and/or" without space.

[tlucas, 2017/08/18 15:10:34]

Done.

+                # We want to adhere to that / preserve the addtional info
                 if sourceFile.endswith('.json'):
                     with io.open(sourceFile, 'r', encoding='utf-8') as handle:
-                        sourceData = {k: v['message'] for k, v in json.load(handle).iteritems()}
+                        sourceData = json.load(handle)
+
+                    set_translation = set_translation_and_additional_info
                 else:
                     sourceData = localeTools.readFile(sourceFile)
 
+                    set_translation = set_translation_without_access_keys
+
                 # Resolve wildcard imports
                 if keys == '*' or keys == '=*':
                     importList = sourceData.keys()
                     importList = filter(lambda k: not k.startswith('_'), importList)
                     if keys == '=*':
                         importList = map(lambda k: '=' + k, importList)
                     keys = ' '.join(importList)
 
                 for stringID in keys.split():
                     noMangling = False
                     if stringID.startswith('='):
                         stringID = stringID[1:]
                         noMangling = True
 
                     if stringID in sourceData:
                         if noMangling:
                             key = re.sub(r'\W', '_', stringID)
                         else:
                             key = re.sub(r'\..*', '', parts[-1]) + '_' + re.sub(r'\W', '_', stringID)
                         if key in data:
                             print 'Warning: locale string %s defined multiple times' % key
 
-                        # Remove access keys
-                        value = sourceData[stringID]
-                        match = re.search(r'^(.*?)\s*\(&.\)$', value)
-                        if match:
-                            value = match.group(1)
-                        else:
-                            index = value.find('&')
-                            if index >= 0:
-                                value = value[0:index] + value[index + 1:]
-                        data[key] = {'message': value}
+                        set_translation(data, key, sourceData, stringID)
             except Exception as e:
                 print 'Warning: error importing locale data from %s: %s' % (sourceFile, e)
 
             files[targetFile] = toJson(data)
 
 
 def truncate(text, length_limit):
     if len(text) <= length_limit:
         return text
     return text[:length_limit - 1].rstrip() + u'\u2026'
@@ skipping 95 matching lines @@
     if metadata.has_section('convert_js'):
         convertJS(params, files)
 
     if metadata.has_section('preprocess'):
         files.preprocess(
             [f for f, _ in metadata.items('preprocess')],
             {'needsExt': True}
         )
 
     if metadata.has_section('import_locales'):
-        importGeckoLocales(params, files)
+        import_locales(params, files)
 
     files['manifest.json'] = createManifest(params, files)
     if type == 'chrome':
         fixTranslationsForCWS(files)
 
     if devenv:
         import buildtools
         import random
         files.read(os.path.join(buildtools.__path__[0], 'chromeDevenvPoller__.js'), relpath='devenvPoller__.js')
         files['devenvVersion__'] = str(random.random())
 
         if metadata.has_option('general', 'testScripts'):
             files['qunit/index.html'] = createScriptPage(
                 params, 'testIndex.html.tmpl', ('general', 'testScripts')
             )
 
     zipdata = files.zipToString()
     signature = None
     pubkey = None
     if keyFile != None:
         signature = signBinary(zipdata, keyFile)
         pubkey = getPublicKey(keyFile)
     writePackage(outFile, pubkey, signature, zipdata)