[FIX] PO translations export

Within the scope of P3 compatibility, a StringIO was replaced by
BytesIO in b4429c2a91, however I missed
that BytesIO is much stricter about types than StringIO *even in
Python 2*. Some bits of text would sneak up in PoFile.write or
PoFile.write_info and blow up the export.

The BytesIO makes sense for output (tgz is definitely not a text
format and base64 is a bytes -> bytes encoding), but ensuring the
entirety of PoFile worked on bytes seemed distasteful (and lots of
changes), it seemed more sensible at all levels that PoFile would make
sure to properly work on *text*, and to add a text adapter wrapper
around the BytesIO buffer.

io.TextIOWrapper was almost perfect (it closes the underlying stream
on close but that can be worked around) *however* it turns out
TextIOWrapper does not work on Python 2's `file` objects... so use the
older codecs.StreamReaderWriter which should be able to do the job
properly.

fixes #17834

odoo/tools/translate.py

@@ -1,10 +1,10 @@
 # -*- coding: utf-8 -*-
 # Part of Odoo. See LICENSE file for full copyright and licensing details.
-
 import codecs
 import csv
 import fnmatch
 import inspect
+import io
 import locale
 import logging
 import os
@@ -474,7 +474,13 @@ def unquote(str):
 # class to handle po files
 class PoFile(object):
     def __init__(self, buffer):
-        self.buffer = buffer
+        # TextIOWrapper closes its underlying buffer on close *and* can't
+        # handle actual file objects (on python 2)
+        self.buffer = codecs.StreamReaderWriter(
+            stream=buffer,
+            Reader=codecs.getreader('utf-8'),
+            Writer=codecs.getwriter('utf-8')
+        )
 
     def __iter__(self):
         self.buffer.seek(0)
@@ -489,7 +495,7 @@ class PoFile(object):
         lines = self.buffer.readlines()
         # remove the BOM (Byte Order Mark):
         if len(lines):
-            lines[0] = unicode(lines[0], 'utf8').lstrip(unicode( codecs.BOM_UTF8, "utf8"))
+            lines[0] = lines[0].lstrip(u"\ufeff")
 
         lines.append('') # ensure that the file ends with at least an empty line
         return lines
@@ -594,7 +600,7 @@ class PoFile(object):
 
     def write_infos(self, modules):
         import odoo.release as release
-        self.buffer.write("# Translation of %(project)s.\n" \
+        self.buffer.write(u"# Translation of %(project)s.\n" \
                           "# This file contains the translation of the following modules:\n" \
                           "%(modules)s" \
                           "#\n" \
@@ -622,30 +628,30 @@ class PoFile(object):
     def write(self, modules, tnrs, source, trad, comments=None):
 
         plurial = len(modules) > 1 and 's' or ''
-        self.buffer.write("#. module%s: %s\n" % (plurial, ', '.join(modules)))
+        self.buffer.write(u"#. module%s: %s\n" % (plurial, ', '.join(modules)))
 
         if comments:
-            self.buffer.write(''.join(('#. %s\n' % c for c in comments)))
+            self.buffer.write(u''.join(('#. %s\n' % c for c in comments)))
 
         code = False
         for typy, name, res_id in tnrs:
-            self.buffer.write("#: %s:%s:%s\n" % (typy, name, res_id))
+            self.buffer.write(u"#: %s:%s:%s\n" % (typy, name, res_id))
             if typy == 'code':
                 code = True
 
         if code:
             # only strings in python code are python formated
-            self.buffer.write("#, python-format\n")
+            self.buffer.write(u"#, python-format\n")
 
         if not isinstance(trad, unicode):
             trad = unicode(trad, 'utf8')
         if not isinstance(source, unicode):
             source = unicode(source, 'utf8')
 
-        msg = "msgid %s\n"      \
-              "msgstr %s\n\n"   \
+        msg = u"msgid %s\n"      \
+              u"msgstr %s\n\n"   \
                   % (quote(source), quote(trad))
-        self.buffer.write(msg.encode('utf8'))
+        self.buffer.write(msg)
 
 
 # Methods to export the translation file