[FIX] P3: text model in various mail bits

No formal decision there, and the stdlib API seems like a PITA as it's
based around native strings, not bytes and not text.

* try to use cross-model API instead of random encode/decode
* unicode all the mail_template things
* formataddr *requires* a text tuple now

  In P2, formataddr((False, False)) "worked" (generated "False
  <False>"), in P3 it will fail and blow up.
* attachments are quite definitely bytes
* remove try_coerce_ascii because it's garbage

8 files changed, 112 insertions, 120 deletions

diff --git a/addons/mail/models/mail_mail.py b/addons/mail/models/mail_mail.py
index 3f206424e07..e53c24fb176 100644
--- a/addons/mail/models/mail_mail.py
+++ b/addons/mail/models/mail_mail.py
@@ -169,7 +169,7 @@ class MailMail(models.Model):
           - else fallback on mail.email_to splitting """
         self.ensure_one()
         if partner:
-            email_to = [formataddr((partner.name, partner.email))]
+            email_to = [formataddr((partner.name or 'False', partner.email or 'False'))]
         else:
             email_to = tools.email_split_and_format(self.email_to)
         return email_to
diff --git a/addons/mail/models/mail_template.py b/addons/mail/models/mail_template.py
index e9bf605c382..f44ee7be09a 100644
--- a/addons/mail/models/mail_template.py
+++ b/addons/mail/models/mail_template.py
@@ -306,8 +306,8 @@ class MailTemplate(models.Model):
         # form a tree
         root = lxml.html.fromstring(html)
         if not len(root) and root.text is None and root.tail is None:
-            html = '<div>%s</div>' % html
-            root = lxml.html.fromstring(html)
+            html = u'<div>%s</div>' % html
+            root = lxml.html.fromstring(html, encoding='unicode')
 
         base_url = self.env['ir.config_parameter'].sudo().get_param('web.base.url')
         base = urls.url_parse(base_url)
@@ -324,12 +324,12 @@ class MailTemplate(models.Model):
         for node in root.iter():
             if node.tag == 'a' and node.get('href'):
                 node.set('href', _process_link(node.get('href')))
-            elif node.tag == 'img' and not node.get('src', 'data').startswith('data'):
+            elif node.tag == 'img' and not node.get('src', 'data').startswith(u'data'):
                 node.set('src', _process_link(node.get('src')))
 
-        html = lxml.html.tostring(root, pretty_print=False, method='html')
+        html = lxml.html.tostring(root, pretty_print=False, method='html', encoding='unicode')
         # this is ugly, but lxml/etree tostring want to put everything in a 'div' that breaks the editor -> remove that
-        if html.startswith('<div>') and html.endswith('</div>'):
+        if html.startswith(u'<div>') and html.endswith(u'</div>'):
             html = html[5:-6]
         return html
 
diff --git a/addons/mail/tests/test_mail_features.py b/addons/mail/tests/test_mail_features.py
index a6fce59c10a..8e2264ad6d3 100644
--- a/addons/mail/tests/test_mail_features.py
+++ b/addons/mail/tests/test_mail_features.py
@@ -131,7 +131,7 @@ class TestMessagePost(TestMail):
         self.assertEqual(set(msg.attachment_ids.mapped('res_id')), set([self.test_pigs.id]),
                          'message_post: all atttachments should be linked to the pigs group')
         self.assertEqual(set([base64.b64decode(x) for x in msg.attachment_ids.mapped('datas')]),
-                         set(['migration test', _attachments[0][1], _attachments[1][1]]))
+                         set([b'migration test', _attachments[0][1], _attachments[1][1]]))
         self.assertTrue(set([_attach_1.id, _attach_2.id]).issubset(msg.attachment_ids.ids),
                         'message_post: mail.message attachments duplicated')
         # notifications
diff --git a/odoo/addons/base/ir/ir_mail_server.py b/odoo/addons/base/ir/ir_mail_server.py
index 91cf0e61e73..eab160b4a84 100644
--- a/odoo/addons/base/ir/ir_mail_server.py
+++ b/odoo/addons/base/ir/ir_mail_server.py
@@ -39,20 +39,8 @@ class WriteToLogger(object):
         _logger.debug(s)
 smtplib.stderr = WriteToLogger()
 
-def try_coerce_ascii(string_utf8):
-    """Attempts to decode the given utf8-encoded string
-       as ASCII after coercing it to UTF-8, then return
-       the confirmed 7-bit ASCII string.
-
-       If the process fails (because the string
-       contains non-ASCII characters) returns ``None``.
-    """
-    try:
-        string_utf8.decode('ascii')
-    except UnicodeDecodeError:
-        return
-    return string_utf8
-
+def is_ascii(s):
+    return all(ord(cp) < 128 for cp in s)
 
 def encode_header(header_text):
     """Returns an appropriate representation of the given header value,
@@ -70,16 +58,10 @@ def encode_header(header_text):
     """
     if not header_text:
         return ""
-    # convert anything to utf-8, suitable for testing ASCIIness, as 7-bit chars are
-    # encoded as ASCII in utf-8
-    header_text_utf8 = ustr(header_text).encode('utf-8')
-    header_text_ascii = try_coerce_ascii(header_text_utf8)
-    # if this header contains non-ASCII characters,
-    # we'll need to wrap it up in a message.header.Header
-    # that will take care of RFC2047-encoding it as
-    # 7-bit string.
-    return header_text_ascii or Header(header_text_utf8, 'utf-8')
-
+    header_text = ustr(header_text) # FIXME: require unicode higher up?
+    if is_ascii(header_text):
+        return pycompat.to_native(header_text)
+    Header(header_text, 'utf-8')
 
 def encode_header_param(param_text):
     """Returns an appropriate RFC2047 encoded representation of the given
@@ -97,10 +79,10 @@ def encode_header_param(param_text):
     # For details see the encode_header() method that uses the same logic
     if not param_text:
         return ""
-    param_text_utf8 = ustr(param_text).encode('utf-8')
-    param_text_ascii = try_coerce_ascii(param_text_utf8)
-    return param_text_ascii or Charset('utf8').header_encode(param_text_utf8)
-
+    param_text = ustr(param_text) # FIXME: require unicode higher up?
+    if is_ascii(param_text):
+        return pycompat.to_native(param_text) # TODO: is that actually necessary?
+    return Charset("utf-8").header_encode(param_text)
 
 address_pattern = re.compile(r'([^ ,<@]+@[^> ,]+)')
 
@@ -111,8 +93,8 @@ def extract_rfc2822_addresses(text):
     """
     if not text:
         return []
-    candidates = address_pattern.findall(ustr(text).encode('utf-8'))
-    return [c for c in candidates if try_coerce_ascii(c)]
+    candidates = address_pattern.findall(ustr(text))
+    return [c for c in candidates if is_ascii(c)]
 
 
 def encode_rfc2822_address_header(header_text):
@@ -124,11 +106,20 @@ def encode_rfc2822_address_header(header_text):
     """
     def encode_addr(addr):
         name, email = addr
-        if not try_coerce_ascii(name):
-            name = str(Header(name, 'utf-8'))
+        # If s is a <text string>, then charset is a hint specifying the
+        # character set of the characters in the string. The Unicode string
+        # will be encoded using the following charsets in order: us-ascii,
+        # the charset hint, utf-8. The first character set to not provoke a
+        # UnicodeError is used.
+        # -> always pass a text string to Header
+
+        # also Header.__str__ in Python 3 "Returns an approximation of the
+        # Header as a string, using an unlimited line length.", the old one
+        # was "A synonym for Header.encode()." so call encode() directly?
+        name = Header(pycompat.to_text(name)).encode()
         return formataddr((name, email))
 
-    addresses = getaddresses([ustr(header_text).encode('utf-8')])
+    addresses = getaddresses([pycompat.to_native(ustr(header_text))])
     return COMMASPACE.join(encode_addr(a) for a in addresses)
 
 
@@ -252,8 +243,8 @@ class IrMailServer(models.Model):
             # The user/password must be converted to bytestrings in order to be usable for
             # certain hashing schemes, like HMAC.
             # See also bug #597143 and python issue #5285
-            smtp_user = ustr(smtp_user).encode('utf-8')
-            smtp_password = ustr(smtp_password).encode('utf-8')
+            smtp_user = pycompat.to_native(ustr(smtp_user))
+            smtp_password = pycompat.to_native(ustr(smtp_password))
             connection.login(smtp_user, smtp_password)
         return connection
 
@@ -301,8 +292,8 @@ class IrMailServer(models.Model):
         email_bcc = email_bcc or []
         body = body or u''
 
-        email_body_utf8 = ustr(body).encode('utf-8')
-        email_text_part = MIMEText(email_body_utf8, _subtype=subtype, _charset='utf-8')
+        email_body = ustr(body)
+        email_text_part = MIMEText(email_body, _subtype=subtype, _charset='utf-8')
         msg = MIMEMultipart()
 
         if not message_id:
@@ -328,20 +319,20 @@ class IrMailServer(models.Model):
         msg['Date'] = formatdate()
         # Custom headers may override normal headers or provide additional ones
         for key, value in pycompat.items(headers):
-            msg[ustr(key).encode('utf-8')] = encode_header(value)
+            msg[pycompat.to_native(ustr(key))] = encode_header(value)
 
         if subtype == 'html' and not body_alternative:
             # Always provide alternative text body ourselves if possible.
-            text_utf8 = html2text.html2text(email_body_utf8.decode('utf-8')).encode('utf-8')
+            text = html2text.html2text(email_body_utf8)
             alternative_part = MIMEMultipart(_subtype="alternative")
-            alternative_part.attach(MIMEText(text_utf8, _charset='utf-8', _subtype='plain'))
+            alternative_part.attach(MIMEText(text, _charset='utf-8', _subtype='plain'))
             alternative_part.attach(email_text_part)
             msg.attach(alternative_part)
         elif body_alternative:
             # Include both alternatives, as specified, within a multipart/alternative part
             alternative_part = MIMEMultipart(_subtype="alternative")
-            body_alternative_utf8 = ustr(body_alternative).encode('utf-8')
-            alternative_body_part = MIMEText(body_alternative_utf8, _subtype=subtype_alternative, _charset='utf-8')
+            body_alternative_ = ustr(body_alternative)
+            alternative_body_part = MIMEText(body_alternative_, _subtype=subtype_alternative, _charset='utf-8')
             alternative_part.attach(alternative_body_part)
             alternative_part.attach(email_text_part)
             msg.attach(alternative_part)
diff --git a/odoo/addons/base/res/res_partner.py b/odoo/addons/base/res/res_partner.py
index 972b74d7f3e..32772fa19a6 100644
--- a/odoo/addons/base/res/res_partner.py
+++ b/odoo/addons/base/res/res_partner.py
@@ -337,7 +337,7 @@ class Partner(models.Model):
     @api.depends('name', 'email')
     def _compute_email_formatted(self):
         for partner in self:
-            partner.email_formatted = formataddr((partner.name, partner.email))
+            partner.email_formatted = formataddr((partner.name or u"False", partner.email or u"False"))
 
     @api.depends('is_company')
     def _compute_company_type(self):
diff --git a/odoo/addons/base/tests/test_mail.py b/odoo/addons/base/tests/test_mail.py
index 1a30ef4690b..84cf892ea30 100644
--- a/odoo/addons/base/tests/test_mail.py
+++ b/odoo/addons/base/tests/test_mail.py
@@ -171,14 +171,14 @@ class TestSanitizer(unittest.TestCase):
         for ext in test_mail_examples.QUOTE_BLOCKQUOTE_IN:
             self.assertIn(ext, html)
         for ext in test_mail_examples.QUOTE_BLOCKQUOTE_OUT:
-            self.assertIn('<span data-o-mail-quote="1">%s' % misc.html_escape(ext.decode('utf-8')), html)
+            self.assertIn(u'<span data-o-mail-quote="1">%s' % misc.html_escape(ext), html)
 
     def test_quote_thunderbird(self):
         html = html_sanitize(test_mail_examples.QUOTE_THUNDERBIRD_1)
         for ext in test_mail_examples.QUOTE_THUNDERBIRD_1_IN:
             self.assertIn(ext, html)
         for ext in test_mail_examples.QUOTE_THUNDERBIRD_1_OUT:
-            self.assertIn('<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext.decode('utf-8')), html)
+            self.assertIn(u'<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext), html)
 
     def test_quote_hotmail_html(self):
         html = html_sanitize(test_mail_examples.QUOTE_HOTMAIL_HTML)
@@ -225,7 +225,7 @@ class TestSanitizer(unittest.TestCase):
             for text in in_lst:
                 self.assertIn(text, new_html)
             for text in out_lst:
-                self.assertIn('<span data-o-mail-quote="1">%s</span>' % misc.html_escape(text), new_html)
+                self.assertIn(u'<span data-o-mail-quote="1">%s</span>' % misc.html_escape(text), new_html)
 
     def test_quote_signature(self):
         test_data = [
@@ -244,27 +244,27 @@ class TestSanitizer(unittest.TestCase):
         for ext in test_mail_examples.GMAIL_1_IN:
             self.assertIn(ext, html)
         for ext in test_mail_examples.GMAIL_1_OUT:
-            self.assertIn('<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext), html)
+            self.assertIn(u'<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext), html)
 
     def test_quote_text(self):
         html = html_sanitize(test_mail_examples.TEXT_1)
         for ext in test_mail_examples.TEXT_1_IN:
             self.assertIn(ext, html)
         for ext in test_mail_examples.TEXT_1_OUT:
-            self.assertIn('<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext), html)
+            self.assertIn(u'<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext), html)
 
         html = html_sanitize(test_mail_examples.TEXT_2)
         for ext in test_mail_examples.TEXT_2_IN:
             self.assertIn(ext, html)
         for ext in test_mail_examples.TEXT_2_OUT:
-            self.assertIn('<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext), html)
+            self.assertIn(u'<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext), html)
 
     def test_quote_bugs(self):
         html = html_sanitize(test_mail_examples.BUG1)
         for ext in test_mail_examples.BUG_1_IN:
             self.assertIn(ext, html)
         for ext in test_mail_examples.BUG_1_OUT:
-            self.assertIn('<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext.decode('utf-8')), html)
+            self.assertIn(u'<span data-o-mail-quote="1">%s</span>' % misc.html_escape(ext), html)
 
     def test_misc(self):
         # False / void should not crash
diff --git a/odoo/addons/base/tests/test_mail_examples.py b/odoo/addons/base/tests/test_mail_examples.py
index dfecfe7c0d4..4f3b5b48300 100644
--- a/odoo/addons/base/tests/test_mail_examples.py
+++ b/odoo/addons/base/tests/test_mail_examples.py
@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 # Part of Odoo. See LICENSE file for full copyright and licensing details.
 
-MISC_HTML_SOURCE = """
+MISC_HTML_SOURCE = u"""
 <font size="2" style="color: rgb(31, 31, 31); font-family: monospace; font-variant: normal; line-height: normal; ">test1</font>
 <div style="color: rgb(31, 31, 31); font-family: monospace; font-variant: normal; line-height: normal; font-size: 12px; font-style: normal; ">
 <b>test2</b></div><div style="color: rgb(31, 31, 31); font-family: monospace; font-variant: normal; line-height: normal; font-size: 12px; ">
@@ -19,7 +19,7 @@ test12</font></div><div><font color="#1f1f1f" face="monospace" size="2"><br></fo
 <a href="javascript:alert('malicious code')">test link</a>
 """
 
-EDI_LIKE_HTML_SOURCE = """<div style="font-family: 'Lucida Grande', Ubuntu, Arial, Verdana, sans-serif; font-size: 12px; color: rgb(34, 34, 34); background-color: #FFF; ">
+EDI_LIKE_HTML_SOURCE = u"""<div style="font-family: 'Lucida Grande', Ubuntu, Arial, Verdana, sans-serif; font-size: 12px; color: rgb(34, 34, 34); background-color: #FFF; ">
     <p>Hello ${object.partner_id.name},</p>
     <p>A new invoice is available for you: </p>
     <p style="border-left: 1px solid #8e0000; margin-left: 30px;">
@@ -63,7 +63,7 @@ EDI_LIKE_HTML_SOURCE = """<div style="font-family: 'Lucida Grande', Ubuntu, Aria
 
 # QUOTES
 
-QUOTE_BLOCKQUOTE = """<html>
+QUOTE_BLOCKQUOTE = u"""<html>
   <head>
     <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
   </head>
@@ -103,14 +103,14 @@ web: <a class="moz-txt-link-freetext" href="https://opener.am">https://opener.am
   </body>
 </html>"""
 
-QUOTE_BLOCKQUOTE_IN = ["""<blockquote cite="mid:CAEJSRZvWvud8c6Qp=wfNG6O1+wK3i_jb33qVrF7XyrgPNjnyUA@mail.gmail.com" type="cite" data-o-mail-quote-node="1" data-o-mail-quote="1">"""]
-QUOTE_BLOCKQUOTE_OUT = ["""-- 
+QUOTE_BLOCKQUOTE_IN = [u"""<blockquote cite="mid:CAEJSRZvWvud8c6Qp=wfNG6O1+wK3i_jb33qVrF7XyrgPNjnyUA@mail.gmail.com" type="cite" data-o-mail-quote-node="1" data-o-mail-quote="1">"""]
+QUOTE_BLOCKQUOTE_OUT = [u"""-- 
 Opener B.V. - Business solutions driven by open source collaboration
 
 Stefan Rijnhart - Consultant/developer"""]
 
 
-QUOTE_THUNDERBIRD_HTML = """<html>
+QUOTE_THUNDERBIRD_HTML = u"""<html>
   <head>
     <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
   </head>
@@ -137,13 +137,13 @@ Raoul Poilvache
 </html>"""
 
 
-QUOTE_THUNDERBIRD_HTML_IN = ["""<blockquote cite="mid:CAP76m_WWFH2KVrbjOxbaozvkmbzZYLWJnQ0n0sy9XpGaCWRf1g@mail.gmail.com" type="cite" data-o-mail-quote-node="1" data-o-mail-quote="1">"""]
-QUOTE_THUNDERBIRD_HTML_OUT = ["""<pre class="moz-signature" cols="72"><span data-o-mail-quote="1">-- 
+QUOTE_THUNDERBIRD_HTML_IN = [u"""<blockquote cite="mid:CAP76m_WWFH2KVrbjOxbaozvkmbzZYLWJnQ0n0sy9XpGaCWRf1g@mail.gmail.com" type="cite" data-o-mail-quote-node="1" data-o-mail-quote="1">"""]
+QUOTE_THUNDERBIRD_HTML_OUT = [u"""<pre class="moz-signature" cols="72"><span data-o-mail-quote="1">-- 
 Raoul Poilvache
 </span></pre>"""]
 
 
-QUOTE_HOTMAIL_HTML = """
+QUOTE_HOTMAIL_HTML = u"""
 <html>
 <head>
 <style><!--
@@ -170,12 +170,12 @@ To: tartelette@example.com grosbedon@example.com<br><br>
 </div>-- <br><div class="ecxgmail_signature">
 Raoul Poilvache</div>
 </div></div></div></body></html>"""
-QUOTE_HOTMAIL_HTML_IN = ["""I don't like that.<br><br>"""]
+QUOTE_HOTMAIL_HTML_IN = [u"""I don't like that.<br><br>"""]
 QUOTE_HOTMAIL_HTML_OUT = [
-    """<hr id="stopSpelling" data-o-mail-quote="1">""",
-    """<div dir="ltr" data-o-mail-quote="1"><b data-o-mail-quote="1"><i data-o-mail-quote="1">Test reply. The suite.</i></b>"""]
+    u"""<hr id="stopSpelling" data-o-mail-quote="1">""",
+    u"""<div dir="ltr" data-o-mail-quote="1"><b data-o-mail-quote="1"><i data-o-mail-quote="1">Test reply. The suite.</i></b>"""]
 
-QUOTE_THUNDERBIRD_1 = """<div>On 11/08/2012 05:29 PM,
+QUOTE_THUNDERBIRD_1 = u"""<div>On 11/08/2012 05:29 PM,
       <a href="mailto:dummy@example.com">dummy@example.com</a> wrote:<br></div>
     <blockquote>
       <div>I contact you about our meeting for tomorrow. Here is the
@@ -203,15 +203,15 @@ Raoul Grosbedonn&#233;e
 </pre>"""
 
 QUOTE_THUNDERBIRD_1_IN = [
-    '<a href="mailto:dummy@example.com">dummy@example.com</a> ',
-    '<blockquote data-o-mail-quote-node="1" data-o-mail-quote="1">',
-    'Ok for me. I am replying directly below your mail, using Thunderbird, with a signature.']
-QUOTE_THUNDERBIRD_1_OUT = ["""-- 
+    u'<a href="mailto:dummy@example.com">dummy@example.com</a> ',
+    u'<blockquote data-o-mail-quote-node="1" data-o-mail-quote="1">',
+    u'Ok for me. I am replying directly below your mail, using Thunderbird, with a signature.']
+QUOTE_THUNDERBIRD_1_OUT = [u"""-- 
 Raoul Grosbedonnée
 """]
 
 
-TEXT_1 = """I contact you about our meeting tomorrow. Here is the schedule I propose:
+TEXT_1 = u"""I contact you about our meeting tomorrow. Here is the schedule I propose:
 9 AM: brainstorming about our new amazing business app
 9.45 AM: summary
 10 AM: meeting with Ignasse to present our app
@@ -219,16 +219,16 @@ Is everything ok for you ?
 --
 MySignature"""
 
-TEXT_1_IN = ["""I contact you about our meeting tomorrow. Here is the schedule I propose:
+TEXT_1_IN = [u"""I contact you about our meeting tomorrow. Here is the schedule I propose:
 9 AM: brainstorming about our new amazing business app
 9.45 AM: summary
 10 AM: meeting with Ignasse to present our app
 Is everything ok for you ?"""]
-TEXT_1_OUT = ["""
+TEXT_1_OUT = [u"""
 --
 MySignature"""]
 
-TEXT_2 = """Salut Raoul!
+TEXT_2 = u"""Salut Raoul!
 Le 28 oct. 2012 à 00:02, Raoul Grosbedon a écrit :
 
 > I contact you about our meeting tomorrow. Here is the schedule I propose: (quote)
@@ -253,8 +253,8 @@ Bert TARTOPOILS
 bert.tartopoils@miam.miam
 """
 
-TEXT_2_IN = ["Salut Raoul!", "Of course. This seems viable."]
-TEXT_2_OUT = ["""
+TEXT_2_IN = [u"Salut Raoul!", "Of course. This seems viable."]
+TEXT_2_OUT = [u"""
 > I contact you about our meeting tomorrow. Here is the schedule I propose: (quote)""",
 """
 > 2012/10/27 Bert Tartopoils :
@@ -272,17 +272,17 @@ TEXT_2_OUT = ["""
 
 # MISC
 
-GMAIL_1 = """Hello,<div><br></div><div>Ok for me. I am replying directly in gmail, without signature.</div><div><br></div><div>Kind regards,</div><div><br></div><div>Demo.<br><br><div>On Thu, Nov 8, 2012 at 5:29 PM,  <span>&lt;<a href="mailto:dummy@example.com">dummy@example.com</a>&gt;</span> wrote:<br><blockquote><div>I contact you about our meeting for tomorrow. Here is the schedule I propose:</div><div><ul><li>9 AM: brainstorming about our new amazing business app&lt;/span&gt;&lt;/li&gt;</li>
+GMAIL_1 = u"""Hello,<div><br></div><div>Ok for me. I am replying directly in gmail, without signature.</div><div><br></div><div>Kind regards,</div><div><br></div><div>Demo.<br><br><div>On Thu, Nov 8, 2012 at 5:29 PM,  <span>&lt;<a href="mailto:dummy@example.com">dummy@example.com</a>&gt;</span> wrote:<br><blockquote><div>I contact you about our meeting for tomorrow. Here is the schedule I propose:</div><div><ul><li>9 AM: brainstorming about our new amazing business app&lt;/span&gt;&lt;/li&gt;</li>
 <li>9.45 AM: summary</li><li>10 AM: meeting with Fabien to present our app</li></ul></div><div>Is everything ok for you ?</div>
 <div><p>-- <br>Administrator</p></div>
 
 <div><p>Log in our portal at: <a href="http://localhost:8069#action=login&amp;db=mail_1&amp;login=demo">http://localhost:8069#action=login&amp;db=mail_1&amp;login=demo</a></p></div>
 </blockquote></div><br></div>"""
 
-GMAIL_1_IN = ['Ok for me. I am replying directly in gmail, without signature.', '<blockquote data-o-mail-quote-node="1" data-o-mail-quote="1">']
+GMAIL_1_IN = [u'Ok for me. I am replying directly in gmail, without signature.', '<blockquote data-o-mail-quote-node="1" data-o-mail-quote="1">']
 GMAIL_1_OUT = []
 
-HOTMAIL_1 = """<div>
+HOTMAIL_1 = u"""<div>
     <div dir="ltr"><br>
         I have an amazing company, i'm learning OpenERP, it is a small company yet, but plannig to grow up quickly.
         <br><br>Kindest regards,<br>xxx<br>
@@ -309,19 +309,19 @@ HOTMAIL_1 = """<div>
         </div>
     </div>
 </div>"""
-HOTMAIL_1_IN = ["""<div dir="ltr"><br>
+HOTMAIL_1_IN = [u"""<div dir="ltr"><br>
         I have an amazing company, i'm learning OpenERP, it is a small company yet, but plannig to grow up quickly.
         <br><br>Kindest regards,<br>xxx<br>"""]
 HOTMAIL_1_OUT = [
-    """<hr id="stopSpelling" data-o-mail-quote="1">""",
-    """<pre data-o-mail-quote="1">
+    u"""<hr id="stopSpelling" data-o-mail-quote="1">""",
+    u"""<pre data-o-mail-quote="1">
                 <a href="http://openerp.com" target="_blank" data-o-mail-quote="1">http://openerp.com</a>
                 Belgium: +32.81.81.37.00
                 U.S.: +1 (650) 307-6736
                 India: +91 (79) 40 500 100
             </pre>"""]
 
-MSOFFICE_1 = """
+MSOFFICE_1 = u"""
 <div>
 <div class="WordSection1">
         <p class="MsoNormal">
@@ -423,8 +423,8 @@ MSOFFICE_1 = """
     </div>
 </div>"""
 
-MSOFFICE_1_IN = ['Our requirements are simple. Just looking to replace some spreadsheets for tracking quotes and possibly using the timecard module.']
-MSOFFICE_1_OUT = ['I noticed you recently downloaded OpenERP.', 'Uou mentioned you wish to use OpenERP in your own company.', 'Belgium: +32.81.81.37.00']
+MSOFFICE_1_IN = [u'Our requirements are simple. Just looking to replace some spreadsheets for tracking quotes and possibly using the timecard module.']
+MSOFFICE_1_OUT = [u'I noticed you recently downloaded OpenERP.', 'Uou mentioned you wish to use OpenERP in your own company.', 'Belgium: +32.81.81.37.00']
 
 
 # ------------------------------------------------------------
@@ -432,7 +432,7 @@ MSOFFICE_1_OUT = ['I noticed you recently downloaded OpenERP.', 'Uou mentioned y
 # ------------------------------------------------------------
 
 # bug: read more not apparent, strange message in read more span
-BUG1 = """<pre>Hi Migration Team,
+BUG1 = u"""<pre>Hi Migration Team,
 
 Paragraph 1, blah blah blah blah blah blah blah blah blah blah blah blah 
 blah blah blah blah blah blah blah blah blah blah blah blah blah blah 
@@ -472,10 +472,10 @@ Tel: +32.81.81.37.00
 Web: http://www.openerp.com</pre>"""
 
 BUG_1_IN = [
-    'Hi Migration Team',
-    'Paragraph 1'
+    u'Hi Migration Team',
+    u'Paragraph 1'
 ]
-BUG_1_OUT = ["""
+BUG_1_OUT = [u"""
 -- 
 Olivier Laurent
 Migration Manager
@@ -486,7 +486,7 @@ Tel: +32.81.81.37.00
 Web: http://www.openerp.com"""]
 
 
-REMOVE_CLASS = """
+REMOVE_CLASS = u"""
 <div style="FONT-SIZE: 12pt; FONT-FAMILY: 'Times New Roman'; COLOR: #000000">
     <div>Hello</div>
     <div>I have just installed Odoo 9 and I've got the following error:</div>
@@ -510,9 +510,9 @@ REMOVE_CLASS = """
 </div>
 """
 REMOVE_CLASS_IN = [
-    '<div style="font-size: 12pt; font-family: \'Times New Roman\'; color: #000000">',
-    'An error occured in a modal and I will send you back the html to try opening one on your end']
+    u'<div style="font-size: 12pt; font-family: \'Times New Roman\'; color: #000000">',
+    u'An error occured in a modal and I will send you back the html to try opening one on your end']
 REMOVE_CLASS_OUT = [
-    '<div class="modal-backdrop in">',
-    '<div class="modal-content openerp">',
-    '<div class="modal-header">']
+    u'<div class="modal-backdrop in">',
+    u'<div class="modal-content openerp">',
+    u'<div class="modal-header">']
diff --git a/odoo/tools/mail.py b/odoo/tools/mail.py
index 530f3e63862..b45de63cc71 100644
--- a/odoo/tools/mail.py
+++ b/odoo/tools/mail.py
@@ -170,7 +170,7 @@ def html_sanitize(src, silent=True, sanitize_tags=True, sanitize_attributes=Fals
     src = ustr(src, errors='replace')
     # html: remove encoding attribute inside tags
     doctype = re.compile(r'(<[^>]*\s)(encoding=(["\'][^"\']*?["\']|[^\s\n\r>]+)(\s[^>]*|/)?>)', re.IGNORECASE | re.DOTALL)
-    src = doctype.sub(r"", src)
+    src = doctype.sub(u"", src)
 
     logger = logging.getLogger(__name__ + '.html_sanitize')
 
@@ -178,10 +178,10 @@ def html_sanitize(src, silent=True, sanitize_tags=True, sanitize_attributes=Fals
     part = re.compile(r"(<(([^a<>]|a[^<>\s])[^<>]*)@[^<>]+>)", re.IGNORECASE | re.DOTALL)
     # remove results containing cite="mid:email_like@address" (ex: blockquote cite)
     # cite_except = re.compile(r"^((?!cite[\s]*=['\"]).)*$", re.IGNORECASE)
-    src = part.sub(lambda m: ('cite=' not in m.group(1) and 'alt=' not in m.group(1)) and misc.html_escape(m.group(1)) or m.group(1), src)
+    src = part.sub(lambda m: (u'cite=' not in m.group(1) and u'alt=' not in m.group(1)) and misc.html_escape(m.group(1)) or m.group(1), src)
     # html encode mako tags <% ... %> to decode them later and keep them alive, otherwise they are stripped by the cleaner
-    src = src.replace('<%', misc.html_escape('<%'))
-    src = src.replace('%>', misc.html_escape('%>'))
+    src = src.replace(u'<%', misc.html_escape(u'<%'))
+    src = src.replace(u'%>', misc.html_escape(u'%>'))
 
     kwargs = {
         'page_structure': True,
@@ -222,33 +222,34 @@ def html_sanitize(src, silent=True, sanitize_tags=True, sanitize_attributes=Fals
         # some corner cases make the parser crash (such as <SCRIPT/XSS SRC=\"http://ha.ckers.org/xss.js\"></SCRIPT> in test_mail)
         cleaner = _Cleaner(**kwargs)
         cleaned = cleaner.clean_html(src)
+        assert isinstance(cleaned, pycompat.text_type)
         # MAKO compatibility: $, { and } inside quotes are escaped, preventing correct mako execution
-        cleaned = cleaned.replace('%24', '$')
-        cleaned = cleaned.replace('%7B', '{')
-        cleaned = cleaned.replace('%7D', '}')
-        cleaned = cleaned.replace('%20', ' ')
-        cleaned = cleaned.replace('%5B', '[')
-        cleaned = cleaned.replace('%5D', ']')
-        cleaned = cleaned.replace('%7C', '|')
-        cleaned = cleaned.replace('&lt;%', '<%')
-        cleaned = cleaned.replace('%&gt;', '%>')
+        cleaned = cleaned.replace(u'%24', u'$')
+        cleaned = cleaned.replace(u'%7B', u'{')
+        cleaned = cleaned.replace(u'%7D', u'}')
+        cleaned = cleaned.replace(u'%20', u' ')
+        cleaned = cleaned.replace(u'%5B', u'[')
+        cleaned = cleaned.replace(u'%5D', u']')
+        cleaned = cleaned.replace(u'%7C', u'|')
+        cleaned = cleaned.replace(u'&lt;%', u'<%')
+        cleaned = cleaned.replace(u'%&gt;', u'%>')
         # html considerations so real html content match database value
-        cleaned.replace(u'\xa0', '&nbsp;')
+        cleaned.replace(u'\xa0', u'&nbsp;')
     except etree.ParserError as e:
-        if 'empty' in str(e):
-            return ""
+        if u'empty' in pycompat.text_type(e):
+            return u""
         if not silent:
             raise
-        logger.warning('ParserError obtained when sanitizing %r', src, exc_info=True)
-        cleaned = '<p>ParserError when sanitizing</p>'
+        logger.warning(u'ParserError obtained when sanitizing %r', src, exc_info=True)
+        cleaned = u'<p>ParserError when sanitizing</p>'
     except Exception:
         if not silent:
             raise
-        logger.warning('unknown error obtained when sanitizing %r', src, exc_info=True)
-        cleaned = '<p>Unknown error when sanitizing</p>'
+        logger.warning(u'unknown error obtained when sanitizing %r', src, exc_info=True)
+        cleaned = u'<p>Unknown error when sanitizing</p>'
 
     # this is ugly, but lxml/etree tostring want to put everything in a 'div' that breaks the editor -> remove that
-    if cleaned.startswith('<div>') and cleaned.endswith('</div>'):
+    if cleaned.startswith(u'<div>') and cleaned.endswith(u'</div>'):
         cleaned = cleaned[5:-6]
 
     return cleaned