[FIX] tools: mail: fixed / improved html_email_clean.
Read more links could not appear when the shorten limit was obtained inside a quote that is removed in the process, removing the read more link node. It now get the first parent that is not inside a quote to be sure the link is displayed. Added the test-case that helped find this bug. bzr revid: tde@openerp.com-20131017110525-etn0u5s91zbnvul1
This commit is contained in:
Thibault Delavallée
@@ -230,6 +230,12 @@ class TestCleaner(unittest2.TestCase):
|
||||
for ext in test_mail_examples.BUG_1_OUT:
|
||||
self.assertNotIn(ext, new_html, 'html_email_cleaner did not removed invalid content')
|
||||
|
||||
new_html = html_email_clean(test_mail_examples.BUG2, remove=True, shorten=True, max_length=100)
|
||||
for ext in test_mail_examples.BUG_2_IN:
|
||||
self.assertIn(ext, new_html, 'html_email_cleaner wrongly removed valid content')
|
||||
for ext in test_mail_examples.BUG_2_OUT:
|
||||
self.assertNotIn(ext, new_html, 'html_email_cleaner did not removed invalid content')
|
||||
|
||||
def test_90_misc(self):
|
||||
# False boolean for text must return empty string
|
||||
new_html = html_email_clean(False)
|
||||
|
||||
@@ -693,3 +693,266 @@ BUG_1_OUT = [
|
||||
'81.81.37.00',
|
||||
'openerp.com',
|
||||
]
|
||||
|
||||
|
||||
BUG2 = """
|
||||
<div>
|
||||
<br>
|
||||
<div class="moz-forward-container"><br>
|
||||
<br>
|
||||
-------- Original Message --------
|
||||
<table class="moz-email-headers-table" border="0" cellpadding="0" cellspacing="0">
|
||||
<tbody>
|
||||
<tr>
|
||||
<th nowrap="" valign="BASELINE" align="RIGHT">Subject:
|
||||
</th>
|
||||
<td>Fwd: TR: OpenERP S.A. Payment Reminder</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<th nowrap="" valign="BASELINE" align="RIGHT">Date: </th>
|
||||
<td>Wed, 16 Oct 2013 14:11:13 +0200</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<th nowrap="" valign="BASELINE" align="RIGHT">From: </th>
|
||||
<td>Christine Herrmann <a class="moz-txt-link-rfc2396E" href="mailto:che@openerp.com"><che@openerp.com></a></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<th nowrap="" valign="BASELINE" align="RIGHT">To: </th>
|
||||
<td><a class="moz-txt-link-abbreviated" href="mailto:online@openerp.com">online@openerp.com</a></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<br>
|
||||
<br>
|
||||
|
||||
<br>
|
||||
<div class="moz-forward-container"><br>
|
||||
<br>
|
||||
-------- Message original --------
|
||||
<table class="moz-email-headers-table" border="0" cellpadding="0" cellspacing="0">
|
||||
<tbody>
|
||||
<tr>
|
||||
<th nowrap="" valign="BASELINE" align="RIGHT">Sujet:
|
||||
</th>
|
||||
<td>TR: OpenERP S.A. Payment Reminder</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<th nowrap="" valign="BASELINE" align="RIGHT">Date :
|
||||
</th>
|
||||
<td>Wed, 16 Oct 2013 10:34:45 -0000</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<th nowrap="" valign="BASELINE" align="RIGHT">De : </th>
|
||||
<td>Ida Siwatala <a class="moz-txt-link-rfc2396E" href="mailto:infos@inzoservices.com"><infos@inzoservices.com></a></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<th nowrap="" valign="BASELINE" align="RIGHT">Répondre
|
||||
|
||||
à : </th>
|
||||
<td><a class="moz-txt-link-abbreviated" href="mailto:catchall@openerp.my.openerp.com">catchall@openerp.my.openerp.com</a></td>
|
||||
</tr>
|
||||
<tr>
|
||||
<th nowrap="" valign="BASELINE" align="RIGHT">Pour :
|
||||
</th>
|
||||
<td>Christine Herrmann (che) <a class="moz-txt-link-rfc2396E" href="mailto:che@openerp.com"><che@openerp.com></a></td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<br>
|
||||
<br>
|
||||
<div>
|
||||
<div class="WordSection1">
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Bonjour,</span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"></span></p>
|
||||
<p> </p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Pourriez-vous
|
||||
|
||||
me faire un retour sur ce point.</span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"></span></p>
|
||||
<p> </p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Cordialement</span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"></span></p>
|
||||
<p> </p>
|
||||
<div>
|
||||
<div style="border:none;border-top:solid #B5C4DF
|
||||
1.0pt;padding:3.0pt 0cm 0cm 0cm">
|
||||
<p class="MsoNormal"><b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">De :</span></b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">
|
||||
Ida Siwatala [<a class="moz-txt-link-freetext" href="mailto:infos@inzoservices.com">mailto:infos@inzoservices.com</a>]
|
||||
<br>
|
||||
<b>Envoyé :</b> vendredi 4 octobre 2013 20:03<br>
|
||||
<b>À :</b> 'Followers of
|
||||
INZO-services-8-all-e-Maxime-Lisbonne-77176-Savigny-le-temple-France'<br>
|
||||
<b>Objet :</b> RE: OpenERP S.A. Payment Reminder</span></p>
|
||||
</div>
|
||||
</div>
|
||||
<p> </p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Bonsoir,</span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"></span></p>
|
||||
<p> </p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Je
|
||||
|
||||
me permets de revenir vers vous par écrit , car j’ai
|
||||
fait 2 appels vers votre service en exposant mon
|
||||
problème, mais je n’ai pas eu de retour.</span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Cela
|
||||
|
||||
fait un mois que j’ai fait la souscription de votre
|
||||
produit, mais je me rends compte qu’il est pas adapté à
|
||||
ma situation ( fonctionnalité manquante et surtout je
|
||||
n’ai pas beaucoup de temps à passer à résoudre des
|
||||
bugs). </span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">C’est
|
||||
|
||||
pourquoi , j’ai demandé qu’un accord soit trouvé avec
|
||||
vous pour annuler le contrat (tout en vous payant le
|
||||
mois d’utilisation de septembre).</span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"></span></p>
|
||||
<p> </p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Pourriez-vous
|
||||
|
||||
me faire un retour sur ce point.</span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"></span></p>
|
||||
<p> </p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Cordialement,</span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"></span></p>
|
||||
<p> </p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D">Ida
|
||||
|
||||
Siwatala</span></p>
|
||||
<p class="MsoNormal"><span style="font-size:11.0pt;font-family:"Calibri","sans-serif";color:#1F497D"></span></p>
|
||||
<p> </p>
|
||||
<p class="MsoNormal"><b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">De :</span></b><span style="font-size:10.0pt;font-family:"Tahoma","sans-serif"">
|
||||
<a href="mailto:che@openerp.com">che@openerp.com</a>
|
||||
[<a href="mailto:che@openerp.com">mailto:che@openerp.com</a>]
|
||||
<br>
|
||||
<b>Envoyé :</b> vendredi 4 octobre 2013 17:41<br>
|
||||
<b>À :</b> <a href="mailto:infos@inzoservices.com">infos@inzoservices.com</a><br>
|
||||
<b>Objet :</b> OpenERP S.A. Payment Reminder</span></p>
|
||||
<p> </p>
|
||||
<div>
|
||||
<p style="background:white"><span style="font-size:9.0pt;font-family:"Arial","sans-serif";color:#222222">Dear
|
||||
|
||||
INZO services,</span></p>
|
||||
<p style="background:white"><span style="font-size:9.0pt;font-family:"Arial","sans-serif";color:#222222">Exception
|
||||
|
||||
made if there was a mistake of ours, it seems that the
|
||||
following amount stays unpaid. Please, take
|
||||
appropriate measures in order to carry out this
|
||||
payment in the next 8 days. </span></p>
|
||||
<p class="MsoNormal" style="background:white"><span style="font-size:9.0pt;font-family:"Arial","sans-serif";color:#222222"></span></p>
|
||||
<p> </p>
|
||||
<table class="MsoNormalTable" style="width:100.0%;border:outset 1.5pt" width="100%" border="1" cellpadding="0">
|
||||
<tbody>
|
||||
<tr>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal">Date de facturation</p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal">Description</p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal">Reference</p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal">Due Date</p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal">Amount (€)</p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal">Lit.</p>
|
||||
</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal"><b>2013-09-24</b></p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal"><b>2013/1121</b></p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal"><b>Enterprise - Inzo Services
|
||||
- Juillet 2013</b></p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal"><b>2013-09-24</b></p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt">
|
||||
<p class="MsoNormal"><b>420.0</b></p>
|
||||
</td>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt"><br>
|
||||
</td>
|
||||
</tr>
|
||||
<tr>
|
||||
<td style="padding:.75pt .75pt .75pt .75pt"><br>
|
||||
</td>
|
||||
<td style="border:none;padding:.75pt .75pt .75pt
|
||||
.75pt"><br>
|
||||
</td>
|
||||
<td style="border:none;padding:.75pt .75pt .75pt
|
||||
.75pt"><br>
|
||||
</td>
|
||||
<td style="border:none;padding:.75pt .75pt .75pt
|
||||
.75pt"><br>
|
||||
</td>
|
||||
<td style="border:none;padding:.75pt .75pt .75pt
|
||||
.75pt"><br>
|
||||
</td>
|
||||
<td style="border:none;padding:.75pt .75pt .75pt
|
||||
.75pt"><br>
|
||||
</td>
|
||||
</tr>
|
||||
</tbody>
|
||||
</table>
|
||||
<p class="MsoNormal" style="text-align:center;background:white" align="center"><span style="font-size:9.0pt;font-family:"Arial","sans-serif";color:#222222">Amount
|
||||
|
||||
due : 420.00 € </span></p>
|
||||
<p style="background:white"><span style="font-size:9.0pt;font-family:"Arial","sans-serif";color:#222222">Would
|
||||
|
||||
your payment have been carried out after this mail was
|
||||
sent, please ignore this message. Do not hesitate to
|
||||
contact our accounting department. </span></p>
|
||||
<p class="MsoNormal" style="background:white"><span style="font-size:9.0pt;font-family:"Arial","sans-serif";color:#222222"><br>
|
||||
Best Regards, <br>
|
||||
Aurore Lesage <br>
|
||||
OpenERP<br>
|
||||
Chaussée de Namur, 40 <br>
|
||||
B-1367 Grand Rosières <br>
|
||||
Tel: +32.81.81.37.00 - Fax: +32.81.73.35.01 <br>
|
||||
E-mail : <a href="mailto:ale@openerp.com">ale@openerp.com</a> <br>
|
||||
Web: <a href="http://www.openerp.com">http://www.openerp.com</a></span></p>
|
||||
</div>
|
||||
</div>
|
||||
</div>
|
||||
--<br>
|
||||
INZO services <small>Sent by <a style="color:inherit" href="http://www.openerp.com">OpenERP
|
||||
S.A.</a> using <a style="color:inherit" href="https://www.openerp.com/">OpenERP</a>.</small>
|
||||
<small>Access your messages and documents <a style="color:inherit" href="https://accounts.openerp.com?db=openerp#action=mail.action_mail_redirect&login=che&message_id=5750830">in
|
||||
|
||||
OpenERP</a></small> <br>
|
||||
<pre class="moz-signature" cols="72">--
|
||||
Christine Herrmann
|
||||
|
||||
OpenERP
|
||||
Chaussée de Namur, 40
|
||||
B-1367 Grand Rosières
|
||||
Tel: +32.81.81.37.00 - Fax: +32.81.73.35.01
|
||||
|
||||
Web: <a class="moz-txt-link-freetext" href="http://www.openerp.com">http://www.openerp.com</a> </pre>
|
||||
<br>
|
||||
</div>
|
||||
<br>
|
||||
<br>
|
||||
</div>
|
||||
<br>
|
||||
|
||||
</div>"""
|
||||
|
||||
BUG_2_IN = [
|
||||
'read more',
|
||||
'...',
|
||||
]
|
||||
BUG_2_OUT = [
|
||||
'Fwd: TR: OpenERP S.A'
|
||||
'fait un mois'
|
||||
]
|
||||
|
||||
+42
-21
@@ -175,6 +175,32 @@ def html_email_clean(html, remove=False, shorten=False, max_length=300):
|
||||
iteration += 1
|
||||
new_node = _insert_new_node(node, -1, new_node_tag, text[idx:] + (cur_node.tail or ''), None, {})
|
||||
|
||||
def _truncate_node(node, position, find_first_blank=True):
|
||||
if node.text is None:
|
||||
node.text = ''
|
||||
# truncate text
|
||||
end_position = position if len(node.text) >= position else len(node.text)
|
||||
innertext = node.text[0:end_position]
|
||||
outertext = node.text[end_position:]
|
||||
if find_first_blank:
|
||||
stop_idx = outertext.find(' ')
|
||||
if stop_idx == -1:
|
||||
stop_idx = len(outertext)
|
||||
else:
|
||||
stop_idx = 0
|
||||
node.text = innertext + outertext[0:stop_idx]
|
||||
# create <span> ... <a href="#">read more</a></span> node
|
||||
read_more_node = _create_node('span', ' ... ', None, {'class': 'oe_mail_expand'})
|
||||
read_more_link_node = _create_node('a', 'read more', None, {'href': '#', 'class': 'oe_mail_expand'})
|
||||
read_more_node.append(read_more_link_node)
|
||||
# create outertext node
|
||||
overtext_node = _create_node('span', outertext[stop_idx:])
|
||||
# tag node
|
||||
overtext_node.set('in_overlength', '1')
|
||||
# add newly created nodes in dom
|
||||
node.append(read_more_node)
|
||||
node.append(overtext_node)
|
||||
|
||||
if not html or not isinstance(html, basestring):
|
||||
return html
|
||||
html = ustr(html)
|
||||
@@ -226,14 +252,16 @@ def html_email_clean(html, remove=False, shorten=False, max_length=300):
|
||||
if 'SkyDrivePlaceholder' in node.get('class', '') or 'SkyDrivePlaceholder' in node.get('id', ''):
|
||||
root.set('hotmail', '1')
|
||||
|
||||
# state of the parsing
|
||||
# state of the parsing: flag quotes and tails to remove
|
||||
if quote_begin:
|
||||
node.set('in_quote', '1')
|
||||
node.set('tail_remove', '1')
|
||||
# state of the parsing: flag when being in over-length content
|
||||
if overlength:
|
||||
node.set('in_overlength', '1')
|
||||
node.set('tail_remove', '1')
|
||||
|
||||
# find quote in msoffice / hotmail / blockquote / text quote and signatures
|
||||
if root.get('msoffice') and node.tag == 'div' and 'border-top:solid' in node.get('style', ''):
|
||||
quote_begin = True
|
||||
node.set('in_quote', '1')
|
||||
@@ -242,35 +270,28 @@ def html_email_clean(html, remove=False, shorten=False, max_length=300):
|
||||
quote_begin = True
|
||||
node.set('in_quote', '1')
|
||||
node.set('tail_remove', '1')
|
||||
if node.tag == 'blockquote' or node.get('text_quote') or node.get('text_signature'):
|
||||
node.set('in_quote', '1')
|
||||
|
||||
# shorten:
|
||||
# 1/ truncate the text at the next available space
|
||||
# 2/ create a 'read more' node, next to current node
|
||||
# 3/ add the truncated text in a new node, next to 'read more' node
|
||||
if shorten and not overlength and cur_char_nbr + len(node.text or '') > max_length:
|
||||
node_to_truncate = node
|
||||
while node.get('in_quote') and node.getparent():
|
||||
node_to_truncate = node.getparent()
|
||||
overlength = True
|
||||
# truncate text
|
||||
innertext = node.text[0:(max_length - cur_char_nbr)]
|
||||
outertext = node.text[(max_length - cur_char_nbr):]
|
||||
stop_idx = outertext.find(' ')
|
||||
if stop_idx == -1:
|
||||
stop_idx = len(outertext)
|
||||
node.text = innertext + outertext[0:stop_idx]
|
||||
# create <span> ... <a href="#">read more</a></span> node
|
||||
read_more_node = _create_node('span', ' ... ', None, {'class': 'oe_mail_expand'})
|
||||
read_more_link_node = _create_node('a', 'read more', None, {'href': '#', 'class': 'oe_mail_expand'})
|
||||
read_more_node.append(read_more_link_node)
|
||||
# create outertext node
|
||||
new_node = _create_node('span', outertext[stop_idx:])
|
||||
# add newly created nodes in dom
|
||||
node.append(read_more_node)
|
||||
# tag node
|
||||
new_node.set('in_overlength', '1')
|
||||
node_to_truncate.set('truncate', '1')
|
||||
node_to_truncate.set('truncate_position', str(max_length - cur_char_nbr))
|
||||
cur_char_nbr += len(node.text or '')
|
||||
|
||||
cur_char_nbr += len(node.text or '')
|
||||
# Tree modification
|
||||
# ------------------------------------------------------------
|
||||
|
||||
if node.tag == 'blockquote' or node.get('text_quote') or node.get('text_signature'):
|
||||
node.set('in_quote', '1')
|
||||
for node in root.iter():
|
||||
if node.get('truncate'):
|
||||
_truncate_node(node, int(node.get('truncate_position', '0')))
|
||||
|
||||
# Post processing
|
||||
# ------------------------------------------------------------
|
||||
|
||||
Reference in New Issue
Block a user