With the introduction of Knowledge Behavior Component, came a need to create
html nodes which would have limited interactions with the editor. i.e. an Odoo
view already has everything it needs to function properly, and when it is
inserted in the editor, any manipulation on the selection or on the style that
could be done with it should be prevented. Another example would be the
/template block (will be renamed /clipboard in the future) that has a
non-editable part (buttons which have a definite action in Odoo, and which
should not be interacted with) as well as an editable part inside of it).
To solve this use case, this commit proposes to mark specific html nodes with
a `data-oe-protected` attribute which could have one of three values:
- "true"
- Only mutations of type "attributes" can be registered on the node itself
which has the `data-oe-protected="true"` attribute
- Prevent mutations of children (and sub-children) from being registered by
the mutationObserver of the editor
- Prevent the selection handling when its anchor is inside a
`data-oe-protected="true"` element, even if it is `contenteditable="false"`
- Prevent the command hint
- Prevent the usage of the wysiwyg toolbar
- Prevent the dblClick tooltip
- Prevent the editor sanitization `Sanitize.js`
- "false"
- Designed to be contained inside a node with `data-oe-protected="true"`
- Re-enable all features disabled by a parent node with
`data-oe-protected="true" for the children of a node with
`data-oe-protected="false"
- ("")
- This is considered equivalent to have the `data-oe-protected` attribute
set to "true" (like other html attributes).
Another attribute is added: `data-oe-transient-content`, with the following
values:
- "true"
- Prevent the serialization of the children of the node, so they are not
shared during a collaboration.
- Transient nodes will be removed during `cleanForSave`, meaning that they
will never be part of the html_field value in the database
- ("")
- equivalent to "true"
The use case is an embedded view: there is a large quantity of nodes that
are not relevant to share nor to save, since it will be recreated with the
lastest data from the database, with the information relevant to the
currently active user each time it has to be rendered.
Note:
This commit does not handle the dynamic switch from a specific value for
`data-oe-protected` to another (i.e. switching from "false" to "" or "true").
This could cause a number of problems like:
- some mutations from when the value was "true" are not yet handled when the
switch (to "false") happens => those mutations will be registered as if they
were always under the "false" value, even though it is not the case.
- in collaborative, some nodes with oids that were not relevant (under the value
"true") won't necessarily have the same oids in between collaborators.
Therefore we cannot suddently listen to their mutations and expect the changes
to be shared by switching to "false".
In conclusion: the `data-oe-protected` attribute value should stay the same
during the entire edition.
Task-2821374
Part-of: odoo/odoo#104680
678 lines
25 KiB
Python
678 lines
25 KiB
Python
# -*- coding: utf-8 -*-
|
|
# Part of Odoo. See LICENSE file for full copyright and licensing details.
|
|
|
|
import base64
|
|
import collections
|
|
import logging
|
|
import random
|
|
import re
|
|
import socket
|
|
import threading
|
|
import time
|
|
from email.utils import getaddresses
|
|
from urllib.parse import urlparse
|
|
|
|
import idna
|
|
import markupsafe
|
|
from lxml import etree, html
|
|
from lxml.html import clean
|
|
from werkzeug import urls
|
|
|
|
import odoo
|
|
from odoo.loglevels import ustr
|
|
from odoo.tools import misc
|
|
|
|
_logger = logging.getLogger(__name__)
|
|
|
|
#----------------------------------------------------------
|
|
# HTML Sanitizer
|
|
#----------------------------------------------------------
|
|
|
|
safe_attrs = clean.defs.safe_attrs | frozenset(
|
|
['style',
|
|
'data-o-mail-quote', # quote detection
|
|
'data-oe-model', 'data-oe-id', 'data-oe-field', 'data-oe-type', 'data-oe-expression', 'data-oe-translation-initial-sha', 'data-oe-nodeid',
|
|
'data-last-history-steps', 'data-oe-protected', 'data-oe-transient-content',
|
|
'data-publish', 'data-id', 'data-res_id', 'data-interval', 'data-member_id', 'data-scroll-background-ratio', 'data-view-id',
|
|
'data-class', 'data-mimetype', 'data-original-src', 'data-original-id', 'data-gl-filter', 'data-quality', 'data-resize-width',
|
|
'data-shape', 'data-shape-colors', 'data-file-name', 'data-original-mimetype',
|
|
'data-behavior-props', 'data-prop-name', # knowledge commands
|
|
])
|
|
SANITIZE_TAGS = {
|
|
# allow new semantic HTML5 tags
|
|
'allow_tags': clean.defs.tags | frozenset('article bdi section header footer hgroup nav aside figure main'.split() + [etree.Comment]),
|
|
'kill_tags': ['base', 'embed', 'frame', 'head', 'iframe', 'link', 'meta',
|
|
'noscript', 'object', 'script', 'style', 'title'],
|
|
'remove_tags': ['html', 'body'],
|
|
}
|
|
|
|
|
|
class _Cleaner(clean.Cleaner):
|
|
|
|
_style_re = re.compile(r'''([\w-]+)\s*:\s*((?:[^;"']|"[^";]*"|'[^';]*')+)''')
|
|
|
|
_style_whitelist = [
|
|
'font-size', 'font-family', 'font-weight', 'font-style', 'background-color', 'color', 'text-align',
|
|
'line-height', 'letter-spacing', 'text-transform', 'text-decoration', 'text-decoration', 'opacity',
|
|
'float', 'vertical-align', 'display',
|
|
'padding', 'padding-top', 'padding-left', 'padding-bottom', 'padding-right',
|
|
'margin', 'margin-top', 'margin-left', 'margin-bottom', 'margin-right',
|
|
'white-space',
|
|
# box model
|
|
'border', 'border-color', 'border-radius', 'border-style', 'border-width', 'border-top', 'border-bottom',
|
|
'height', 'width', 'max-width', 'min-width', 'min-height',
|
|
# tables
|
|
'border-collapse', 'border-spacing', 'caption-side', 'empty-cells', 'table-layout']
|
|
|
|
_style_whitelist.extend(
|
|
['border-%s-%s' % (position, attribute)
|
|
for position in ['top', 'bottom', 'left', 'right']
|
|
for attribute in ('style', 'color', 'width', 'left-radius', 'right-radius')]
|
|
)
|
|
|
|
strip_classes = False
|
|
sanitize_style = False
|
|
|
|
def __call__(self, doc):
|
|
super(_Cleaner, self).__call__(doc)
|
|
|
|
# if we keep attributes but still remove classes
|
|
if not getattr(self, 'safe_attrs_only', False) and self.strip_classes:
|
|
for el in doc.iter(tag=etree.Element):
|
|
self.strip_class(el)
|
|
|
|
# if we keep style attribute, sanitize them
|
|
if not self.style and self.sanitize_style:
|
|
for el in doc.iter(tag=etree.Element):
|
|
self.parse_style(el)
|
|
|
|
def strip_class(self, el):
|
|
if el.attrib.get('class'):
|
|
del el.attrib['class']
|
|
|
|
def parse_style(self, el):
|
|
attributes = el.attrib
|
|
styling = attributes.get('style')
|
|
if styling:
|
|
valid_styles = collections.OrderedDict()
|
|
styles = self._style_re.findall(styling)
|
|
for style in styles:
|
|
if style[0].lower() in self._style_whitelist:
|
|
valid_styles[style[0].lower()] = style[1]
|
|
if valid_styles:
|
|
el.attrib['style'] = '; '.join('%s:%s' % (key, val) for (key, val) in valid_styles.items())
|
|
else:
|
|
del el.attrib['style']
|
|
|
|
|
|
def tag_quote(el):
|
|
def _create_new_node(tag, text, tail=None, attrs=None):
|
|
new_node = etree.Element(tag)
|
|
new_node.text = text
|
|
new_node.tail = tail
|
|
if attrs:
|
|
for key, val in attrs.items():
|
|
new_node.set(key, val)
|
|
return new_node
|
|
|
|
def _tag_matching_regex_in_text(regex, node, tag='span', attrs=None):
|
|
text = node.text or ''
|
|
if not re.search(regex, text):
|
|
return
|
|
|
|
child_node = None
|
|
idx, node_idx = 0, 0
|
|
for item in re.finditer(regex, text):
|
|
new_node = _create_new_node(tag, text[item.start():item.end()], None, attrs)
|
|
if child_node is None:
|
|
node.text = text[idx:item.start()]
|
|
new_node.tail = text[item.end():]
|
|
node.insert(node_idx, new_node)
|
|
else:
|
|
child_node.tail = text[idx:item.start()]
|
|
new_node.tail = text[item.end():]
|
|
node.insert(node_idx, new_node)
|
|
child_node = new_node
|
|
idx = item.end()
|
|
node_idx = node_idx + 1
|
|
|
|
el_class = el.get('class', '') or ''
|
|
el_id = el.get('id', '') or ''
|
|
|
|
# gmail or yahoo // # outlook, html // # msoffice
|
|
if 'gmail_extra' in el_class or \
|
|
'divRplyFwdMsg' in el_id or \
|
|
('SkyDrivePlaceholder' in el_class or 'SkyDrivePlaceholder' in el_class):
|
|
el.set('data-o-mail-quote', '1')
|
|
if el.getparent() is not None:
|
|
el.getparent().set('data-o-mail-quote-container', '1')
|
|
|
|
if (el.tag == 'hr' and ('stopSpelling' in el_class or 'stopSpelling' in el_id)) or \
|
|
'yahoo_quoted' in el_class:
|
|
# Quote all elements after this one
|
|
el.set('data-o-mail-quote', '1')
|
|
for sibling in el.itersiblings(preceding=False):
|
|
sibling.set('data-o-mail-quote', '1')
|
|
|
|
# html signature (-- <br />blah)
|
|
signature_begin = re.compile(r"((?:(?:^|\n)[-]{2}[\s]?$))")
|
|
if el.text and el.find('br') is not None and re.search(signature_begin, el.text):
|
|
el.set('data-o-mail-quote', '1')
|
|
if el.getparent() is not None:
|
|
el.getparent().set('data-o-mail-quote-container', '1')
|
|
|
|
# text-based quotes (>, >>) and signatures (-- Signature)
|
|
text_complete_regex = re.compile(r"((?:\n[>]+[^\n\r]*)+|(?:(?:^|\n)[-]{2}[\s]?[\r\n]{1,2}[\s\S]+))")
|
|
if not el.get('data-o-mail-quote'):
|
|
_tag_matching_regex_in_text(text_complete_regex, el, 'span', {'data-o-mail-quote': '1'})
|
|
|
|
if el.tag == 'blockquote':
|
|
# remove single node
|
|
el.set('data-o-mail-quote-node', '1')
|
|
el.set('data-o-mail-quote', '1')
|
|
if el.getparent() is not None and (el.getparent().get('data-o-mail-quote') or el.getparent().get('data-o-mail-quote-container')) and not el.getparent().get('data-o-mail-quote-node'):
|
|
el.set('data-o-mail-quote', '1')
|
|
|
|
|
|
def html_normalize(src, filter_callback=None):
|
|
""" Normalize `src` for storage as an html field value.
|
|
|
|
The string is parsed as an html tag soup, made valid, then decorated for
|
|
"email quote" detection, and prepared for an optional filtering.
|
|
The filtering step (e.g. sanitization) should be performed by the
|
|
`filter_callback` function (to avoid multiple parsing operations, and
|
|
normalize the result).
|
|
|
|
:param src: the html string to normalize
|
|
:param filter_callback: optional callable taking a single `etree._Element`
|
|
document parameter, to be called during normalization in order to
|
|
filter the output document
|
|
"""
|
|
|
|
if not src:
|
|
return src
|
|
|
|
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(u"", src)
|
|
|
|
try:
|
|
doc = html.fromstring(src)
|
|
except etree.ParserError as e:
|
|
# HTML comment only string, whitespace only..
|
|
if 'empty' in str(e):
|
|
return u""
|
|
raise
|
|
|
|
# perform quote detection before cleaning and class removal
|
|
if doc is not None:
|
|
for el in doc.iter(tag=etree.Element):
|
|
tag_quote(el)
|
|
|
|
if filter_callback:
|
|
doc = filter_callback(doc)
|
|
|
|
src = html.tostring(doc, encoding='unicode')
|
|
|
|
# this is ugly, but lxml/etree tostring want to put everything in a
|
|
# 'div' that breaks the editor -> remove that
|
|
if src.startswith('<div>') and src.endswith('</div>'):
|
|
src = src[5:-6]
|
|
|
|
# html considerations so real html content match database value
|
|
src = src.replace(u'\xa0', u' ')
|
|
|
|
return src
|
|
|
|
|
|
def html_sanitize(src, silent=True, sanitize_tags=True, sanitize_attributes=False, sanitize_style=False, sanitize_form=True, strip_style=False, strip_classes=False):
|
|
if not src:
|
|
return src
|
|
|
|
logger = logging.getLogger(__name__ + '.html_sanitize')
|
|
|
|
def sanitize_handler(doc):
|
|
kwargs = {
|
|
'page_structure': True,
|
|
'style': strip_style, # True = remove style tags/attrs
|
|
'sanitize_style': sanitize_style, # True = sanitize styling
|
|
'forms': sanitize_form, # True = remove form tags
|
|
'remove_unknown_tags': False,
|
|
'comments': False,
|
|
'processing_instructions': False
|
|
}
|
|
if sanitize_tags:
|
|
kwargs.update(SANITIZE_TAGS)
|
|
|
|
if sanitize_attributes: # We keep all attributes in order to keep "style"
|
|
if strip_classes:
|
|
current_safe_attrs = safe_attrs - frozenset(['class'])
|
|
else:
|
|
current_safe_attrs = safe_attrs
|
|
kwargs.update({
|
|
'safe_attrs_only': True,
|
|
'safe_attrs': current_safe_attrs,
|
|
})
|
|
else:
|
|
kwargs.update({
|
|
'safe_attrs_only': False, # keep oe-data attributes + style
|
|
'strip_classes': strip_classes, # remove classes, even when keeping other attributes
|
|
})
|
|
|
|
cleaner = _Cleaner(**kwargs)
|
|
cleaner(doc)
|
|
return doc
|
|
|
|
try:
|
|
sanitized = html_normalize(src, filter_callback=sanitize_handler)
|
|
except etree.ParserError:
|
|
if not silent:
|
|
raise
|
|
logger.warning(u'ParserError obtained when sanitizing %r', src, exc_info=True)
|
|
sanitized = '<p>ParserError when sanitizing</p>'
|
|
except Exception:
|
|
if not silent:
|
|
raise
|
|
logger.warning(u'unknown error obtained when sanitizing %r', src, exc_info=True)
|
|
sanitized = '<p>Unknown error when sanitizing</p>'
|
|
|
|
return markupsafe.Markup(sanitized)
|
|
|
|
# ----------------------------------------------------------
|
|
# HTML/Text management
|
|
# ----------------------------------------------------------
|
|
|
|
URL_REGEX = r'(\bhref=[\'"](?!mailto:|tel:|sms:)([^\'"]+)[\'"])'
|
|
TEXT_URL_REGEX = r'https?://[\w@:%.+&~#=/-]+(?:\?\S+)?'
|
|
# retrieve inner content of the link
|
|
HTML_TAG_URL_REGEX = URL_REGEX + r'([^<>]*>([^<>]+)<\/)?'
|
|
HTML_TAGS_REGEX = re.compile('<.*?>')
|
|
HTML_NEWLINES_REGEX = re.compile('<(div|p|br|tr)[^>]*>|\n')
|
|
|
|
|
|
def validate_url(url):
|
|
if urls.url_parse(url).scheme not in ('http', 'https', 'ftp', 'ftps'):
|
|
return 'http://' + url
|
|
|
|
return url
|
|
|
|
|
|
def is_html_empty(html_content):
|
|
"""Check if a html content is empty. If there are only formatting tags with style
|
|
attributes or a void content return True. Famous use case if a
|
|
'<p style="..."><br></p>' added by some web editor.
|
|
|
|
:param str html_content: html content, coming from example from an HTML field
|
|
:returns: bool, True if no content found or if containing only void formatting tags
|
|
"""
|
|
if not html_content:
|
|
return True
|
|
tag_re = re.compile(r'\<\s*\/?(?:p|div|span|br|b|i|font)(?:(?=\s+\w*)[^/>]*|\s*)/?\s*\>')
|
|
return not bool(re.sub(tag_re, '', html_content).strip())
|
|
|
|
def html_keep_url(text):
|
|
""" Transform the url into clickable link with <a/> tag """
|
|
idx = 0
|
|
final = ''
|
|
link_tags = re.compile(r"""(?<!["'])((ftp|http|https):\/\/(\w+:{0,1}\w*@)?([^\s<"']+)(:[0-9]+)?(\/|\/([^\s<"']))?)(?![^\s<"']*["']|[^\s<"']*</a>)""")
|
|
for item in re.finditer(link_tags, text):
|
|
final += text[idx:item.start()]
|
|
final += '<a href="%s" target="_blank" rel="noreferrer noopener">%s</a>' % (item.group(0), item.group(0))
|
|
idx = item.end()
|
|
final += text[idx:]
|
|
return final
|
|
|
|
|
|
def html_to_inner_content(html):
|
|
"""Returns unformatted text after removing html tags and excessive whitespace from a
|
|
string/Markup. Passed strings will first be sanitized.
|
|
"""
|
|
if is_html_empty(html):
|
|
return ''
|
|
if not isinstance(html, markupsafe.Markup):
|
|
html = html_sanitize(html)
|
|
processed = re.sub(HTML_NEWLINES_REGEX, ' ', html)
|
|
processed = re.sub(HTML_TAGS_REGEX, '', processed)
|
|
processed = re.sub(r' {2,}|\t', ' ', processed)
|
|
processed = processed.strip()
|
|
return processed
|
|
|
|
|
|
def html2plaintext(html, body_id=None, encoding='utf-8'):
|
|
""" From an HTML text, convert the HTML to plain text.
|
|
If @param body_id is provided then this is the tag where the
|
|
body (not necessarily <body>) starts.
|
|
"""
|
|
## (c) Fry-IT, www.fry-it.com, 2007
|
|
## <peter@fry-it.com>
|
|
## download here: http://www.peterbe.com/plog/html2plaintext
|
|
|
|
html = ustr(html)
|
|
|
|
if not html.strip():
|
|
return ''
|
|
|
|
tree = etree.fromstring(html, parser=etree.HTMLParser())
|
|
|
|
if body_id is not None:
|
|
source = tree.xpath('//*[@id=%s]' % (body_id,))
|
|
else:
|
|
source = tree.xpath('//body')
|
|
if len(source):
|
|
tree = source[0]
|
|
|
|
url_index = []
|
|
i = 0
|
|
for link in tree.findall('.//a'):
|
|
url = link.get('href')
|
|
if url:
|
|
i += 1
|
|
link.tag = 'span'
|
|
link.text = '%s [%s]' % (link.text, i)
|
|
url_index.append(url)
|
|
|
|
html = ustr(etree.tostring(tree, encoding=encoding))
|
|
# \r char is converted into , must remove it
|
|
html = html.replace(' ', '')
|
|
|
|
html = html.replace('<strong>', '*').replace('</strong>', '*')
|
|
html = html.replace('<b>', '*').replace('</b>', '*')
|
|
html = html.replace('<h3>', '*').replace('</h3>', '*')
|
|
html = html.replace('<h2>', '**').replace('</h2>', '**')
|
|
html = html.replace('<h1>', '**').replace('</h1>', '**')
|
|
html = html.replace('<em>', '/').replace('</em>', '/')
|
|
html = html.replace('<tr>', '\n')
|
|
html = html.replace('</p>', '\n')
|
|
html = re.sub('<br\s*/?>', '\n', html)
|
|
html = re.sub('<.*?>', ' ', html)
|
|
html = html.replace(' ' * 2, ' ')
|
|
html = html.replace('>', '>')
|
|
html = html.replace('<', '<')
|
|
html = html.replace('&', '&')
|
|
|
|
# strip all lines
|
|
html = '\n'.join([x.strip() for x in html.splitlines()])
|
|
html = html.replace('\n' * 2, '\n')
|
|
|
|
for i, url in enumerate(url_index):
|
|
if i == 0:
|
|
html += '\n\n'
|
|
html += ustr('[%s] %s\n') % (i + 1, url)
|
|
|
|
return html.strip()
|
|
|
|
def plaintext2html(text, container_tag=None):
|
|
r"""Convert plaintext into html. Content of the text is escaped to manage
|
|
html entities, using :func:`~odoo.tools.misc.html_escape`.
|
|
|
|
- all ``\n``, ``\r`` are replaced by ``<br/>``
|
|
- enclose content into ``<p>``
|
|
- convert url into clickable link
|
|
- 2 or more consecutive ``<br/>`` are considered as paragraph breaks
|
|
|
|
:param str text: plaintext to convert
|
|
:param str container_tag: container of the html; by default the content is
|
|
embedded into a ``<div>``
|
|
:rtype: markupsafe.Markup
|
|
"""
|
|
text = misc.html_escape(ustr(text))
|
|
|
|
# 1. replace \n and \r
|
|
text = re.sub(r'(\r\n|\r|\n)', '<br/>', text)
|
|
|
|
# 2. clickable links
|
|
text = html_keep_url(text)
|
|
|
|
# 3-4: form paragraphs
|
|
idx = 0
|
|
final = '<p>'
|
|
br_tags = re.compile(r'(([<]\s*[bB][rR]\s*/?[>]\s*){2,})')
|
|
for item in re.finditer(br_tags, text):
|
|
final += text[idx:item.start()] + '</p><p>'
|
|
idx = item.end()
|
|
final += text[idx:] + '</p>'
|
|
|
|
# 5. container
|
|
if container_tag: # FIXME: validate that container_tag is just a simple tag?
|
|
final = '<%s>%s</%s>' % (container_tag, final, container_tag)
|
|
return markupsafe.Markup(final)
|
|
|
|
def append_content_to_html(html, content, plaintext=True, preserve=False, container_tag=None):
|
|
""" Append extra content at the end of an HTML snippet, trying
|
|
to locate the end of the HTML document (</body>, </html>, or
|
|
EOF), and converting the provided content in html unless ``plaintext``
|
|
is False.
|
|
|
|
Content conversion can be done in two ways:
|
|
|
|
- wrapping it into a pre (``preserve=True``)
|
|
- use plaintext2html (``preserve=False``, using ``container_tag`` to
|
|
wrap the whole content)
|
|
|
|
A side-effect of this method is to coerce all HTML tags to
|
|
lowercase in ``html``, and strip enclosing <html> or <body> tags in
|
|
content if ``plaintext`` is False.
|
|
|
|
:param str html: html tagsoup (doesn't have to be XHTML)
|
|
:param str content: extra content to append
|
|
:param bool plaintext: whether content is plaintext and should
|
|
be wrapped in a <pre/> tag.
|
|
:param bool preserve: if content is plaintext, wrap it into a <pre>
|
|
instead of converting it into html
|
|
:param str container_tag: tag to wrap the content into, defaults to `div`.
|
|
:rtype: markupsafe.Markup
|
|
"""
|
|
html = ustr(html)
|
|
if plaintext and preserve:
|
|
content = u'\n<pre>%s</pre>\n' % misc.html_escape(ustr(content))
|
|
elif plaintext:
|
|
content = '\n%s\n' % plaintext2html(content, container_tag)
|
|
else:
|
|
content = re.sub(r'(?i)(</?(?:html|body|head|!\s*DOCTYPE)[^>]*>)', '', content)
|
|
content = u'\n%s\n' % ustr(content)
|
|
# Force all tags to lowercase
|
|
html = re.sub(r'(</?)(\w+)([ >])',
|
|
lambda m: '%s%s%s' % (m.group(1), m.group(2).lower(), m.group(3)), html)
|
|
insert_location = html.find('</body>')
|
|
if insert_location == -1:
|
|
insert_location = html.find('</html>')
|
|
if insert_location == -1:
|
|
return markupsafe.Markup('%s%s' % (html, content))
|
|
return markupsafe.Markup('%s%s%s' % (html[:insert_location], content, html[insert_location:]))
|
|
|
|
|
|
def prepend_html_content(html_body, html_content):
|
|
"""Prepend some HTML content at the beginning of an other HTML content."""
|
|
html_content = type(html_content)(re.sub(r'(?i)(</?(?:html|body|head|!\s*DOCTYPE)[^>]*>)', '', html_content))
|
|
html_content = html_content.strip()
|
|
|
|
body_match = re.search(r'<body[^>]*>', html_body) or re.search(r'<html[^>]*>', html_body)
|
|
insert_index = body_match.end() if body_match else 0
|
|
|
|
return html_body[:insert_index] + html_content + html_body[insert_index:]
|
|
|
|
#----------------------------------------------------------
|
|
# Emails
|
|
#----------------------------------------------------------
|
|
|
|
# matches any email in a body of text
|
|
email_re = re.compile(r"""([a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,63})""", re.VERBOSE)
|
|
|
|
# matches a string containing only one email
|
|
single_email_re = re.compile(r"""^[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-]+\.[a-zA-Z]{2,63}$""", re.VERBOSE)
|
|
|
|
mail_header_msgid_re = re.compile('<[^<>]+>')
|
|
|
|
email_addr_escapes_re = re.compile(r'[\\"]')
|
|
|
|
|
|
def generate_tracking_message_id(res_id):
|
|
"""Returns a string that can be used in the Message-ID RFC822 header field
|
|
|
|
Used to track the replies related to a given object thanks to the "In-Reply-To"
|
|
or "References" fields that Mail User Agents will set.
|
|
"""
|
|
try:
|
|
rnd = random.SystemRandom().random()
|
|
except NotImplementedError:
|
|
rnd = random.random()
|
|
rndstr = ("%.15f" % rnd)[2:]
|
|
return "<%s.%.15f-openerp-%s@%s>" % (rndstr, time.time(), res_id, socket.gethostname())
|
|
|
|
def email_split_tuples(text):
|
|
""" Return a list of (name, email) address tuples found in ``text`` . Note
|
|
that text should be an email header or a stringified email list as it may
|
|
give broader results than expected on actual text. """
|
|
if not text:
|
|
return []
|
|
return [(addr[0], addr[1]) for addr in getaddresses([text])
|
|
# getaddresses() returns '' when email parsing fails, and
|
|
# sometimes returns emails without at least '@'. The '@'
|
|
# is strictly required in RFC2822's `addr-spec`.
|
|
if addr[1]
|
|
if '@' in addr[1]]
|
|
|
|
def email_split(text):
|
|
""" Return a list of the email addresses found in ``text`` """
|
|
if not text:
|
|
return []
|
|
return [email for (name, email) in email_split_tuples(text)]
|
|
|
|
def email_split_and_format(text):
|
|
""" Return a list of email addresses found in ``text``, formatted using
|
|
formataddr. """
|
|
if not text:
|
|
return []
|
|
return [formataddr((name, email)) for (name, email) in email_split_tuples(text)]
|
|
|
|
def email_normalize(text, strict=True):
|
|
""" Sanitize and standardize email address entries.
|
|
A normalized email is considered as :
|
|
- having a left part + @ + a right part (the domain can be without '.something')
|
|
- being lower case
|
|
- having no name before the address. Typically, having no 'Name <>'
|
|
Ex:
|
|
- Possible Input Email : 'Name <NaMe@DoMaIn.CoM>'
|
|
- Normalized Output Email : 'name@domain.com'
|
|
|
|
:param bool strict: text should contain exactly one email (default behavior
|
|
and unique behavior before Odoo16);
|
|
|
|
:return: False if no email found (or if more than 1 email found when being
|
|
in strict mode); normalized email otherwise;
|
|
"""
|
|
emails = email_split(text)
|
|
if not emails or (strict and len(emails) != 1):
|
|
return False
|
|
return emails[0].lower()
|
|
|
|
def email_domain_extract(email):
|
|
""" Extract the company domain to be used by IAP services notably. Domain
|
|
is extracted from email information e.g:
|
|
|
|
- info@proximus.be -> proximus.be
|
|
"""
|
|
normalized_email = email_normalize(email)
|
|
if normalized_email:
|
|
return normalized_email.split('@')[1]
|
|
return False
|
|
|
|
def email_domain_normalize(domain):
|
|
"""Return the domain normalized or False if the domain is invalid."""
|
|
if not domain or '@' in domain:
|
|
return False
|
|
|
|
return domain.lower()
|
|
|
|
def url_domain_extract(url):
|
|
""" Extract the company domain to be used by IAP services notably. Domain
|
|
is extracted from an URL e.g:
|
|
|
|
- www.info.proximus.be -> proximus.be
|
|
"""
|
|
parser_results = urlparse(url)
|
|
company_hostname = parser_results.hostname
|
|
if company_hostname and '.' in company_hostname:
|
|
return '.'.join(company_hostname.split('.')[-2:]) # remove subdomains
|
|
return False
|
|
|
|
def email_escape_char(email_address):
|
|
""" Escape problematic characters in the given email address string"""
|
|
return email_address.replace('\\', '\\\\').replace('%', '\\%').replace('_', '\\_')
|
|
|
|
# was mail_thread.decode_header()
|
|
def decode_message_header(message, header, separator=' '):
|
|
return separator.join(h for h in message.get_all(header, []) if h)
|
|
|
|
def formataddr(pair, charset='utf-8'):
|
|
"""Pretty format a 2-tuple of the form (realname, email_address).
|
|
|
|
If the first element of pair is falsy then only the email address
|
|
is returned.
|
|
|
|
Set the charset to ascii to get a RFC-2822 compliant email. The
|
|
realname will be base64 encoded (if necessary) and the domain part
|
|
of the email will be punycode encoded (if necessary). The local part
|
|
is left unchanged thus require the SMTPUTF8 extension when there are
|
|
non-ascii characters.
|
|
|
|
>>> formataddr(('John Doe', 'johndoe@example.com'))
|
|
'"John Doe" <johndoe@example.com>'
|
|
|
|
>>> formataddr(('', 'johndoe@example.com'))
|
|
'johndoe@example.com'
|
|
"""
|
|
name, address = pair
|
|
local, _, domain = address.rpartition('@')
|
|
|
|
try:
|
|
domain.encode(charset)
|
|
except UnicodeEncodeError:
|
|
# rfc5890 - Internationalized Domain Names for Applications (IDNA)
|
|
domain = idna.encode(domain).decode('ascii')
|
|
|
|
if name:
|
|
try:
|
|
name.encode(charset)
|
|
except UnicodeEncodeError:
|
|
# charset mismatch, encode as utf-8/base64
|
|
# rfc2047 - MIME Message Header Extensions for Non-ASCII Text
|
|
name = base64.b64encode(name.encode('utf-8')).decode('ascii')
|
|
return f"=?utf-8?b?{name}?= <{local}@{domain}>"
|
|
else:
|
|
# ascii name, escape it if needed
|
|
# rfc2822 - Internet Message Format
|
|
# #section-3.4 - Address Specification
|
|
name = email_addr_escapes_re.sub(r'\\\g<0>', name)
|
|
return f'"{name}" <{local}@{domain}>'
|
|
return f"{local}@{domain}"
|
|
|
|
|
|
def encapsulate_email(old_email, new_email):
|
|
"""Change the FROM of the message and use the old one as name.
|
|
|
|
e.g.
|
|
* Old From: "Admin" <admin@gmail.com>
|
|
* New From: notifications@odoo.com
|
|
* Output: "Admin" <notifications@odoo.com>
|
|
"""
|
|
old_email_split = getaddresses([old_email])
|
|
if not old_email_split or not old_email_split[0]:
|
|
return old_email
|
|
|
|
new_email_split = getaddresses([new_email])
|
|
if not new_email_split or not new_email_split[0]:
|
|
return
|
|
|
|
old_name, old_email = old_email_split[0]
|
|
if old_name:
|
|
name_part = old_name
|
|
else:
|
|
name_part = old_email.split("@")[0]
|
|
|
|
return formataddr((
|
|
name_part,
|
|
new_email_split[0][1],
|
|
))
|