def iter_threads(self):
table = self.parser.select(self.document.getroot(), "table#listeMessages", 1)
for tr in table.xpath("./tr"):
if tr.attrib.get("class", "") not in ("msgLu", "msgNonLu"):
continue
author = unicode(self.parser.select(tr, "td.colEmetteur", 1).text)
link = self.parser.select(tr, "td.colObjet a", 1)
date_raw = self.parser.select(tr, "td.colDate1", 1).attrib["data"]
jsparams = re.search("\((.+)\)", link.attrib["onclick"]).groups()[0]
jsparams = [i.strip("'\" ") for i in jsparams.split(",")]
page_id, _id, unread = jsparams
# this means unread on the website
unread = False if unread == "false" else True
# 2012/02/29:01h30min45sec
dt_match = re.match("(\d+)/(\d+)/(\d+):(\d+)h(\d+)min(\d+)sec", date_raw).groups()
dt_match = [int(d) for d in dt_match]
thread = Thread(_id)
thread._link_id = (page_id, unread)
thread.date = datetime(*dt_match)
thread.title = unicode(link.text)
message = Message(thread, 0)
message.set_empty_fields(None)
message.flags = message.IS_HTML
message.title = thread.title
message.date = thread.date
message.sender = author
message.content = NotLoaded # This is the only thing we are missing
thread.root = message
yield thread
def iter_threads(self):
table = self.parser.select(self.document.getroot(),
'table#listeMessages', 1)
for tr in table.xpath('./tr'):
if tr.attrib.get('class', '') not in ('msgLu', 'msgNonLu'):
continue
author = unicode(self.parser.select(tr, 'td.colEmetteur', 1).text)
link = self.parser.select(tr, 'td.colObjet a', 1)
date_raw = self.parser.select(tr, 'td.colDate1', 1).attrib['data']
jsparams = re.search('\((.+)\)',
link.attrib['onclick']).groups()[0]
jsparams = [i.strip('\'" ') for i in jsparams.split(',')]
page_id, _id, unread = jsparams
# this means unread on the website
unread = False if unread == "false" else True
# 2012/02/29:01h30min45sec
dt_match = re.match('(\d+)/(\d+)/(\d+):(\d+)h(\d+)min(\d+)sec',
date_raw).groups()
dt_match = [int(d) for d in dt_match]
thread = Thread(_id)
thread._link_id = (page_id, unread)
thread.date = datetime(*dt_match)
thread.title = unicode(link.text)
message = Message(thread, 0)
message.set_empty_fields(None)
message.flags = message.IS_HTML
message.title = thread.title
message.date = thread.date
message.sender = author
message.content = NotLoaded # This is the only thing we are missing
thread.root = message
yield thread
def build_date(self, k):
ret = Message(id='%s.%s' % k)
ret.title = '%s/%s' % k
ret.content = ''
ret.date = date(int(k[0]), int(k[1]), 1)
ret._type = 'date'
ret._key = k
return ret
def _get_messages(self, thread):
thread_div = self.document.find(True, 'PADpost_txt')
used_ids = set()
rcpt = self.document.find('input', attrs={'type': 'hidden', 'name': 'Dest'})['value']
sender_to_receiver = {rcpt: self.browser.username, self.browser.username: rcpt}
# site is sorted from latest to oldest message
for message_table in reversed(thread_div.findAll('table')):
for td in message_table.findAll('td'):
profile_a = td.find('a', href=re.compile(r'profil_read.php\?.*'))
if not profile_a:
continue
first_br = td.find('br')
assert first_br.nextSibling.name == 'br'
text_nodes = ovsparse.all_next_siblings(first_br.nextSibling.nextSibling) # TODO
#~ print text_nodes
# date will be used as id
sitedate = profile_a.findParent('div').find(text=re.compile(',.*')).replace(', ', '')
sysdate = parse_french_date(sitedate)
compactdate = datetime.datetime.strftime(sysdate, '%Y%m%dT%H%M%S')
# but make it unique
msg_id = ovsparse.create_unique_id(compactdate, used_ids)
used_ids.add(msg_id)
message = Message(thread, msg_id)
message.sender = re.search(r'\?(.+)', profile_a['href']).group(1)
message.receivers = [sender_to_receiver[message.sender]]
message.date = sysdate
message.content = ovsparse.html_message_to_text(text_nodes)
notread_self = bool(td.find('span', 'ColorSurligne'))
notread_other = bool(td.find('span', 'new_sortiepartenaire'))
if notread_other or notread_self:
message.flags |= Message.IS_NOT_RECEIVED
else:
message.flags |= Message.IS_RECEIVED
yield message
def _get_messages(self, thread):
thread_div = self.document.find(True, "PADpost_txt")
used_ids = set()
rcpt = self.document.find("input", attrs={"type": "hidden", "name": "Dest"})["value"]
sender_to_receiver = {rcpt: self.browser.username, self.browser.username: rcpt}
# site is sorted from latest to oldest message
for message_table in reversed(thread_div.findAll("table")):
for td in message_table.findAll("td"):
profile_a = td.find("a", href=re.compile(r"profil_read.php\?.*"))
if not profile_a:
continue
first_br = td.find("br")
assert first_br.nextSibling.name == "br"
text_nodes = ovsparse.all_next_siblings(first_br.nextSibling.nextSibling) # TODO
# ~ print text_nodes
# date will be used as id
sitedate = profile_a.findParent("div").find(text=re.compile(",.*")).replace(", ", "")
sysdate = parse_french_date(sitedate)
compactdate = datetime.datetime.strftime(sysdate, "%Y%m%dT%H%M%S")
# but make it unique
msg_id = ovsparse.create_unique_id(compactdate, used_ids)
used_ids.add(msg_id)
message = Message(thread, msg_id)
message.sender = re.search(r"\?(.+)", profile_a["href"]).group(1)
message.receivers = [sender_to_receiver[message.sender]]
message.date = sysdate
message.content = ovsparse.html_message_to_text(text_nodes)
notread_self = bool(td.find("span", "ColorSurligne"))
notread_other = bool(td.find("span", "new_sortiepartenaire"))
if notread_other or notread_self:
message.flags |= Message.IS_NOT_RECEIVED
else:
message.flags |= Message.IS_RECEIVED
yield message
