Implement better handling of corner cases
This commit is contained in:
parent
54e5e7d0e4
commit
d19c02228e
1 changed files with 17 additions and 4 deletions
|
@ -1,5 +1,6 @@
|
||||||
import re
|
import re
|
||||||
import enum
|
import enum
|
||||||
|
import datetime
|
||||||
|
|
||||||
from dateparser.search import search_dates
|
from dateparser.search import search_dates
|
||||||
from email.header import decode_header
|
from email.header import decode_header
|
||||||
|
@ -44,7 +45,7 @@ class MBoxReaderBuffer():
|
||||||
return self.lines[line][0:5] == 'From '
|
return self.lines[line][0:5] == 'From '
|
||||||
|
|
||||||
def is_header_line(self, line):
|
def is_header_line(self, line):
|
||||||
return re.match('^([^:]+): (.*)$', self.lines[line]) is not None
|
return re.match('^([A-Za-z0-9\\-]+): (.*)$', self.lines[line]) is not None
|
||||||
|
|
||||||
def is_start(self):
|
def is_start(self):
|
||||||
if self.lines[0] is None or self.lines[1] is None:
|
if self.lines[0] is None or self.lines[1] is None:
|
||||||
|
@ -67,9 +68,12 @@ def decode(text: str):
|
||||||
decoded = decode_header(text)[0]
|
decoded = decode_header(text)[0]
|
||||||
|
|
||||||
if decoded[1] is None:
|
if decoded[1] is None:
|
||||||
return decoded[0]
|
return str(decoded[0])
|
||||||
|
|
||||||
|
try:
|
||||||
return str(decoded[0], decoded[1])
|
return str(decoded[0], decoded[1])
|
||||||
|
except:
|
||||||
|
return str(decoded[0])
|
||||||
|
|
||||||
class MBoxMessage():
|
class MBoxMessage():
|
||||||
__slots__ = 'state', 'headers', 'line', 'content', 'body', 'key',
|
__slots__ = 'state', 'headers', 'line', 'content', 'body', 'key',
|
||||||
|
@ -113,7 +117,16 @@ class MBoxMessage():
|
||||||
return self.headers.get(key.lower())
|
return self.headers.get(key.lower())
|
||||||
|
|
||||||
def date(self):
|
def date(self):
|
||||||
|
try:
|
||||||
return search_dates(self.headers['date'])[0][1]
|
return search_dates(self.headers['date'])[0][1]
|
||||||
|
except:
|
||||||
|
return datetime.datetime.fromtimestamp(0)
|
||||||
|
|
||||||
|
def sender(self):
|
||||||
|
return self.headers.get('from', 'Unknown')
|
||||||
|
|
||||||
|
def subject(self):
|
||||||
|
return self.headers.get('subject', '(no subject)')
|
||||||
|
|
||||||
def is_first_line(self):
|
def is_first_line(self):
|
||||||
return len(self.headers) == 1 and (self.body == '' or self.body is None)
|
return len(self.headers) == 1 and (self.body == '' or self.body is None)
|
||||||
|
|
Loading…
Add table
Reference in a new issue