Add more fuzzy trailer matching logic

Will now also accept trailers if:

- there's a full match for local part and partial match for domain part
of the email in from/trailer
- we find a comma in the name and match the trailer after rearranging
it

Signed-off-by: Konstantin Ryabitsev <konstantin@linuxfoundation.org>

1 files changed, 32 insertions, 2 deletions

diff --git a/get-lore-mbox.py b/get-lore-mbox.py
index 5160fb5..58244bc 100755
--- a/get-lore-mbox.py
+++ b/get-lore-mbox.py
@@ -30,7 +30,7 @@ charset.add_charset('utf-8', None)
 emlpolicy = email.policy.EmailPolicy(utf8=True, cte_type='8bit', max_line_length=None)
 logger = logging.getLogger('get-lore-mbox')
 
-VERSION = '0.2.14'
+VERSION = '0.2.15'
 
 # You can use bash-style globbing here
 WANTHDRS = [
@@ -431,10 +431,40 @@ class LoreMessage:
                 # Basic sanity checking -- the trailer must match the name or the email
                 # in the From header, to avoid false-positive trailer parsing errors
                 for tname, tvalue in matches:
+                    tmatch = False
                     namedata = email.utils.getaddresses([tvalue])[0]
                     tfrom = re.sub(r'\+[^@]+@', '@', namedata[1].lower())
                     hfrom = re.sub(r'\+[^@]+@', '@', self.fromemail.lower())
-                    if tfrom == hfrom or namedata[0].lower() == self.fromname.lower():
+                    tlname = namedata[0].lower()
+                    hlname = self.fromname.lower()
+                    tchunks = tfrom.split('@')
+                    hchunks = hfrom.split('@')
+                    if tfrom == hfrom:
+                        logger.debug('  trailer exact email match')
+                        tmatch = True
+                    # See if domain part of one of the addresses is a subset of the other one,
+                    # which should match cases like @linux.intel.com and @intel.com
+                    elif (len(tchunks) == 2 and len(hchunks) == 2
+                            and tchunks[0] == hchunks[0]
+                            and (tchunks[1].find(hchunks[1]) >= 0 or hchunks[1].find(tchunks[1]) >= 0)):
+                        logger.debug('  trailer fuzzy email match')
+                        tmatch = True
+                    # Does the name match, at least?
+                    elif tlname == hlname:
+                        logger.debug('  trailer exact name match')
+                        tmatch = True
+                    # Finally, see if the header From has a comma in it and try to find all
+                    # parts in the trailer name
+                    elif hlname.find(',') > 0:
+                        nmatch = True
+                        for nchunk in hlname.split(','):
+                            if hlname.find(nchunk.strip()) < 0:
+                                nmatch = False
+                                break
+                        if nmatch:
+                            logger.debug('  trailer fuzzy name match')
+                            tmatch = True
+                    if tmatch:
                         self.trailers.append((tname, tvalue))
                     else:
                         logger.debug('  ignoring "%s: %s" due to from mismatch (from: %s %s)', tname, tvalue,