xenofem/python-zulip-api
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Wiki Activity

zephyr_mirror: Improve different-paragraph detection algorithm.

Browse source 
(imported from commit 789651dd5b9af58bbc0e59a64dd565f06585134d)
This commit is contained in:
Tim Abbott 2012-11-05 15:39:09 -05:00
parent 403b728d3c
commit 77bb41b3dd
1 changed files with 12 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
bots/zephyr_mirror.py
View file

@ -96,9 +96,18 @@ def to_zephyr_username(humbug_username):
raise Exception("Could not parse Zephyr realm for cross-realm user %s" % (humbug_username,)) raise Exception("Could not parse Zephyr realm for cross-realm user %s" % (humbug_username,))
return match_user.group(1).lower() + "@" + match_user.group(2).upper() return match_user.group(1).lower() + "@" + match_user.group(2).upper()
def early_indent(line, next_line): # Checks whether the pair of adjacent lines would have been
# linewrapped together, had they been intended to be parts of the same
# paragraph. Our check is whether if you move the first word on the
# 2nd line onto the first line, the resulting line is either (1)
# significantly shorter than the following line (which, if they were
# in the same paragraph, should have been wrapped in a way consistent
# with how the previous line was wrapped) or (2) shorter than 60
# characters (our assumed minimum linewrapping threshhold for Zephyr)
def different_paragraph(line, next_line):
words = next_line.split() words = next_line.split()
return len(line + " " + words[0]) < len(next_line) * 0.8 return (len(line + " " + words[0]) < len(next_line) * 0.8 or
len(line + " " + words[0]) < 60)
# Linewrapping algorithm based on: # Linewrapping algorithm based on:
# http://gcbenison.wordpress.com/2011/07/03/a-program-to-intelligently-remove-carriage-returns-so-you-can-paste-text-without-having-it-look-awful/ # http://gcbenison.wordpress.com/2011/07/03/a-program-to-intelligently-remove-carriage-returns-so-you-can-paste-text-without-having-it-look-awful/
@ -111,7 +120,7 @@ def unwrap_lines(body):
if (line == "" or if (line == "" or
previous_line == "" or previous_line == "" or
not re.match(r'^[\w]', line, flags=re.UNICODE) or not re.match(r'^[\w]', line, flags=re.UNICODE) or
early_indent(previous_line, line)): different_paragraph(previous_line, line)):
# Use 2 newlines to separate sections so that we # Use 2 newlines to separate sections so that we
# trigger proper Markdown processing on things like # trigger proper Markdown processing on things like
# bulleted lists # bulleted lists