def correct(source, target):
"""Runs a set of easy and automatic corrections
Current corrections include:
- Ellipses - align target to use source form of ellipses (either three dots or the Unicode ellipses characters)
- Missing whitespace and start or end of the target
- Missing punction (.:?) at the end of the target
"""
if target == "":
return target
if "..." in source and u"…" in target:
return target.replace(u"…", "...")
if u"…" in source and "..." in target:
return target.replace("...", u"…")
if decoration.spacestart(source) != decoration.spacestart(target) or decoration.spaceend(
source
) != decoration.spaceend(target):
return decoration.spacestart(source) + target.strip() + decoration.spaceend(source)
punctuation = (".", ":", ". ", ": ", "?")
puncendid = decoration.puncend(source, punctuation)
puncendstr = decoration.puncend(target, punctuation)
if puncendid != puncendstr:
if not puncendstr:
return target + puncendid
if source[:1].isalpha() and target[:1].isalpha():
if source[:1].isupper() and target[:1].islower():
return target[:1].upper() + target[1:]
elif source[:1].islower() and target[:1].isupper():
return target[:1].lower() + target[1:]
return None
def correct(source, target):
"""Runs a set of easy and automatic corrections
Current corrections include:
- Ellipses - align target to use source form of ellipses (either three dots or the Unicode ellipses characters)
- Missing whitespace and start or end of the target
- Missing punction (.:?) at the end of the target
"""
if target == "":
return target
if "..." in source and u"…" in target:
return target.replace(u"…", "...")
if u"…" in source and "..." in target:
return target.replace("...", u"…")
if decoration.spacestart(source) != decoration.spacestart(
target) or decoration.spaceend(source) != decoration.spaceend(
target):
return decoration.spacestart(
source) + target.strip() + decoration.spaceend(source)
punctuation = (".", ":", ". ", ": ", "?")
puncendid = decoration.puncend(source, punctuation)
puncendstr = decoration.puncend(target, punctuation)
if puncendid != puncendstr:
if not puncendstr:
return target + puncendid
if source[:1].isalpha() and target[:1].isalpha():
if source[:1].isupper() and target[:1].islower():
return target[:1].upper() + target[1:]
elif source[:1].islower() and target[:1].isupper():
return target[:1].lower() + target[1:]
return None
def test_spacestart():
"""test operation of spacestart()"""
assert decoration.spacestart(" Start") == " "
assert decoration.spacestart(u"\u0020\u00a0Start") == u"\u0020\u00a0"
# non-breaking space
assert decoration.spacestart(u"\u00a0\u202fStart") == u"\u00a0\u202f"
# Some exotic spaces
assert decoration.spacestart(u"\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200aStart") == u"\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200a"
def test_spacestart():
"""test operation of spacestart()"""
assert decoration.spacestart(" Start") == " "
assert decoration.spacestart(u"\u0020\u00a0Start") == u"\u0020\u00a0"
# non-breaking space
assert decoration.spacestart(u"\u00a0\u202fStart") == u"\u00a0\u202f"
# Some exotic spaces
assert decoration.spacestart(u"\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200aStart") == u"\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200a"
