def __init__(self, uid, basename, logger):
self.logger = logger
self.uid = uid
self.basename = basename
if Char.isdefined(uid):
self.general = Char.charType(uid)
self.cc = Char.getCombiningClass(uid)
self.icuGC = Char.charType(uid)
self.icuJT = Char.getIntPropertyValue(uid, UProperty.JOINING_TYPE)
self.icuJG = Char.getIntPropertyValue(uid, UProperty.JOINING_GROUP)
else:
self.logger.log('USV %04X not in ICU; no properties known' % uid,
'W')
self.feats = set() # feat tags that affect this char
self.langs = set() # lang tags that affect this char
# Additional info from UFO:
self.takesMarks = self.isMark = self.isBase = False
def is_exemplar_wordbreak(char):
"""True if the character has the Word_Break properties Katakana, ALetter, or MidLetter."""
# The following should be exposed by PyICU, but does not seem to be implemented.
# There are other values, but these are the ones need for this function.
WB_ALETTER = 1
WB_KATAKANA = 3
WB_MIDLETTER = 4
numeric_wordbreak_type = Char.getIntPropertyValue(char, UProperty.WORD_BREAK)
if (numeric_wordbreak_type == WB_KATAKANA or
numeric_wordbreak_type == WB_ALETTER or
numeric_wordbreak_type == WB_MIDLETTER):
return True
return False
def is_exemplar_wordbreak(char):
"""True if the character has the Word_Break properties Katakana, ALetter, or MidLetter."""
# The following should be exposed by PyICU, but does not seem to be implemented.
# There are other values, but these are the ones need for this function.
WB_ALETTER = 1
WB_KATAKANA = 3
# WB_MIDLETTER = 4
numeric_wordbreak_type = Char.getIntPropertyValue(char, UProperty.WORD_BREAK)
if (numeric_wordbreak_type == WB_KATAKANA or
# numeric_wordbreak_type == WB_MIDLETTER or
numeric_wordbreak_type == WB_ALETTER):
return True
return False
def render(self, uids, ftml, keyUID=0, addBreaks=True, rtl=None):
""" general purpose (but not required) function to generate ftml for a character sequence """
if len(uids) == 0:
return
# Make a copy so we don't affect caller
uids = list(uids)
# Remember first uid and original length for later
startUID = uids[0]
uidLen = len(uids)
# if keyUID wasn't supplied, use startUID
if keyUID == 0: keyUID = startUID
# Construct label from uids:
label = '\n'.join(['U+{0:04X}'.format(u) for u in uids])
# Construct comment from glyph names:
comment = ' '.join([self._charFromUID[u].basename for u in uids])
# see if uid list includes a mirrored char
hasMirrored = bool(len([x for x in uids if Char.isMirrored(x)]))
# Analyze first and last joining char
joiningChars = [
x for x in uids if
Char.getIntPropertyValue(x, UProperty.JOINING_TYPE) != TRANSPARENT
]
if len(joiningChars):
# If first or last non-TRANSPARENT char is a joining char, then we need to emit examples with zwj
uid = joiningChars[0]
zwjBefore = Char.getIntPropertyValue(
uid, UProperty.JOINING_TYPE) == DUAL_JOINING or (
Char.charDirection(uid) == UCharDirection.LEFT_TO_RIGHT
and Char.getIntPropertyValue(uid, UProperty.JOINING_TYPE)
== LEFT_JOINING) or (
Char.charDirection(uid) != UCharDirection.LEFT_TO_RIGHT
and Char.getIntPropertyValue(
uid, UProperty.JOINING_TYPE) == RIGHT_JOINING)
uid = joiningChars[-1]
zwjAfter = Char.getIntPropertyValue(
uid, UProperty.JOINING_TYPE) == DUAL_JOINING or (
Char.charDirection(uid) == UCharDirection.LEFT_TO_RIGHT
and Char.getIntPropertyValue(uid, UProperty.JOINING_TYPE)
== RIGHT_JOINING) or (
Char.charDirection(uid) != UCharDirection.LEFT_TO_RIGHT
and Char.getIntPropertyValue(
uid, UProperty.JOINING_TYPE) == LEFT_JOINING)
else:
zwjBefore = zwjAfter = False
if Char.charType(startUID) == UCharCategory.NON_SPACING_MARK:
# First char is a NSM... prefix a suitable base
uids.insert(0, self.diacBase)
zwjBefore = False # No longer any need to put zwj before
elif Char.isUWhiteSpace(startUID):
# First char is whitespace -- prefix with baseline brackets:
uids.insert(0, 0xF130)
lastNonMark = [
x for x in uids
if Char.charType(x) != UCharCategory.NON_SPACING_MARK
][-1]
if Char.isUWhiteSpace(lastNonMark):
# Last non-mark is whitespace -- append baseline brackets:
uids.append(0xF131)
s = ''.join([chr(uid) for uid in uids])
if zwjBefore or zwjAfter:
# Show contextual forms:
t = u'{0} '.format(s)
if zwjAfter:
t += u'{0}\u200D '.format(s)
if zwjBefore:
t += u'\u200D{0}\u200D '.format(s)
if zwjBefore:
t += u'\u200D{0} '.format(s)
if zwjBefore and zwjAfter:
t += u'{0}{0}{0}'.format(s)
if addBreaks: ftml.closeTest()
ftml.addToTest(keyUID, t, label=label, comment=comment, rtl=rtl)
if addBreaks: ftml.closeTest()
elif hasMirrored and self.rtlEnable:
# Contains mirrored and rtl enabled:
if addBreaks: ftml.closeTest()
ftml.addToTest(
keyUID,
u'{0} LTR: \u202A{0}\u202C RTL: \u202B{0}\u202C'.format(s),
label=label,
comment=comment,
rtl=rtl)
if addBreaks: ftml.closeTest()
# elif is LRE, RLE, PDF
# elif is LRI, RLI, FSI, PDI
elif uidLen > 1:
ftml.addToTest(keyUID, s, label=label, comment=comment, rtl=rtl)
else:
ftml.addToTest(keyUID, s, comment=comment, rtl=rtl)