def getNumberSystems(cache={}):
"""Cached look-up of number system information.
Pass no arguments. Returns a mapping from number system names to,
for each system, a mapping with keys u'digits', u'type' and
u'id'\n"""
if not cache:
for ns in findTagsInFile(os.path.join(cldr_dir, '..', 'supplemental',
'numberingSystems.xml'),
'numberingSystems'):
# ns has form: [u'numberingSystem', [(u'digits', u'0123456789'), (u'type', u'numeric'), (u'id', u'latn')]]
entry = dict(ns[1])
cache[entry[u'id']] = entry
return cache
def getNumberSystems(cache={}):
"""Cached look-up of number system information.
Pass no arguments. Returns a mapping from number system names to,
for each system, a mapping with keys u'digits', u'type' and
u'id'\n"""
if not cache:
for ns in findTagsInFile(os.path.join(cldr_dir, '..', 'supplemental',
'numberingSystems.xml'),
'numberingSystems'):
# ns has form: [u'numberingSystem', [(u'digits', u'0123456789'), (u'type', u'numeric'), (u'id', u'latn')]]
entry = dict(ns[1])
name = entry[u'id']
if u'digits' in entry and ord(entry[u'digits'][0]) > 0xffff:
# FIXME, QTBUG-69324: make this redundant:
# omit number system if zero doesn't fit in single-char16 UTF-16 :-(
sys.stderr.write('skipping number system "%s" [can\'t represent its zero, U+%X]\n'
% (name, ord(entry[u'digits'][0])))
else:
cache[name] = entry
return cache
def from_supplement(tag,
path=os.path.join(dir_name, '..', 'supplemental',
'supplementalData.xml')):
return findTagsInFile(path, tag)
def _generateLocaleInfo(path, language_code, script_code, country_code, variant_code=""):
if not path.endswith(".xml"):
return {}
if language_code == 'root':
# just skip it
return {}
# we do not support variants
# ### actually there is only one locale with variant: en_US_POSIX
# does anybody care about it at all?
if variant_code:
raise xpathlite.Error('we do not support variants ("%s")' % variant_code)
language_id = enumdata.languageCodeToId(language_code)
if language_id <= 0:
raise xpathlite.Error('unknown language code "%s"' % language_code)
script_id = enumdata.scriptCodeToId(script_code)
if script_id == -1:
raise xpathlite.Error('unknown script code "%s"' % script_code)
# we should handle fully qualified names with the territory
if not country_code:
return {}
country_id = enumdata.countryCodeToId(country_code)
if country_id <= 0:
raise xpathlite.Error('unknown country code "%s"' % country_code)
# So we say we accept only those values that have "contributed" or
# "approved" resolution. see http://www.unicode.org/cldr/process.html
# But we only respect the resolution for new datas for backward
# compatibility.
draft = DraftResolution.contributed
result = dict(
language=enumdata.language_list[language_id][0],
language_code=language_code, language_id=language_id,
script=enumdata.script_list[script_id][0],
script_code=script_code, script_id=script_id,
country=enumdata.country_list[country_id][0],
country_code=country_code, country_id=country_id,
variant_code=variant_code)
(dir_name, file_name) = os.path.split(path)
def from_supplement(tag,
path=os.path.join(dir_name, '..', 'supplemental',
'supplementalData.xml')):
return findTagsInFile(path, tag)
currencies = from_supplement('currencyData/region[iso3166=%s]' % country_code)
result['currencyIsoCode'] = ''
result['currencyDigits'] = 2
result['currencyRounding'] = 1
if currencies:
for e in currencies:
if e[0] == 'currency':
t = [x[1] == 'false' for x in e[1] if x[0] == 'tender']
if t and t[0]:
pass
elif not any(x[0] == 'to' for x in e[1]):
result['currencyIsoCode'] = (x[1] for x in e[1] if x[0] == 'iso4217').next()
break
if result['currencyIsoCode']:
t = from_supplement("currencyData/fractions/info[iso4217=%s]"
% result['currencyIsoCode'])
if t and t[0][0] == 'info':
result['currencyDigits'] = (int(x[1]) for x in t[0][1] if x[0] == 'digits').next()
result['currencyRounding'] = (int(x[1]) for x in t[0][1] if x[0] == 'rounding').next()
numbering_system = None
try:
numbering_system = findEntry(path, "numbers/defaultNumberingSystem")
except:
pass
def findEntryDef(path, xpath, value=''):
try:
return findEntry(path, xpath)
except xpathlite.Error:
return value
def get_number_in_system(path, xpath, numbering_system):
if numbering_system:
try:
return findEntry(path, xpath + "[numberSystem=" + numbering_system + "]")
except xpathlite.Error:
# in CLDR 1.9 number system was refactored for numbers (but not for currency)
# so if previous findEntry doesn't work we should try this:
try:
return findEntry(path, xpath.replace("/symbols/", "/symbols[numberSystem=" + numbering_system + "]/"))
except xpathlite.Error:
# fallback to default
pass
return findEntry(path, xpath)
result['decimal'] = get_number_in_system(path, "numbers/symbols/decimal", numbering_system)
result['group'] = get_number_in_system(path, "numbers/symbols/group", numbering_system)
result['list'] = get_number_in_system(path, "numbers/symbols/list", numbering_system)
result['percent'] = get_number_in_system(path, "numbers/symbols/percentSign", numbering_system)
try:
numbering_systems = {}
for ns in findTagsInFile(os.path.join(cldr_dir, '..', 'supplemental',
'numberingSystems.xml'),
'numberingSystems'):
tmp = {}
id = ""
for data in ns[1:][0]: # ns looks like this: [u'numberingSystem', [(u'digits', u'0123456789'), (u'type', u'numeric'), (u'id', u'latn')]]
tmp[data[0]] = data[1]
if data[0] == u"id":
id = data[1]
numbering_systems[id] = tmp
result['zero'] = numbering_systems[numbering_system][u"digits"][0]
except e:
sys.stderr.write("Native zero detection problem:\n" + str(e) + "\n")
result['zero'] = get_number_in_system(path, "numbers/symbols/nativeZeroDigit", numbering_system)
result['minus'] = get_number_in_system(path, "numbers/symbols/minusSign", numbering_system)
result['plus'] = get_number_in_system(path, "numbers/symbols/plusSign", numbering_system)
result['exp'] = get_number_in_system(path, "numbers/symbols/exponential", numbering_system).lower()
result['quotationStart'] = findEntry(path, "delimiters/quotationStart")
result['quotationEnd'] = findEntry(path, "delimiters/quotationEnd")
result['alternateQuotationStart'] = findEntry(path, "delimiters/alternateQuotationStart")
result['alternateQuotationEnd'] = findEntry(path, "delimiters/alternateQuotationEnd")
result['listPatternPartStart'] = parse_list_pattern_part_format(findEntry(path, "listPatterns/listPattern/listPatternPart[start]"))
result['listPatternPartMiddle'] = parse_list_pattern_part_format(findEntry(path, "listPatterns/listPattern/listPatternPart[middle]"))
result['listPatternPartEnd'] = parse_list_pattern_part_format(findEntry(path, "listPatterns/listPattern/listPatternPart[end]"))
result['listPatternPartTwo'] = parse_list_pattern_part_format(findEntry(path, "listPatterns/listPattern/listPatternPart[2]"))
result['am'] = findEntry(path, "dates/calendars/calendar[gregorian]/dayPeriods/dayPeriodContext[format]/dayPeriodWidth[wide]/dayPeriod[am]", draft)
result['pm'] = findEntry(path, "dates/calendars/calendar[gregorian]/dayPeriods/dayPeriodContext[format]/dayPeriodWidth[wide]/dayPeriod[pm]", draft)
result['longDateFormat'] = convert_date(findEntry(path, "dates/calendars/calendar[gregorian]/dateFormats/dateFormatLength[full]/dateFormat/pattern"))
result['shortDateFormat'] = convert_date(findEntry(path, "dates/calendars/calendar[gregorian]/dateFormats/dateFormatLength[short]/dateFormat/pattern"))
result['longTimeFormat'] = convert_date(findEntry(path, "dates/calendars/calendar[gregorian]/timeFormats/timeFormatLength[full]/timeFormat/pattern"))
result['shortTimeFormat'] = convert_date(findEntry(path, "dates/calendars/calendar[gregorian]/timeFormats/timeFormatLength[short]/timeFormat/pattern"))
endonym = None
if country_code and script_code:
endonym = findEntryDef(path, "localeDisplayNames/languages/language[type=%s_%s_%s]" % (language_code, script_code, country_code))
if not endonym and script_code:
endonym = findEntryDef(path, "localeDisplayNames/languages/language[type=%s_%s]" % (language_code, script_code))
if not endonym and country_code:
endonym = findEntryDef(path, "localeDisplayNames/languages/language[type=%s_%s]" % (language_code, country_code))
if not endonym:
endonym = findEntryDef(path, "localeDisplayNames/languages/language[type=%s]" % (language_code))
result['language_endonym'] = endonym
result['country_endonym'] = findEntryDef(path, "localeDisplayNames/territories/territory[type=%s]" % (country_code))
currency_format = get_number_in_system(path, "numbers/currencyFormats/currencyFormatLength/currencyFormat/pattern", numbering_system)
currency_format = parse_number_format(currency_format, result)
result['currencyFormat'] = currency_format[0]
result['currencyNegativeFormat'] = ''
if len(currency_format) > 1:
result['currencyNegativeFormat'] = currency_format[1]
result['currencySymbol'] = ''
result['currencyDisplayName'] = ''
if result['currencyIsoCode']:
result['currencySymbol'] = findEntryDef(path, "numbers/currencies/currency[%s]/symbol" % result['currencyIsoCode'])
result['currencyDisplayName'] = ';'.join(
findEntryDef(path, 'numbers/currencies/currency[' + result['currencyIsoCode']
+ ']/displayName' + tail)
for tail in ['',] + [
'[count=%s]' % x for x in ('zero', 'one', 'two', 'few', 'many', 'other')
]) + ';'
def findUnitDef(path, stem, fallback=''):
# The displayName for a quantified unit in en.xml is kByte
# instead of kB (etc.), so prefer any unitPattern provided:
for count in ('many', 'few', 'two', 'other', 'zero', 'one'):
try:
ans = findEntry(path, stem + 'unitPattern[count=%s]' % count)
except xpathlite.Error:
continue
# TODO: epxloit count-handling, instead of discarding placeholders
if ans.startswith('{0}'):
ans = ans[3:].lstrip()
if ans:
return ans
return findEntryDef(path, stem + 'displayName', fallback)
# First without quantifier, then quantified each way:
result['byte_unit'] = findEntryDef(
path, 'units/unitLength[type=long]/unit[type=digital-byte]/displayName',
'bytes')
stem = 'units/unitLength[type=short]/unit[type=digital-%sbyte]/'
known = [] # cases where we *do* have a given version:
result['byte_si_quantified'] = ';'.join(unit_quantifiers(findUnitDef, path, stem, 'B', known))
# IEC 60027-2
# http://physics.nist.gov/cuu/Units/binary.html
result['byte_iec_quantified'] = ';'.join(unit_quantifiers(findUnitDef, path, stem % '%sbi', 'iB', known))
# Used for month and day data:
namings = (
('standaloneLong', 'stand-alone', 'wide'),
('standaloneShort', 'stand-alone', 'abbreviated'),
('standaloneNarrow', 'stand-alone', 'narrow'),
('long', 'format', 'wide'),
('short', 'format', 'abbreviated'),
('narrow', 'format', 'narrow'),
)
# Month data:
for cal in ('gregorian',): # We shall want to add to this
stem = 'dates/calendars/calendar[' + cal + ']/months/'
for (key, mode, size) in namings:
prop = 'monthContext[' + mode + ']/monthWidth[' + size + ']/'
result[key + 'Months'] = ';'.join(
findEntry(path, stem + prop + "month[%d]" % i)
for i in range(1, 13)) + ';'
# Day data (for Gregorian, at least):
stem = 'dates/calendars/calendar[gregorian]/days/'
days = ('sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat')
for (key, mode, size) in namings:
prop = 'dayContext[' + mode + ']/dayWidth[' + size + ']/day'
result[key + 'Days'] = ';'.join(
findEntry(path, stem + prop + '[' + day + ']')
for day in days) + ';'
return Locale(result)
if len(sys.argv) != 2:
usage()
cldr_dir = sys.argv[1]
if not os.path.isdir(cldr_dir):
usage()
cldr_files = os.listdir(cldr_dir)
locale_database = {}
# see http://www.unicode.org/reports/tr35/tr35-info.html#Default_Content
defaultContent_locales = {}
for ns in findTagsInFile(os.path.join(cldr_dir, '..', 'supplemental',
'supplementalMetadata.xml'),
'metadata/defaultContent'):
for data in ns[1:][0]:
if data[0] == u"locales":
defaultContent_locales = data[1].split()
for file in defaultContent_locales:
items = file.split("_")
if len(items) == 3:
language_code = items[0]
script_code = items[1]
country_code = items[2]
else:
if len(items) != 2:
sys.stderr.write('skipping defaultContent locale "' + file + '" [neither lang_script_country nor lang_country]\n')
continue
def from_supplement(tag,
path=os.path.join(dir_name, '..', 'supplemental',
'supplementalData.xml')):
return findTagsInFile(path, tag)
def _generateLocaleInfo(path, language_code, script_code, country_code, variant_code=""):
if not path.endswith(".xml"):
return {}
if language_code == 'root':
# just skip it
return {}
# we do not support variants
# ### actually there is only one locale with variant: en_US_POSIX
# does anybody care about it at all?
if variant_code:
raise xpathlite.Error('we do not support variants ("%s")' % variant_code)
language_id = enumdata.languageCodeToId(language_code)
if language_id <= 0:
raise xpathlite.Error('unknown language code "%s"' % language_code)
script_id = enumdata.scriptCodeToId(script_code)
if script_id == -1:
raise xpathlite.Error('unknown script code "%s"' % script_code)
# we should handle fully qualified names with the territory
if not country_code:
return {}
country_id = enumdata.countryCodeToId(country_code)
if country_id <= 0:
raise xpathlite.Error('unknown country code "%s"' % country_code)
# So we say we accept only those values that have "contributed" or
# "approved" resolution. see http://www.unicode.org/cldr/process.html
# But we only respect the resolution for new datas for backward
# compatibility.
draft = DraftResolution.contributed
result = dict(
language=enumdata.language_list[language_id][0],
language_code=language_code, language_id=language_id,
script=enumdata.script_list[script_id][0],
script_code=script_code, script_id=script_id,
country=enumdata.country_list[country_id][0],
country_code=country_code, country_id=country_id,
variant_code=variant_code)
(dir_name, file_name) = os.path.split(path)
def from_supplement(tag,
path=os.path.join(dir_name, '..', 'supplemental',
'supplementalData.xml')):
return findTagsInFile(path, tag)
currencies = from_supplement('currencyData/region[iso3166=%s]' % country_code)
result['currencyIsoCode'] = ''
result['currencyDigits'] = 2
result['currencyRounding'] = 1
if currencies:
for e in currencies:
if e[0] == 'currency':
t = [x[1] == 'false' for x in e[1] if x[0] == 'tender']
if t and t[0]:
pass
elif not any(x[0] == 'to' for x in e[1]):
result['currencyIsoCode'] = (x[1] for x in e[1] if x[0] == 'iso4217').next()
break
if result['currencyIsoCode']:
t = from_supplement("currencyData/fractions/info[iso4217=%s]"
% result['currencyIsoCode'])
if t and t[0][0] == 'info':
result['currencyDigits'] = (int(x[1]) for x in t[0][1] if x[0] == 'digits').next()
result['currencyRounding'] = (int(x[1]) for x in t[0][1] if x[0] == 'rounding').next()
numbering_system = None
try:
numbering_system = findEntry(path, "numbers/defaultNumberingSystem")
except:
pass
def findEntryDef(path, xpath, value=''):
try:
return findEntry(path, xpath)
except xpathlite.Error:
return value
def get_number_in_system(path, xpath, numbering_system):
if numbering_system:
try:
return findEntry(path, xpath + "[numberSystem=" + numbering_system + "]")
except xpathlite.Error:
# in CLDR 1.9 number system was refactored for numbers (but not for currency)
# so if previous findEntry doesn't work we should try this:
try:
return findEntry(path, xpath.replace("/symbols/", "/symbols[numberSystem=" + numbering_system + "]/"))
except xpathlite.Error:
# fallback to default
pass
return findEntry(path, xpath)
result['decimal'] = get_number_in_system(path, "numbers/symbols/decimal", numbering_system)
result['group'] = get_number_in_system(path, "numbers/symbols/group", numbering_system)
result['list'] = get_number_in_system(path, "numbers/symbols/list", numbering_system)
result['percent'] = get_number_in_system(path, "numbers/symbols/percentSign", numbering_system)
try:
numbering_systems = {}
for ns in findTagsInFile(os.path.join(cldr_dir, '..', 'supplemental',
'numberingSystems.xml'),
'numberingSystems'):
tmp = {}
id = ""
for data in ns[1:][0]: # ns looks like this: [u'numberingSystem', [(u'digits', u'0123456789'), (u'type', u'numeric'), (u'id', u'latn')]]
tmp[data[0]] = data[1]
if data[0] == u"id":
id = data[1]
numbering_systems[id] = tmp
result['zero'] = numbering_systems[numbering_system][u"digits"][0]
except e:
sys.stderr.write("Native zero detection problem:\n" + str(e) + "\n")
result['zero'] = get_number_in_system(path, "numbers/symbols/nativeZeroDigit", numbering_system)
result['minus'] = get_number_in_system(path, "numbers/symbols/minusSign", numbering_system)
result['plus'] = get_number_in_system(path, "numbers/symbols/plusSign", numbering_system)
result['exp'] = get_number_in_system(path, "numbers/symbols/exponential", numbering_system).lower()
result['quotationStart'] = findEntry(path, "delimiters/quotationStart")
result['quotationEnd'] = findEntry(path, "delimiters/quotationEnd")
result['alternateQuotationStart'] = findEntry(path, "delimiters/alternateQuotationStart")
result['alternateQuotationEnd'] = findEntry(path, "delimiters/alternateQuotationEnd")
result['listPatternPartStart'] = parse_list_pattern_part_format(findEntry(path, "listPatterns/listPattern/listPatternPart[start]"))
result['listPatternPartMiddle'] = parse_list_pattern_part_format(findEntry(path, "listPatterns/listPattern/listPatternPart[middle]"))
result['listPatternPartEnd'] = parse_list_pattern_part_format(findEntry(path, "listPatterns/listPattern/listPatternPart[end]"))
result['listPatternPartTwo'] = parse_list_pattern_part_format(findEntry(path, "listPatterns/listPattern/listPatternPart[2]"))
result['am'] = findEntry(path, "dates/calendars/calendar[gregorian]/dayPeriods/dayPeriodContext[format]/dayPeriodWidth[wide]/dayPeriod[am]", draft)
result['pm'] = findEntry(path, "dates/calendars/calendar[gregorian]/dayPeriods/dayPeriodContext[format]/dayPeriodWidth[wide]/dayPeriod[pm]", draft)
result['longDateFormat'] = convert_date(findEntry(path, "dates/calendars/calendar[gregorian]/dateFormats/dateFormatLength[full]/dateFormat/pattern"))
result['shortDateFormat'] = convert_date(findEntry(path, "dates/calendars/calendar[gregorian]/dateFormats/dateFormatLength[short]/dateFormat/pattern"))
result['longTimeFormat'] = convert_date(findEntry(path, "dates/calendars/calendar[gregorian]/timeFormats/timeFormatLength[full]/timeFormat/pattern"))
result['shortTimeFormat'] = convert_date(findEntry(path, "dates/calendars/calendar[gregorian]/timeFormats/timeFormatLength[short]/timeFormat/pattern"))
endonym = None
if country_code and script_code:
endonym = findEntryDef(path, "localeDisplayNames/languages/language[type=%s_%s_%s]" % (language_code, script_code, country_code))
if not endonym and script_code:
endonym = findEntryDef(path, "localeDisplayNames/languages/language[type=%s_%s]" % (language_code, script_code))
if not endonym and country_code:
endonym = findEntryDef(path, "localeDisplayNames/languages/language[type=%s_%s]" % (language_code, country_code))
if not endonym:
endonym = findEntryDef(path, "localeDisplayNames/languages/language[type=%s]" % (language_code))
result['language_endonym'] = endonym
result['country_endonym'] = findEntryDef(path, "localeDisplayNames/territories/territory[type=%s]" % (country_code))
currency_format = get_number_in_system(path, "numbers/currencyFormats/currencyFormatLength/currencyFormat/pattern", numbering_system)
currency_format = parse_number_format(currency_format, result)
result['currencyFormat'] = currency_format[0]
result['currencyNegativeFormat'] = ''
if len(currency_format) > 1:
result['currencyNegativeFormat'] = currency_format[1]
result['currencySymbol'] = ''
result['currencyDisplayName'] = ''
if result['currencyIsoCode']:
result['currencySymbol'] = findEntryDef(path, "numbers/currencies/currency[%s]/symbol" % result['currencyIsoCode'])
result['currencyDisplayName'] = ';'.join(
findEntryDef(path, 'numbers/currencies/currency[' + result['currencyIsoCode']
+ ']/displayName' + tail)
for tail in ['',] + [
'[count=%s]' % x for x in ('zero', 'one', 'two', 'few', 'many', 'other')
]) + ';'
def findUnitDef(path, stem, fallback=''):
# The displayName for a quantified unit in en.xml is kByte
# instead of kB (etc.), so prefer any unitPattern provided:
for count in ('many', 'few', 'two', 'other', 'zero', 'one'):
try:
ans = findEntry(path, stem + 'unitPattern[count=%s]' % count)
except xpathlite.Error:
continue
# TODO: epxloit count-handling, instead of discarding placeholders
if ans.startswith('{0}'):
ans = ans[3:].lstrip()
if ans:
return ans
return findEntryDef(path, stem + 'displayName', fallback)
# First without quantifier, then quantified each way:
result['byte_unit'] = findEntryDef(
path, 'units/unitLength[type=long]/unit[type=digital-byte]/displayName',
'bytes')
stem = 'units/unitLength[type=short]/unit[type=digital-%sbyte]/'
known = [] # cases where we *do* have a given version:
result['byte_si_quantified'] = ';'.join(unit_quantifiers(findUnitDef, path, stem, 'B', known))
# IEC 60027-2
# http://physics.nist.gov/cuu/Units/binary.html
result['byte_iec_quantified'] = ';'.join(unit_quantifiers(findUnitDef, path, stem % '%sbi', 'iB', known))
# Used for month and day data:
namings = (
('standaloneLong', 'stand-alone', 'wide'),
('standaloneShort', 'stand-alone', 'abbreviated'),
('standaloneNarrow', 'stand-alone', 'narrow'),
('long', 'format', 'wide'),
('short', 'format', 'abbreviated'),
('narrow', 'format', 'narrow'),
)
# Month data:
for cal in ('gregorian',): # We shall want to add to this
stem = 'dates/calendars/calendar[' + cal + ']/months/'
for (key, mode, size) in namings:
prop = 'monthContext[' + mode + ']/monthWidth[' + size + ']/'
result[key + 'Months'] = ';'.join(
findEntry(path, stem + prop + "month[%d]" % i)
for i in range(1, 13)) + ';'
# Day data (for Gregorian, at least):
stem = 'dates/calendars/calendar[gregorian]/days/'
days = ('sun', 'mon', 'tue', 'wed', 'thu', 'fri', 'sat')
for (key, mode, size) in namings:
prop = 'dayContext[' + mode + ']/dayWidth[' + size + ']/day'
result[key + 'Days'] = ';'.join(
findEntry(path, stem + prop + '[' + day + ']')
for day in days) + ';'
return Locale(result)
if len(sys.argv) != 2:
usage()
cldr_dir = sys.argv[1]
if not os.path.isdir(cldr_dir):
usage()
cldr_files = os.listdir(cldr_dir)
locale_database = {}
# see http://www.unicode.org/reports/tr35/tr35-info.html#Default_Content
defaultContent_locales = {}
for ns in findTagsInFile(os.path.join(cldr_dir, '..', 'supplemental',
'supplementalMetadata.xml'),
'metadata/defaultContent'):
for data in ns[1:][0]:
if data[0] == u"locales":
defaultContent_locales = data[1].split()
for file in defaultContent_locales:
items = file.split("_")
if len(items) == 3:
language_code = items[0]
script_code = items[1]
country_code = items[2]
else:
if len(items) != 2:
sys.stderr.write('skipping defaultContent locale "' + file + '" [neither lang_script_country nor lang_country]\n')
continue
