def testSemiIdenticalCliques(self):
messages = [
tclib.Message(
text='Hello USERNAME',
placeholders=[tclib.Placeholder('USERNAME', '$1', 'Joi')]),
tclib.Message(
text='Hello USERNAME',
placeholders=[tclib.Placeholder('USERNAME', '%s', 'Joi')]),
]
self.failUnless(messages[0].GetId() == messages[1].GetId())
# Both of the above would share a translation.
translation = tclib.Translation(
id=messages[0].GetId(),
text='Bonjour USERNAME',
placeholders=[tclib.Placeholder('USERNAME', '$1', 'Joi')])
factory = clique.UberClique()
cliques = [factory.MakeClique(msg) for msg in messages]
for clq in cliques:
clq.AddTranslation(translation, 'fr')
self.failUnless(cliques[0].MessageForLanguage('fr').GetRealContent() ==
'Bonjour $1')
self.failUnless(cliques[1].MessageForLanguage('fr').GetRealContent() ==
'Bonjour %s')
def GetPlaceholders(msg):
tag_list = []
ph_names = re.findall(
'%{([^}]+)}|<(([a-zA-Z]+)[^>]*)(?<!/)>|</(([a-zA-Z]+)[^>]*)>|<(([a-zA-Z]+)[^>]*)/>',
msg)
placeholders = []
for (gettext_ph, open_tag_contents, open_tag, close_tag_contents,
close_tag, unary_tag_contents, unary_tag) in ph_names:
if gettext_ph != '':
placeholders.append(
tclib.Placeholder(gettext_ph.upper(), '%%{%s}' % gettext_ph,
'(replaceable)'))
elif open_tag != '':
tag_list.append(open_tag)
placeholders.append(
tclib.Placeholder(
MakeHtmlPlaceholderName(open_tag, 'begin', tag_list),
'<%s>' % open_tag_contents, '(HTML code)'))
elif close_tag != '':
placeholders.append(
tclib.Placeholder(
MakeHtmlPlaceholderName(close_tag, 'end', tag_list),
'</%s>' % close_tag_contents, '(HTML code)'))
elif unary_tag != '':
tag_list.append(unary_tag)
placeholders.append(
tclib.Placeholder(
MakeHtmlPlaceholderName(unary_tag, None, tag_list),
'<%s/>' % unary_tag_contents, '(HTML code)'))
return placeholders
def testConstruct(self):
msg = tclib.Message(text=" Hello USERNAME, how are you? BINGO\t\t",
placeholders=[tclib.Placeholder('USERNAME', '%s', 'Joi'),
tclib.Placeholder('BINGO', '%d', '11')])
msg_node = message.MessageNode.Construct(None, msg, 'BINGOBONGO')
self.failUnless(msg_node.children[0].name == 'ph')
self.failUnless(msg_node.children[0].children[0].name == 'ex')
self.failUnless(msg_node.children[0].children[0].GetCdata() == 'Joi')
self.failUnless(msg_node.children[1].children[0].GetCdata() == '11')
self.failUnless(msg_node.ws_at_start == ' ')
self.failUnless(msg_node.ws_at_end == '\t\t')
def testXmlFormatContentWithEntities(self):
'''Tests a bug where would not be escaped correctly.'''
from grit import tclib
msg_node = message.MessageNode.Construct(None, tclib.Message(
text = 'BEGIN_BOLDHelloWHITESPACEthere!END_BOLD Bingo!',
placeholders = [
tclib.Placeholder('BEGIN_BOLD', '<b>', 'bla'),
tclib.Placeholder('WHITESPACE', ' ', 'bla'),
tclib.Placeholder('END_BOLD', '</b>', 'bla')]),
'BINGOBONGO')
xml = msg_node.FormatXml()
self.failUnless(xml.find(' ') == -1, 'should have no entities')
def testRegressionTranslationInherited(self):
'''Regression tests a bug that was caused by grit.tclib.Translation
inheriting from the translation console's Translation object
instead of only owning an instance of it.
'''
msg = tclib.Message(text=u"BLA1\r\nFrom: BLA2 \u00fe BLA3",
placeholders=[
tclib.Placeholder('BLA1', '%s', '%s'),
tclib.Placeholder('BLA2', '%s', '%s'),
tclib.Placeholder('BLA3', '%s', '%s')
])
transl = tclib.Translation(text=msg.GetPresentableContent(),
placeholders=msg.GetPlaceholders())
content = transl.GetContent()
self.failUnless(isinstance(content[3], types.UnicodeType))
def AddMessage(self, msgtext, description, meaning, translateable):
if msgtext == '':
return
msg = tclib.Message(description=description, meaning=meaning)
unescaped_text = self.UnEscape(msgtext)
parts = PLACEHOLDER_RE.split(unescaped_text)
in_placeholder = False
for part in parts:
if part == '':
continue
elif part == '[![':
in_placeholder = True
elif part == ']!]':
in_placeholder = False
else:
if in_placeholder:
msg.AppendPlaceholder(
tclib.Placeholder(part, '[![%s]!]' % part,
'(placeholder)'))
else:
msg.AppendText(part)
self.skeleton_.append(
self.uberclique.MakeClique(msg, translateable=translateable))
# if statement needed because this is supposed to be idempotent (so never
# set back to false)
if translateable:
self.translatable_chunk_ = True
def SubstituteMessage(self, msg):
'''Apply substitutions to a tclib.Message object.
Text of the form [message_name] will be replaced by a new placeholder,
whose presentation will take the form the message_name_{UsageCount}, and
whose example will be the message's value. Existing placeholders are
not affected.
Args:
msg: A tclib.Message object.
Returns:
A tclib.Message object, with substitutions done.
'''
from grit import tclib # avoid circular import
counts = {}
text = msg.GetPresentableContent()
placeholders = []
newtext = ''
for f in self.GetExp().split(text):
sub = self._SubFragment(f)
if f != sub:
f = str(f)
count = counts.get(f, 0) + 1
counts[f] = count
name = "%s_%d" % (f[1:-1], count)
placeholders.append(tclib.Placeholder(name, f, sub))
newtext += name
else:
newtext += f
if placeholders:
return tclib.Message(newtext, msg.GetPlaceholders() + placeholders,
msg.GetDescription(), msg.GetMeaning())
else:
return msg
def Callback(id, structure):
if id not in self.cliques_:
if debug: print "Ignoring translation #%s" % id
return
if debug: print "Adding translation #%s" % id
# We fetch placeholder information from the original message (the XTB file
# only contains placeholder names).
original_msg = self.BestClique(id).GetMessage()
translation = tclib.Translation(id=id)
for is_ph,text in structure:
if not is_ph:
translation.AppendText(text)
else:
found_placeholder = False
for ph in original_msg.GetPlaceholders():
if ph.GetPresentation() == text:
translation.AppendPlaceholder(tclib.Placeholder(
ph.GetPresentation(), ph.GetOriginal(), ph.GetExample()))
found_placeholder = True
break
if not found_placeholder:
raise exception.MismatchingPlaceholders(
'Translation for message ID %s had <ph name="%s%/>, no match\n'
'in original message' % (id, text))
self.FindCliqueAndAddTranslation(translation, lang)
def CreateTclibMessage(self, android_string):
"""Transforms a <string/> element from strings.xml into a tclib.Message.
Interprets whitespace, quotes, and escaped characters in the android_string
according to Android's formatting and styling rules for strings. Also
converts <xliff:g> placeholders into <ph> placeholders, e.g.:
<xliff:g id="website" example="google.com">%s</xliff:g>
becomes
<ph name="website"><ex>google.com</ex>%s</ph>
Returns:
The tclib.Message.
"""
msg = tclib.Message()
current_text = '' # Accumulated text that hasn't yet been added to msg.
nodes = android_string.childNodes
for i, node in enumerate(nodes):
# Handle text nodes.
if node.nodeType in (Node.TEXT_NODE, Node.CDATA_SECTION_NODE):
current_text += node.data
# Handle <xliff:g> and other tags.
elif node.nodeType == Node.ELEMENT_NODE:
if node.tagName == 'xliff:g':
assert node.hasAttribute(
'id'), 'missing id: ' + node.data()
placeholder_id = node.getAttribute('id')
placeholder_text = self.__FormatPlaceholderText(node)
placeholder_example = node.getAttribute('example')
if not placeholder_example:
print(
'Info: placeholder does not contain an example: %s'
% node.toxml())
placeholder_example = placeholder_id.upper()
msg.AppendPlaceholder(
tclib.Placeholder(placeholder_id, placeholder_text,
placeholder_example))
else:
print(
'Warning: removing tag <%s> which must be inside a '
'placeholder: %s' % (node.tagName, node.toxml()))
msg.AppendText(self.__FormatPlaceholderText(node))
# Handle other nodes.
elif node.nodeType != Node.COMMENT_NODE:
assert False, 'Unknown node type: %s' % node.nodeType
is_last_node = (i == len(nodes) - 1)
if (current_text and
(is_last_node or nodes[i + 1].nodeType == Node.ELEMENT_NODE)):
# For messages containing just text and comments (no xml tags) Android
# strips leading and trailing whitespace. We mimic that behavior.
if not msg.GetContent() and is_last_node:
current_text = current_text.strip()
msg.AppendText(self.__FormatAndroidString(current_text))
current_text = ''
return msg
def _ParsePlaceholder(self, placeholder, msg):
'''Extracts a placeholder from a DOM node and adds it to a tclib Message.
Args:
placeholder: A DOM node of the form:
<ph name="PLACEHOLDER_NAME">Placeholder text<ex>Example value</ex></ph>
msg: The placeholder is added to this message.
'''
text = []
example_text = []
for node1 in placeholder.childNodes:
if (node1.nodeType == minidom.Node.TEXT_NODE):
text.append(node1.data)
elif (node1.nodeType == minidom.Node.ELEMENT_NODE
and node1.tagName == 'ex'):
for node2 in node1.childNodes:
example_text.append(node2.toxml())
else:
raise Exception('Unexpected element inside a placeholder: ' +
node2.toxml())
if example_text == []:
# In such cases the original text is okay for an example.
example_text = text
replaced_text = self.Escape(''.join(text).strip())
replaced_text = replaced_text.replace('$1', self._config['app_name'])
replaced_text = replaced_text.replace('$2', self._config['os_name'])
replaced_text = replaced_text.replace('$3', self._config['frame_name'])
msg.AppendPlaceholder(
tclib.Placeholder(placeholder.attributes['name'].value,
replaced_text, ''.join(example_text).strip()))
def _AddNontranslateableChunk(self, chunk):
'''Adds a nontranslateable chunk.'''
if self.single_message_:
ph = tclib.Placeholder('XX%02dXX' % self.ph_counter_, chunk, chunk)
self.ph_counter_ += 1
self.single_message_.AppendPlaceholder(ph)
else:
self.skeleton_.append(chunk)
def testPseudoMessage(self):
msg = tclib.Message(
text='Hello USERNAME, how are you?',
placeholders=[tclib.Placeholder('USERNAME', '%s', 'Joi')])
trans = pseudo.PseudoMessage(msg)
# TODO(joi) It would be nicer if 'you' -> 'youPou' instead of
# 'you' -> 'youPyou' and if we handled the silent e in 'are'
self.failUnless(trans.GetPresentableContent() == pseudo.MapVowels(
u'HePelloPo USERNAME, hoPow aParePe youPyou?', 1))
def EndParsing(self):
super(type(self), self).EndParsing()
# Make the text (including placeholder references) and list of placeholders,
# then strip and store leading and trailing whitespace and create the
# tclib.Message() and a clique to contain it.
text = ''
placeholders = []
for item in self.mixed_content:
if isinstance(item, types.StringTypes):
text += item
else:
presentation = item.attrs['name'].upper()
text += presentation
ex = ' '
if len(item.children):
ex = item.children[0].GetCdata()
original = item.GetCdata()
placeholders.append(
tclib.Placeholder(presentation, original, ex))
m = _WHITESPACE.match(text)
if m:
self.ws_at_start = m.group('start')
self.ws_at_end = m.group('end')
text = m.group('body')
self.shortcut_groups_ = self._SPLIT_RE.split(
self.attrs['shortcut_groups'])
self.shortcut_groups_ = [i for i in self.shortcut_groups_ if i != '']
description_or_id = self.attrs['desc']
if description_or_id == '' and 'name' in self.attrs:
description_or_id = 'ID: %s' % self.attrs['name']
assigned_id = None
if (self.attrs['use_name_for_id'] == 'true'
and self.SatisfiesOutputCondition()):
assigned_id = self.attrs['name']
message = tclib.Message(text=text,
placeholders=placeholders,
description=description_or_id,
meaning=self.attrs['meaning'],
assigned_id=assigned_id)
self.clique = self.UberClique().MakeClique(message,
self.IsTranslateable())
for group in self.shortcut_groups_:
self.clique.AddToShortcutGroup(group)
if self.attrs['custom_type'] != '':
self.clique.SetCustomType(
util.NewClassInstance(self.attrs['custom_type'],
clique.CustomType))
elif self.attrs['validation_expr'] != '':
self.clique.SetCustomType(
clique.OneOffCustomType(self.attrs['validation_expr']))
def testAll(self):
text = u'Howdie USERNAME'
phs = [tclib.Placeholder(u'USERNAME', u'%s', 'Joi')]
msg = tclib.Message(text=text, placeholders=phs)
self.failUnless(msg.GetPresentableContent() == 'Howdie USERNAME')
trans = tclib.Translation(text=text, placeholders=phs)
self.failUnless(trans.GetPresentableContent() == 'Howdie USERNAME')
self.failUnless(
isinstance(trans.GetPresentableContent(), types.StringTypes))
def testPseudolocales(self):
p1 = tclib.Placeholder(u'USERNAME', '%s', 'foo')
p2 = tclib.Placeholder(u'EMAIL', '%s', 'bar')
msg = tclib.Message()
msg.AppendText('h_')
msg.AppendPlaceholder(p1)
msg.AppendPlaceholder(p2)
msg.AppendText('w')
self.assertEqual(
pl.PseudoLongStringMessage(msg).GetContent(),
[u'\u0125_', p1, p2, u'\u0175', ' - one two three four'])
msg.AppendText('hello world')
self.assertEqual(
pl.PseudoRTLMessage(msg).GetContent(), [
u'\u202eh\u202c_', p1, p2,
u'\u202ewhello\u202c \u202eworld\u202c'
])
def testClique(self):
factory = clique.UberClique()
msg = tclib.Message(
text='Hello USERNAME, how are you?',
placeholders=[tclib.Placeholder('USERNAME', '%s', 'Joi')])
c = factory.MakeClique(msg)
self.failUnless(c.GetMessage() == msg)
self.failUnless(c.GetId() == msg.GetId())
msg_fr = tclib.Translation(
text='Bonjour USERNAME, comment ca va?',
id=msg.GetId(),
placeholders=[tclib.Placeholder('USERNAME', '%s', 'Joi')])
msg_de = tclib.Translation(
text='Guten tag USERNAME, wie geht es dir?',
id=msg.GetId(),
placeholders=[tclib.Placeholder('USERNAME', '%s', 'Joi')])
c.AddTranslation(msg_fr, 'fr')
factory.FindCliqueAndAddTranslation(msg_de, 'de')
# sort() sorts lists in-place and does not return them
for lang in ('en', 'fr', 'de'):
self.failUnless(lang in c.clique)
self.failUnless(
c.MessageForLanguage('fr').GetRealContent() ==
msg_fr.GetRealContent())
try:
c.MessageForLanguage('zh-CN', False)
self.fail('Should have gotten exception')
except:
pass
self.failUnless(c.MessageForLanguage('zh-CN', True) != None)
rex = re.compile('fr|de|bingo')
self.failUnless(len(c.AllMessagesThatMatch(rex, False)) == 2)
self.failUnless(
c.AllMessagesThatMatch(rex, True)[pseudo.PSEUDO_LANG] is not None)
def testTagsWithCommonSubstring(self):
word = 'ABCDEFGHIJ'
text = ' '.join([word[:i] for i in range(1, 11)])
phs = [
tclib.Placeholder(word[:i], str(i), str(i)) for i in range(1, 11)
]
try:
msg = tclib.Message(text=text, placeholders=phs)
self.failUnless(msg.GetRealContent() == '1 2 3 4 5 6 7 8 9 10')
except:
self.fail('tclib.Message() should handle placeholders that are '
'substrings of each other')
def testBuildAndUnbuildTree(self):
p1 = tclib.Placeholder(u'USERNAME', '%s', 'foo')
p2 = tclib.Placeholder(u'EMAIL', '%s', 'bar')
msg = tclib.Message()
msg.AppendText('hello')
msg.AppendPlaceholder(p1)
msg.AppendPlaceholder(p2)
msg.AppendText('world')
tree, placeholders = pl.BuildTreeFromMessage(msg)
self.assertTreesEqual(
tree,
pl.NodeSequence([
pl.RawText('hello'), PLACEHOLDER_NODE, PLACEHOLDER_NODE,
pl.RawText('world')
]))
self.assertEqual(placeholders, [p1, p2])
transl = pl.ToTranslation(tree, placeholders)
self.assertEqual(transl.GetContent(), ['hello', p1, p2, 'world'])
def EndParsing(self):
super(MessageNode, self).EndParsing()
# Make the text (including placeholder references) and list of placeholders,
# then strip and store leading and trailing whitespace and create the
# tclib.Message() and a clique to contain it.
text = ''
placeholders = []
for item in self.mixed_content:
if isinstance(item, types.StringTypes):
text += item
else:
presentation = item.attrs['name'].upper()
text += presentation
ex = ' '
if len(item.children):
ex = item.children[0].GetCdata()
original = item.GetCdata()
placeholders.append(
tclib.Placeholder(presentation, original, ex))
m = _WHITESPACE.match(text)
if m:
self.ws_at_start = m.group('start')
self.ws_at_end = m.group('end')
text = m.group('body')
self.shortcut_groups_ = self._SPLIT_RE.split(
self.attrs['shortcut_groups'])
self.shortcut_groups_ = [i for i in self.shortcut_groups_ if i != '']
description_or_id = self.attrs['desc']
if description_or_id == '' and 'name' in self.attrs:
description_or_id = 'ID: %s' % self.attrs['name']
assigned_id = None
if self.attrs['use_name_for_id'] == 'true':
assigned_id = self.attrs['name']
message = tclib.Message(text=text,
placeholders=placeholders,
description=description_or_id,
meaning=self.attrs['meaning'],
assigned_id=assigned_id)
self.InstallMessage(message)
def Placeholderize(self, text):
'''Creates a tclib.Message object from 'text', attempting to recognize
a few different formats of text that can be automatically placeholderized
(HTML code, printf-style format strings, and FormatMessage-style format
strings).
'''
try:
# First try HTML placeholderizing.
# TODO(joi) Allow use of non-TotalRecall flavors of HTML placeholderizing
msg = tr_html.HtmlToMessage(text, True)
for item in msg.GetContent():
if not isinstance(item, types.StringTypes):
return msg # Contained at least one placeholder, so we're done
# HTML placeholderization didn't do anything, so try to find printf or
# FormatMessage format specifiers and change them into placeholders.
msg = tclib.Message()
parts = _FORMAT_SPECIFIER.split(text)
todo_counter = 1 # We make placeholder IDs 'TODO_0001' etc.
for part in parts:
if _FORMAT_SPECIFIER.match(part):
msg.AppendPlaceholder(
tclib.Placeholder('TODO_%04d' % todo_counter, part,
'TODO'))
todo_counter += 1
elif part != '':
msg.AppendText(part)
if self.role_model and len(
parts) > 1: # there are TODO placeholders
role_model_msg = self.role_model.UberClique(
).BestCliqueByOriginalText(msg.GetRealContent(), '')
if role_model_msg:
# replace wholesale to get placeholder names and examples
msg = role_model_msg
return msg
except:
print 'Exception processing message with text "%s"' % text
raise
def testPlaceholderNameChecking(self):
try:
ph = tclib.Placeholder('BINGO BONGO', 'bla', 'bla')
raise Exception("We shouldn't get here")
except exception.InvalidPlaceholderName:
pass # Expect exception to be thrown because presentation contained space
def EndParsing(self):
super(MessageNode, self).EndParsing()
# Make the text (including placeholder references) and list of placeholders,
# verify placeholder formats, then strip and store leading and trailing
# whitespace and create the tclib.Message() and a clique to contain it.
text = ''
placeholders = []
for item in self.mixed_content:
if isinstance(item, six.string_types):
# Not a <ph> element: fail if any <ph> formatters are detected.
if _FORMATTERS.search(item):
print(_BAD_PLACEHOLDER_MSG % (item, self.source))
raise exception.PlaceholderNotInsidePhNode
text += item
else:
# Extract the <ph> element components.
presentation = item.attrs['name'].upper()
text += presentation
ex = ' ' # <ex> example element cdata if present.
if len(item.children):
ex = item.children[0].GetCdata()
original = item.GetCdata()
# Sanity check the <ph> element content.
cdata = original
# Replace all HTML tag tokens in cdata.
match = _HTMLTOKEN.search(cdata)
while match:
cdata = cdata.replace(match.group(0), '_')
match = _HTMLTOKEN.search(cdata)
# Replace all HTML entities in cdata.
match = _HTMLENTITY.search(cdata)
while match:
cdata = cdata.replace(match.group(0), '_')
match = _HTMLENTITY.search(cdata)
# Remove first matching formatter from cdata.
match = _FORMATTERS.search(cdata)
if match:
cdata = cdata.replace(match.group(0), '')
# Fail if <ph> special chars remain in cdata.
if re.search(r'[%\$]', cdata):
message_id = self.attrs['name'] + ' ' + original
print(_INVALID_PH_CHAR_MSG % (message_id, self.source))
raise exception.InvalidCharactersInsidePhNode
# Otherwise, accept this <ph> placeholder.
placeholders.append(
tclib.Placeholder(presentation, original, ex))
m = _WHITESPACE.match(text)
if m:
self.ws_at_start = m.group('start')
self.ws_at_end = m.group('end')
text = m.group('body')
self.shortcut_groups_ = self._SPLIT_RE.split(
self.attrs['shortcut_groups'])
self.shortcut_groups_ = [i for i in self.shortcut_groups_ if i != '']
description_or_id = self.attrs['desc']
if description_or_id == '' and 'name' in self.attrs:
description_or_id = 'ID: %s' % self.attrs['name']
assigned_id = None
if self.attrs['use_name_for_id'] == 'true':
assigned_id = self.attrs['name']
message = tclib.Message(text=text,
placeholders=placeholders,
description=description_or_id,
meaning=self.attrs['meaning'],
assigned_id=assigned_id)
self.InstallMessage(message)
def HtmlToMessage(html, include_block_tags=False, description=''):
'''Takes a bit of HTML, which must contain only "inline" HTML elements,
and changes it into a tclib.Message. This involves escaping any entities and
replacing any HTML code with placeholders.
If include_block_tags is true, no error will be given if block tags (e.g.
<p> or <br>) are included in the HTML.
Args:
html: 'Hello <b>[USERNAME]</b>, how <i>are</i> you?'
include_block_tags: False
Return:
tclib.Message('Hello START_BOLD1USERNAMEEND_BOLD, '
'howNBSPSTART_ITALICareEND_ITALIC you?',
[ Placeholder('START_BOLD', '<b>', ''),
Placeholder('USERNAME', '[USERNAME]', ''),
Placeholder('END_BOLD', '</b>', ''),
Placeholder('START_ITALIC', '<i>', ''),
Placeholder('END_ITALIC', '</i>', ''), ])
'''
# Approach is:
# - first placeholderize, finding <elements>, [REPLACEABLES] and
# - then escape all character entities in text in-between placeholders
parts = [] # List of strings (for text chunks) and tuples (ID, original)
# for placeholders
count_names = {} # Map of base names to number of times used
end_names = {
} # Map of base names to stack of end tags (for correct nesting)
def MakeNameClosure(base, type=''):
'''Returns a closure that can be called once all names have been allocated
to return the final name of the placeholder. This allows us to minimally
number placeholders for non-overlap.
Also ensures that END_XXX_Y placeholders have the same Y as the
corresponding BEGIN_XXX_Y placeholder when we have nested tags of the same
type.
Args:
base: 'phname'
type: '' | 'begin' | 'end'
Return:
Closure()
'''
name = base.upper()
if type != '':
name = ('%s_%s' % (type, base)).upper()
if name in count_names.keys():
count_names[name] += 1
else:
count_names[name] = 1
def MakeFinalName(name_=name, index=count_names[name] - 1):
if (type.lower() == 'end' and base in end_names.keys()
and len(end_names[base])):
return end_names[base].pop(-1) # For correct nesting
if count_names[name_] != 1:
name_ = '%s_%s' % (name_, _SUFFIXES[index])
# We need to use a stack to ensure that the end-tag suffixes match
# the begin-tag suffixes. Only needed when more than one tag of the
# same type.
if type == 'begin':
end_name = ('END_%s_%s' % (base, _SUFFIXES[index])).upper()
if base in end_names.keys():
end_names[base].append(end_name)
else:
end_names[base] = [end_name]
return name_
return MakeFinalName
current = 0
last_nobreak = False
while current < len(html):
m = _MESSAGE_NO_BREAK_COMMENT.match(html[current:])
if m:
last_nobreak = True
current += m.end()
continue
m = _NBSP.match(html[current:])
if m:
parts.append((MakeNameClosure('SPACE'), m.group()))
current += m.end()
continue
m = _REPLACEABLE.match(html[current:])
if m:
# Replaceables allow - but placeholders don't, so replace - with _
ph_name = MakeNameClosure('X_%s_X' %
m.group('name').replace('-', '_'))
parts.append((ph_name, m.group()))
current += m.end()
continue
m = _SPECIAL_ELEMENT.match(html[current:])
if m:
if not include_block_tags:
if last_nobreak:
last_nobreak = False
else:
raise exception.BlockTagInTranslateableChunk(html)
element_name = 'block' # for simplification
# Get the appropriate group name
for group in m.groupdict().keys():
if m.groupdict()[group]:
break
parts.append((MakeNameClosure(element_name, 'begin'),
html[current:current + m.start(group)]))
parts.append(m.group(group))
parts.append((MakeNameClosure(element_name, 'end'),
html[current + m.end(group):current + m.end()]))
current += m.end()
continue
m = _ELEMENT.match(html[current:])
if m:
element_name = m.group('element').lower()
if not include_block_tags and not element_name in _INLINE_TAGS:
if last_nobreak:
last_nobreak = False
else:
raise exception.BlockTagInTranslateableChunk(
html[current:])
if element_name in _HTML_PLACEHOLDER_NAMES: # use meaningful names
element_name = _HTML_PLACEHOLDER_NAMES[element_name]
# Make a name for the placeholder
type = ''
if not m.group('empty'):
if m.group('closing'):
type = 'end'
else:
type = 'begin'
parts.append((MakeNameClosure(element_name, type), m.group()))
current += m.end()
continue
if len(parts) and isinstance(parts[-1], types.StringTypes):
parts[-1] += html[current]
else:
parts.append(html[current])
current += 1
msg_text = ''
placeholders = []
for part in parts:
if isinstance(part, types.TupleType):
final_name = part[0]()
original = part[1]
msg_text += final_name
placeholders.append(
tclib.Placeholder(final_name, original, '(HTML code)'))
else:
msg_text += part
msg = tclib.Message(text=msg_text,
placeholders=placeholders,
description=description)
content = msg.GetContent()
for ix in range(len(content)):
if isinstance(content[ix], types.StringTypes):
content[ix] = util.UnescapeHtml(content[ix], replace_nbsp=False)
return msg
def testProdFailures(self):
p1 = tclib.Placeholder(u'USERNAME', '%s', 'foo')
msg = tclib.Message()
msg.AppendText(u'{LINE_COUNT, plural,\n =1 {<1 line not shown>}\n'
' other {<')
msg.AppendPlaceholder(p1)
msg.AppendText(u' lines not shown>}\n}')
tree, _ = pl.BuildTreeFromMessage(msg)
self.assertTreesEqual(
tree,
pl.Plural('{LINE_COUNT, plural,\n ', [
pl.PluralOption('=1 {', [pl.RawText('<1 line not shown>')]),
pl.PluralOption('other {', [
pl.RawText('<'), PLACEHOLDER_NODE,
pl.RawText(' lines not shown>')
])
]))
msg = tclib.Message()
msg.AppendText(u'{1, plural,\n \n =1 {Rated ')
msg.AppendPlaceholder(p1)
msg.AppendText(u' by one user.}\n other{Rated ')
msg.AppendPlaceholder(p1)
msg.AppendText(u' by # users.}}')
tree, _ = pl.BuildTreeFromMessage(msg)
self.assertTreesEqual(
tree,
pl.Plural('{1, plural,\n \n ', [
pl.PluralOption('=1 {', [
pl.RawText('Rated '), PLACEHOLDER_NODE,
pl.RawText(' by one user.')
]),
pl.PluralOption('other{', [
pl.RawText('Rated '), PLACEHOLDER_NODE,
pl.RawText(' by # users.')
]),
]))
self.assertBuildTree(
'{count, plural, offset:2\n'
' =1 {{VAR}}\n'
' =2 {{VAR}, {VAR}}\n'
' other {{VAR}, {VAR}, and # more}\n'
' }',
pl.Plural('{count, plural, offset:2\n ', [
pl.PluralOption('=1 {', [VAR_NODE]),
pl.PluralOption(
'=2 {', [VAR_NODE, pl.RawText(', '), VAR_NODE]),
pl.PluralOption('other {', [
VAR_NODE,
pl.RawText(', '), VAR_NODE,
pl.RawText(', and # more')
]),
]))
self.assertBuildTree(
'{NUM_POPUPS,plural,=1{Pop-up blocked} other{# pop-ups blocked}}',
pl.Plural('{NUM_POPUPS,plural,', [
pl.PluralOption('=1{', [pl.RawText('Pop-up blocked')]),
pl.PluralOption('other{', [pl.RawText('# pop-ups blocked')])
]))
self.assertBuildTree(
'Open ${url}',
pl.NodeSequence([pl.RawText('Open '),
pl.BasicVariable('${url}')]))
