def __parse_sentence(sent, doc_offset, tokenizer):
"""
Perform tokenization on sent
:param string sent:
:param tuple doc_offset: (start, end)
:param tokenizer:
:return: Sentence object
"""
sent_obj = models.Sentence(content=sent, doc_offset=doc_offset)
sent_obj.tokens = []
tokens = tokenizer.parse(sent)
current_pos = 0
for tok in tokens:
t = tok.string.strip()
start_offset = sent.find(t, current_pos)
end_offset = start_offset + len(t)
token_obj = models.Token(content=t,
sent_offset=(
start_offset, end_offset),
doc_offset=(
doc_offset[0] + start_offset,
doc_offset[0] + end_offset),
metadata={'POS': tok.tag_})
current_pos = end_offset
sent_obj.tokens.append(token_obj)
return sent_obj
def parse_sentence(sent, doc_offset, tokenizer):
"""
Perform tokenization on sent
:param string sent:
:param tuple doc_offset: (start, end)
:param tokenizer:
:return: models.Sentence object
"""
sent_obj = models.Sentence(content=sent, doc_offset=doc_offset)
sent_obj.tokens = []
tokens = tokenizer.tokenize(sent)
current_pos = 0
for tok in tokens:
start_offset = sent.find(tok, current_pos)
end_offset = start_offset + len(tok)
token_obj = models.Token(
content=tok,
sent_offset=(start_offset, end_offset),
doc_offset=(doc_offset[0] + start_offset,
doc_offset[0] + end_offset),
# metadata={'pos_tag': pos}
)
current_pos = end_offset
sent_obj.tokens.append(token_obj)
return sent_obj
def check_if_user_edited(user, sentenceID):
''' This filter checks if the user edited the sentence
:param string user: the user
:param string sentenceID: the sentence's id
:returns: a boolean saying if the user edited the sentence
'''
s = models.Sentence(oid=sentenceID)
return models.User(username=user)._id == s.state['userID']
def check_if_user_approved(user, sentenceID):
''' This filter checks if the user approved the sentence
:param string user: the user's username
:param string sentenceID: the sentence's id
:returns: a boolean saying if the user approved the sentence
'''
s = models.Sentence(oid=sentenceID)
return models.User(username=user)._id in s.state['approvers']
def check_if_user_edited(user, sentenceID):
''' This filter checks if the user edited the sentence
:Parameters:
- 'user': the user
- 'sentenceID': the sentence
:Returns:
- a boolean saying if the user edited the sentence
'''
s=models.Sentence(oid=sentenceID)
return models.User(username=user)._id == s.state['userID']
def unapprove_translation():
''' This function is called when a user posts an edit.
It first validates the edit and then it it submits it to the database
'''
j = fix_json(request.json)
t = models.Sentence(oid=j[u'old'][u'_id'])
unapprover = models.User(username=j[u'new'][u'unapprover'])
try:
t.unapprove(unapprover)
return json.dumps({ "code:":200, "msg":"Unapproval Succeeded"}),200
except models.MyError as e:
return json.dumps({ "code": e.code , "msg": e.msg }), e.code
except models.LockError as e:
return json.dumps({ "code": e.code , "msg": e.msg , "file_path": e.file_path, "username": e.username, "target_language": e.target_language}), e.code
def edit_translation():
''' This function is called when a user posts an edit.
It first validates the edit and then it it submits it to the database
'''
try:
j = fix_json(request.json)
t = models.Sentence(oid=j[u'old'][u'_id'])
editor = models.User(username=j[u'new'][u'editor'])
t.edit(editor, j[u'new'][u'new_target_sentence'])
return json.dumps({ "code": 200 , "msg": "Edit Succeeded" }), 200
except KeyError:
return json.dumps({ "code": 401 , "msg": "Edit Failed" }), 500
except models.MyError as e:
return json.dumps({ "code": e.code , "msg": e.msg }), e.code
except models.LockError as e:
return json.dumps({ "code": e.code , "msg": e.msg , "file_path": e.file_path, "username": e.username, "target_language": e.target_language}), e.code
def write_mongo(po_fn, userID, status, language, po_root, db):
'''write a po_file to mongodb
:Parameters:
- 'po_fn': the file name of the current pofile
- 'userID': the ID of the user that translated the po file
- 'status': the status of the translations
- 'language': The target_language of the translations (source assumed to be english)
- 'po_root': the root of the po_files
'''
po = polib.pofile(po_fn)
rel_fn = os.path.relpath(po_fn, po_root)
rel_fn = os.path.splitext(rel_fn)[0]
f = models.File(
{
u'file_path': rel_fn,
u'priority': 0,
u'source_language': u'en',
u'target_language': language
},
curr_db=db)
i = 0
reg = re.compile('^:[a-zA-Z0-9]+:`(?!.*<.*>.*)[^`]*`$')
for entry in po.translated_entries():
sentence_status = status
match = re.match(reg, entry.msgstr.encode('utf-8'))
if match is not None and match.group() == entry.msgstr.encode('utf-8'):
sentence_status = "approved"
t = models.Sentence(
{
u'source_language': u'en',
u'source_sentence': entry.msgid.encode('utf-8'),
u'sentenceID': entry.tcomment.encode('utf-8'),
u'sentence_num': i,
u'fileID': f._id,
u'target_sentence': entry.msgstr.encode('utf-8'),
u'target_language': language,
u'userID': userID,
u'status': sentence_status,
u'update_number': 0
},
curr_db=db)
t.save()
i += 1
f.get_num_sentences()
def sentence(self, id=None):
'''This method wraps around the sentence creator to provide the correct db'''
s = models.Sentence(oid=id, curr_db=self.db)
return s