def compile_metadata(dirty_metadata):
number_pattern = re.compile(r'[0-9]+')
metadata = {}
# Detect duplicate keys, and prepare dummy list assign values to ~~~~~~~~~~>
key_count = Counter(x.split('[')[0] for x in dirty_metadata.keys())
duplicate_keys = dict((k,v) for k,v in key_count.items() if v > 1)
for k,v in duplicate_keys.items():
metadata[k] = [None for x in range(v)]
# Merge duplicate items, and clean out empty values ~~~~~~~~~~~~~~~~~~~~~~~>
for k,v in dirty_metadata.items():
if v == '' or v is None:
continue
unique_key = k.split('[')[0]
if unique_key in duplicate_keys:
number_match = number_pattern.search(k)
if not number_match:
k_index = 0
else:
k_index = int(number_match.group())
metadata[unique_key][k_index] = v
else:
metadata[k] = v
# Filter out None sub-values ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~>
for k,v in metadata.items():
if type(v) == list:
metadata[k] = [x for x in v if x is not None]
# Filter out None values, and return clean dictionary ~~~~~~~~~~~~~~~~~~~~~>
return dict((k,v) for k,v in metadata.items() if v)
def format_text(dict_format, entries, accent=None, phoneset=None, encoding='windows-1252'):
fmt = dict_formats[dict_format]
if not accent:
accent = fmt['accent']
if not phoneset:
phoneset = fmt['phoneset']
if phoneset == 'ipa':
encoding = 'utf-8'
phonemeset = load_phonemes(accent, phoneset)
for word, context, phonemes, comment, metadata, error in entries:
if error:
print(error, file=sys.stderr)
continue
components = []
if word:
components.append('entry')
word = fmt['word'](word)
if context:
components.append('context')
if comment != None:
if metadata != None:
meta = []
for key, values in sorted(metadata.items()):
meta.extend(['{0}={1}'.format(key, value) for value in values])
comment = '@@ {0} @@{1}'.format(' '.join(meta), comment)
components.append('comment')
if phonemes:
phonemes = phonemeset.format(phonemes)
if len(components) == 0:
print()
else:
printf(fmt['-'.join(components)], encoding, word, context, phonemes, comment)
def format_text(dict_format,
entries,
accent=None,
phoneset=None,
encoding='windows-1252'):
fmt = dict_formats[dict_format]
if not accent:
accent = fmt['accent']
if not phoneset:
phoneset = fmt['phoneset']
if phoneset == 'ipa':
encoding = 'utf-8'
phonemeset = load_phonemes(accent, phoneset)
for word, context, phonemes, comment, metadata, error in entries:
if error:
print(error, file=sys.stderr)
continue
components = []
if word:
components.append('entry')
word = fmt['word'](word)
if context:
components.append('context')
if comment != None:
if metadata != None:
meta = []
for key, values in sorted(metadata.items()):
meta.extend(
['{0}={1}'.format(key, value) for value in values])
comment = '@@ {0} @@{1}'.format(' '.join(meta), comment)
components.append('comment')
if phonemes:
phonemes = phonemeset.format(phonemes)
if len(components) == 0:
print()
else:
printf(fmt['-'.join(components)], encoding, word, context,
phonemes, comment)
def merge(metadatas):
"""merge
given a list of metadatas
return a dictionary of the union of all of them
raise an exception if there is a conflict
"""
all_items = []
for metadata in metadatas:
try:
all_items += metadata.items()
except AttributeError:
continue
if all_items == []:
return None
all_items.sort()
cur_item = all_items[0]
for item in all_items:
if cur_item == item:
continue
if cur_item[0] == item[0]:
raise MergeConflict(cur_item, item)
cur_item = item
return dict(all_items)
def merge(metadatas):
"""merge
given a list of metadatas
return a dictionary of the union of all of them
raise an exception if there is a conflict
"""
all_items = []
for metadata in metadatas:
try:
all_items += metadata.items()
except AttributeError:
continue
if all_items == []:
return None
all_items.sort()
cur_item = all_items[0]
for item in all_items:
if cur_item == item:
continue
if cur_item[0] == item[0]:
raise MergeConflict(cur_item, item)
cur_item = item
return dict(all_items)
