def _AddFullPattern(full_pattern, metre_name):
if full_pattern in known_full_patterns:
# TODO(shreevatsa): Figure out what exactly to do in this case
Print('Error: full pattern already present')
Print(metre_name)
Print(full_pattern)
Print(known_full_patterns[full_pattern])
return False
assert full_pattern not in known_full_patterns
known_full_patterns[full_pattern] = {metre_name: True}
return True
def _AddPatternForMetre(metre_name, pada_patterns):
if metre_name in pattern_for_metre:
if pattern_for_metre[metre_name] != pada_patterns:
Print('Mismatch for %s' % metre_name)
Print(pattern_for_metre[metre_name])
Print(' vs ')
Print(pada_patterns)
assert False
# Print('Not adding duplicate as already present: %s' % metre_name)
return
pattern_for_metre[metre_name] = pada_patterns
def _AddArdhasamavrttaPattern(metre_name, odd_and_even_pada_patterns):
"""Given an ardha-sama-vṛtta's pattern, add it."""
(odd_pada_pattern, even_pada_pattern) = odd_and_even_pada_patterns
clean_odd = _CleanUpPattern(odd_pada_pattern)
assert re.match(r'^[LG]*$', clean_odd)
clean_even = _CleanUpPattern(even_pada_pattern)
if clean_even.endswith('L'):
Print('Not adding %s for now, as %s ends with laghu' %
(metre_name, clean_even))
return
assert re.match(r'^[LG]*G$', clean_even), (metre_name, clean_even)
_AddPatternForMetre(metre_name, [clean_odd, clean_even] * 2)
patterns_odd = [clean_odd[:-1] + 'G', clean_odd[:-1] + 'L']
patterns_even = [clean_even[:-1] + 'G', clean_even[:-1] + 'L']
for (a, b, c, d) in itertools.product(patterns_odd,
patterns_even,
repeat=2):
_AddFullPattern(a + b + c + d, metre_name)
for (a, b) in itertools.product(patterns_odd, patterns_even):
_AddHalfPattern(a + b, metre_name, {1, 2})
for a in patterns_odd:
_AddPadaPattern(a, metre_name, {1, 3})
for b in patterns_even:
_AddPadaPattern(b, metre_name, {2, 4})
def _AddPatternForMetre(metre_name, pada_patterns):
if metre_name in pattern_for_metre:
if pattern_for_metre[metre_name] != pada_patterns:
# Print('Mismatch for %s' % metre_name)
# Print(pattern_for_metre[metre_name])
# Print(' vs ')
# Print(pada_patterns)
# assert False
Print('Not overwriting as already present: %s' % metre_name)
return
pattern_for_metre[metre_name] = pada_patterns
def set_up_logger(input_file_name):
logger = logging.getLogger()
log_file = tempfile.NamedTemporaryFile(prefix='read_gretil_%s_' %
os.path.basename(input_file_name),
delete=False)
Print('Logging to %s' % log_file.name)
handler = logging.FileHandler(log_file.name)
handler.setFormatter(
logging.Formatter(
'%(levelname)s\t%(asctime)s %(filename)s:%(lineno)d] %(message)s'))
logger.addHandler(handler)
logger.setLevel(logging.DEBUG)
def post(self):
verse_text = self.request.get('verse_text')
assert type(verse_text) == type('unicode literal')
metre_name = self.request.get('metre_name')
Print('Type of metre_name is ' +
('bytes' if type(metre_name) else 'unicode'))
if isinstance(metre_name, bytes):
metre_name = metre_name.decode('utf-8')
print('Type of metre_name is now ', type(metre_name))
(alignment, table) = find_alignment(verse_text, metre_name)
ret = {'alignment': alignment, 'table': table}
self.response.content_type = b'application/json'
self.response.write(json.dumps(ret))
'%(levelname)s\t%(asctime)s %(filename)s:%(lineno)d] %(message)s'))
logger.addHandler(handler)
logger.setLevel(logging.DEBUG)
if __name__ == '__main__':
args = get_args()
input_file_name = args.input_file
set_up_logger(input_file_name)
text = codecs.open(input_file_name, 'r', 'utf-8').read()
custom_splitter = read.split_gretil.mss_splitter if 'msubhs_u.htm' in input_file_name else None
(verses, text) = read.split_gretil.split(text,
custom_splitter=custom_splitter)
# blocks = list(read.split_gretil.blocks_of_verses_in_text(verses, text))
Print('There are %d verses.' % len(verses))
identifier = identifier_pipeline.IdentifierPipeline()
table = {}
for (verse_number, verse) in enumerate(verses):
verse_number += 1
Print('\nVerse %d is:' % verse_number)
Print('\n '.join((' ' + verse).splitlines()))
Print('End Verse %d' % verse_number)
ok_and_results = identifier.IdentifyFromText(verse)
if not ok_and_results: # None for lines that contain no syllables
continue
(perfect, results) = ok_and_results
if not results:
table['unknown'] = table.get('unknown', 0) + 1
if args.print_unidentified_verses != 'none':
def __init__(self):
global Print
Print = Print()
os.system('clear')
self.do_initial_setup()
self.show_option_list()