def mapper_init(self):
#sys.stderr.write('work=%s plain=%s\n'%(self.work_filename,self.plain_text))
work = None
if self.work_filename and os.path.exists(
os.path.join(os.getcwd(), self.work_filename)):
with open(os.path.join(os.getcwd(), self.work_filename),
'r') as work_file:
work = work_file.read()
self.quote_matcher = QuoteMatcher(work=work,
plain_text=self.plain_text)
def mapper_init(self):
#sys.stderr.write('work=%s plain=%s\n'%(self.work_filename,self.plain_text))
work = None
if self.work_filename and os.path.exists(os.path.join(os.getcwd(),self.work_filename)):
with open(os.path.join(os.getcwd(),self.work_filename),'r') as work_file:
work = work_file.read()
self.quote_matcher = QuoteMatcher(work=work, plain_text=self.plain_text)
class MatchQuotedText(MRJob):
OUTPUT_PROTOCOL = JSONValueProtocol
def configure_options(self):
super(MatchQuotedText, self).configure_options()
self.add_file_option('--work', type='str', default='', help='Path to work doc')
self.add_passthrough_option('--plain', type='str', default='false', help='Source text is plain text')
def load_options(self, args):
super(MatchQuotedText, self).load_options(args)
self.work_filename = self.options.work
self.plain_text = self.options.plain == 'true'
def mapper_init(self):
#sys.stderr.write('work=%s plain=%s\n'%(self.work_filename,self.plain_text))
work = None
if self.work_filename and os.path.exists(os.path.join(os.getcwd(),self.work_filename)):
with open(os.path.join(os.getcwd(),self.work_filename),'r') as work_file:
work = work_file.read()
self.quote_matcher = QuoteMatcher(work=work, plain_text=self.plain_text)
def mapper(self, _, line):
try:
self.increment_counter('counters', 'documents_evaluated', 1)
quotes_data = json.loads(line)
if quotes_data['quotes']:
for quote in quotes_data['quotes']:
quote['id'] = quotes_data['id']
self.increment_counter('counters', 'quotes_evaluated', 1)
if self.quote_matcher.match_quote(quote):
self.increment_counter('counters', 'matches', 1)
if len(quote['matched_text']) >= 20 and quote['similarity'] >= 0.90: self.increment_counter('counters', 'high_confidence_matches', 1)
yield (quote['id'], quote)
except KeyboardInterrupt:
raise
except:
self.increment_counter('mapper1', 'fail', 1)
sys.stderr.write(traceback.format_exc()+'\n')
sys.stderr.write('%s\n'%id)
#raise
def reducer(self, id, quotesgen):
try:
has_high_confidence_match = False
quotes = []
for q in quotesgen:
quotes.append(q)
if q['similarity'] >= 0.90: has_high_confidence_match = True
doc = {'id': id, 'quotes': quotes}
yield (None, doc)
self.increment_counter('reducer2', 'matched_docs', 1)
if has_high_confidence_match: self.increment_counter('counters', 'high_confidence_docs', 1)
except:
self.increment_counter('reducer', 'fail', 1)
class MatchQuotedText(MRJob):
OUTPUT_PROTOCOL = JSONValueProtocol
def configure_options(self):
super(MatchQuotedText, self).configure_options()
self.add_file_option('--work',
type='str',
default='',
help='Path to work doc')
self.add_passthrough_option('--plain',
type='str',
default='false',
help='Source text is plain text')
def load_options(self, args):
super(MatchQuotedText, self).load_options(args)
self.work_filename = self.options.work
self.plain_text = self.options.plain == 'true'
def mapper_init(self):
#sys.stderr.write('work=%s plain=%s\n'%(self.work_filename,self.plain_text))
work = None
if self.work_filename and os.path.exists(
os.path.join(os.getcwd(), self.work_filename)):
with open(os.path.join(os.getcwd(), self.work_filename),
'r') as work_file:
work = work_file.read()
self.quote_matcher = QuoteMatcher(work=work,
plain_text=self.plain_text)
def mapper(self, _, line):
try:
self.increment_counter('counters', 'documents_evaluated', 1)
quotes_data = json.loads(line)
if quotes_data['quotes']:
for quote in quotes_data['quotes']:
quote['id'] = quotes_data['id']
self.increment_counter('counters', 'quotes_evaluated', 1)
if self.quote_matcher.match_quote(quote):
self.increment_counter('counters', 'matches', 1)
if len(quote['matched_text']
) >= 20 and quote['similarity'] >= 0.90:
self.increment_counter('counters',
'high_confidence_matches',
1)
yield (quote['id'], quote)
except KeyboardInterrupt:
raise
except:
self.increment_counter('mapper1', 'fail', 1)
sys.stderr.write(traceback.format_exc() + '\n')
sys.stderr.write('%s\n' % id)
#raise
def reducer(self, id, quotesgen):
try:
has_high_confidence_match = False
quotes = []
for q in quotesgen:
quotes.append(q)
if q['similarity'] >= 0.90: has_high_confidence_match = True
doc = {'id': id, 'quotes': quotes}
yield (None, doc)
self.increment_counter('reducer2', 'matched_docs', 1)
if has_high_confidence_match:
self.increment_counter('counters', 'high_confidence_docs', 1)
except:
self.increment_counter('reducer', 'fail', 1)
