#!/usr/bin/env python3
from nlptools import *
from nlptools.tools import *
from nlptools.resources import *
from nlptools.run import full_run
import os
import time
import logging
import sys
import json
import plac
# exemple :
# sed -e '1d; $d' ../data/data.json | sed 's/\,$/ /g' | python3 analyze.py stemmer -o doc -lang fr -log analyze.log
@plac.annotations(
analyzer=(
"Name oh the NLPpipe ",
"positional",
None,
str,
["stemmer", "termMatcher", "ner", "NPchunker", "POStagger", "gazetteer", "NPchunkerDP", "lefff_tagger"]
),
language=("language", "option", "lang", str, ["fr", "en"]),
output=("Format result ", "option", "o", str, ["doc", "json"]),
param=("initialisation param in json", "option", "param", str),
init_file=("initialisation file [default config.ini]", "option", "ini", str),
log=("log file", "option", "log", str)
)
def main(analyzer, init_file, log, language, param, output="doc"):
compteur = 0
# init d un pipe nlptools
pipe = full_run(analyzer, language, init_file, param, output)
# doc = corpus traité sur la sortie standard
# init_file = fichier de param par defaut
logging.basicConfig(filename=log, level=logging.DEBUG)
t1 = time.time()
# boucle de traitement sur le champ "text" de chaque document
# label et keywords sont extraits, puis replacés
for json_line in sys.stdin:
compteur += 1
try:
data = json.loads(json_line)
except json.decoder.JSONDecodeError:
logging.error("Input format problem line :{} : String could not be converted to JSON".format(compteur))
exit(1)
#print("in".format(compteur))
data["value"] = pipe.pipe_analyse(data["value"])
#print("ou".format(compteur))
sys.stdout.write(json.dumps(data))
sys.stdout.write("\n")
# calcul du temps execution
t2 = time.time()
logging.info("TRACE::Executing times %.3f " % (t2 - t1))
if __name__ == "__main__":
if False:
import cProfile
import pstats
cProfile.runctx("plac.call(main)", globals(), locals(), "Profile.prof")
s = pstats.Stats("Profile.prof")
s.strip_dirs().sort_stats("time").print_stats()
else:
plac.call(main)