nlp/spacy_nlp
2019-05-23 10:09:01 +02:00

72 lines
2.0 KiB
Python
Executable File

#!/usr/bin/env python3.5
# coding=utf-8
import argparse
import os
import spacy
import textwrap
parser = argparse.ArgumentParser(
description='Tag a text file with spaCy and save it as a verticalized text file.'
)
parser.add_argument(
'i',
metavar='txt-sourcefile',
)
parser.add_argument(
'-l',
choices=['de', 'el', 'en', 'es', 'fr', 'it', 'nl', 'pt'],
dest='lang',
required=True
)
parser.add_argument(
'o',
metavar='vrt-destfile',
)
args = parser.parse_args()
SPACY_MODELS = {
'de': 'de_core_news_sm', 'el': 'el_core_news_sm', 'en': 'en_core_web_sm',
'es': 'es_core_news_sm', 'fr': 'fr_core_news_sm', 'it': 'it_core_news_sm',
'nl': 'nl_core_news_sm', 'pt': 'pt_core_news_sm'
}
# Set the language model for spacy
nlp = spacy.load(SPACY_MODELS[args.lang])
# Read text from the input file and if neccessary split it into parts with a
# length of less than 1 million characters.
with open(args.i) as input_file:
text = input_file.read()
texts = textwrap.wrap(text, 1000000, break_long_words=False)
text = None
# Create and open the output file
output_file = open(args.o, 'w+')
output_file.write(
'<?xml version="1.0" encoding="UTF-8"?>\n<corpus>\n<text id="%s">\n' % (
os.path.basename(args.i).rsplit(".", 1)[0]
)
)
for text in texts:
# Run spacy nlp over the text (partial string if above 1 million chars)
doc = nlp(text)
for sent in doc.sents:
output_file.write('<s>\n')
for token in sent:
# Skip whitespace tokens like "\n" or "\t"
if token.text.isspace():
continue
# Write all information in .vrt style to the output file
# text, lemma, simple_pos, pos, ner
output_file.write(
token.text + '\t' + token.lemma_ + '\t'
+ token.pos_ + '\t' + token.tag_ + '\t'
+ (token.ent_type_ if token.ent_type_ != '' else 'NULL') + '\n'
)
output_file.write('</s>\n')
output_file.write('</text>\n</corpus>')
output_file.close()