une_graph/UNE_NER.py at main · SPabloCalleja/une_graph · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Tue Oct  4 11:37:36 2022

@author: Pablo
"""


import os

import spacy
from transformers import pipeline
import sys
#!pip install spacy
#!pip uninstall spacy -y
#!python -m spacy download es_core_news_sm
#!pip uninstall transformers -y
#!pip install transformers

#!pip uninstall keras -y
#!pip install keras
#!pip uninstall tensorflow -y
#!pip install tensorflow

nlp = spacy.load("es_core_news_sm")

def get_sentences(text):

    doc = nlp(text)
    ls=[]
    for s in doc.sents:
        ls.append(s)
    return ls


def ner_text(text):

    set_persons=[]
    set_locations=[]
    set_organizations=[]
    set_misc=[]

    global nlp_ner
    general_offset=0
    sentences= get_sentences(text)
    for sent in sentences:
        #print(sent)
        try:
            entities= nlp_ner(str(sent))

            for entity in entities:
                #print(entity)

                start_position = int(entity['start']) + general_offset
                end_position=    int(entity['end']  ) + general_offset
                entity_text= entity['word']+'['+str(start_position)+':'+str(end_position)+']'
                group = entity['entity_group']
                if group == 'ORG':
                    set_organizations.append(entity_text)
                if group == 'LOC':
                    set_locations.append(entity_text)

                if group == 'PER':
                    set_persons.append(entity_text)
                if group == 'MISC':
                    set_misc.append(entity_text)
            ##update offsets
            general_offset=general_offset+len(sent)
        except:
            general_offset=general_offset+len(sent)
            continue

    return set_persons,set_locations,set_organizations,set_misc


def prepare_row(doc_id,section_id,parr_id, per_entities, loc_entities, org_entities,misc_entities):

    row=[doc_id,section_id,parr_id, '|'.join(per_entities), '|'.join(loc_entities), '|'.join(org_entities),'|'.join(misc_entities)]
    return row


def write_tsv(name,content):
    with open(name, "w", encoding='utf8') as write_file:
        for row in content:
            write_file.write('\t'.join(row))
            write_file.write('\n')


nlp_ner = pipeline(
"ner",
model="mrm8488/bert-spanish-cased-finetuned-ner",
tokenizer=(
    'mrm8488/bert-spanish-cased-finetuned-ner',
     {"use_fast": True},

),
 aggregation_strategy= 'max'
)


def process_txt_file(file,filename):

    file_id= filename.replace('.txt','').split('_')[0]
    section_id= filename.replace('.txt','').split('_')[1]
    par_counter=0
    rows=[]
    f = open(file,encoding='utf-8')     # This is a big file
    for line in f:                # Using 'for ... in' on file object

        res1,res2,res3,res4=ner_text(line)
        row= prepare_row(file_id,section_id,str(par_counter),res1,res2,res3,res4)
        rows.append(row)

        par_counter+=1

    f.close()

    return rows


def process_txt_folder(folder_name,output_folder):

    total_rows=[]
    counter=0
    for (root, dirs, files) in os.walk(folder_name):
        total=len(files)
        for f in files:

            if  f.endswith('.txt' ):
                print(f)
                print((counter/total)*100)
                try:
                    rows= process_txt_file(os.path.join(root, f),f)
                    total_rows.extend(rows)


                except  Exception as e:
                    print('Error in: '+f+str(e))
                    #break
            counter+=1
            #if counter>4:
                #break
            #break
    write_tsv(os.path.join(output_folder, 'UNE_ner_results_offsets.tsv'),total_rows)


def main(argv):
    input_folder = argv[0]
    output_folder = argv[1]


    process_txt_folder(input_folder,output_folder)


if __name__ == '__main__':
    main(sys.argv[1:])