SinaTools 0.1.13__py2.py3-none-any.whl → 0.1.14__py2.py3-none-any.whl

{SinaTools-0.1.13.dist-info → SinaTools-0.1.14.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: SinaTools
-Version: 0.1.13
+Version: 0.1.14
 Summary: Open-source Python toolkit for Arabic Natural Understanding, allowing people to integrate it in their system workflow.
 Home-page: https://github.com/SinaLab/sinatools
 License: MIT license

{SinaTools-0.1.13.dist-info → SinaTools-0.1.14.dist-info}/RECORD

@@ -1,5 +1,5 @@
-SinaTools-0.1.13.data/data/sinatools/environment.yml,sha256=OzilhLjZbo_3nU93EQNUFX-6G5O3newiSWrwxvMH2Os,7231
-sinatools/VERSION,sha256=7lGv2l4eJuZteaVLIUnlbwoi4W41EwZ01RPRCjudlCI,6
+SinaTools-0.1.14.data/data/sinatools/environment.yml,sha256=OzilhLjZbo_3nU93EQNUFX-6G5O3newiSWrwxvMH2Os,7231
+sinatools/VERSION,sha256=IMD191UlikGLN9r9ghiHCkd_kk6Nd5gGarueeGgwL-0,6
 sinatools/__init__.py,sha256=bEosTU1o-FSpyytS6iVP_82BXHF2yHnzpJxPLYRbeII,135
 sinatools/environment.yml,sha256=OzilhLjZbo_3nU93EQNUFX-6G5O3newiSWrwxvMH2Os,7231
 sinatools/install_env.py,sha256=EODeeE0ZzfM_rz33_JSIruX03Nc4ghyVOM5BHVhsZaQ,404
@@ -8,7 +8,7 @@ sinatools/CLI/DataDownload/download_files.py,sha256=tkH293ZUSvlvyZClkJmxfNk1x-C3
 sinatools/CLI/morphology/ALMA_multi_word.py,sha256=ZImJ1vtcpSHydI1BjJmK3KcMJbGBZX16kO4L6rxvBvA,2086
 sinatools/CLI/morphology/morph_analyzer.py,sha256=ieIM47QK9Nct3MtCS9uq3h2rZN5r4qNhsLmlVeE6wiE,3503
 sinatools/CLI/ner/corpus_entity_extractor.py,sha256=_o0frMSgpsFVXPoztS3mQTK7LjHsgzUv9gfs6iJL424,4024
-sinatools/CLI/ner/entity_extractor.py,sha256=QFGkavZz8ZZGetMTXiTH_OeoN9B2Iyx60EKCYdFtoDY,2811
+sinatools/CLI/ner/entity_extractor.py,sha256=zn0Jd37BEDE1wHE5HOAK0_N2tURAznFNj7WDd6WGLIw,2932
 sinatools/CLI/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 sinatools/CLI/utils/arStrip.py,sha256=NLyp8vOu2xv80tL9jiKRvyptmbkRZVg-wcAr-9YyvNY,3264
 sinatools/CLI/utils/corpus_tokenizer.py,sha256=nH0T4h6urr_0Qy6-wN3PquOtnwybj0REde5Ts_OE4U8,1650
@@ -80,7 +80,7 @@ sinatools/ner/__init__.py,sha256=gSs0x6veWJ8j3_iOs79tynBd_hJP0t44CGpJ0xzoiW4,104
 sinatools/ner/data.py,sha256=lvOW86dXse8SC75Q0supQaE0rrRffoxNjIA0Qbv5WZY,4354
 sinatools/ner/data_format.py,sha256=7Yt0aOicOn9_YuuyCkM_IYi_rgjGYxR9bCuUaNGM73o,4341
 sinatools/ner/datasets.py,sha256=mG1iwqSm3lXCFHLqE-b4wNi176cpuzNBz8tKaBU6z6M,5059
-sinatools/ner/entity_extractor.py,sha256=fwaPmbg3RaohQu9uu9rMXlvamCnv3am1EYjQMG6tuyY,2270
+sinatools/ner/entity_extractor.py,sha256=Stuemcv4uDBI-SfcefEkGMDcruZQhVm63f78lshwaF8,2271
 sinatools/ner/helpers.py,sha256=dnOoDY5JMyOLTUWVIZLMt8mBn2IbWlVaqHhQyjs1voo,2343
 sinatools/ner/metrics.py,sha256=Irz6SsIvpOzGIA2lWxrEV86xnTnm0TzKm9SUVT4SXUU,2734
 sinatools/ner/transforms.py,sha256=vti3mDdi-IRP8i0aTQ37QqpPlP9hdMmJ6_bAMa0uL-s,4871
@@ -113,10 +113,10 @@ sinatools/wsd/__init__.py,sha256=5Ondsp-Xe9YxVjRlTc4nLrxu6xiyML7B3bQ3EZ44uEM,327
 sinatools/wsd/disambiguator.py,sha256=BUiIXLd8b9tdZqThBiwacfSZtTkRx9LNnqegibmlbFA,20008
 sinatools/wsd/settings.py,sha256=b_AqTxVWALuGXnsMd9KhnnwIo9-JEoWOTekB-7_xJCU,1111
 sinatools/wsd/wsd.py,sha256=gHIBUFXegoY1z3rRnIlK6TduhYq2BTa_dHakOjOlT4k,4434
-SinaTools-0.1.13.dist-info/AUTHORS.rst,sha256=aTWeWlIdfLi56iLJfIUAwIrmqDcgxXKLji75_Fjzjyg,174
-SinaTools-0.1.13.dist-info/LICENSE,sha256=uwsKYG4TayHXNANWdpfMN2lVW4dimxQjA_7vuCVhD70,1088
-SinaTools-0.1.13.dist-info/METADATA,sha256=OAlD6n0C6DAbu4Kf9Foys8qSgyPzcwk-jwkklU6QkzA,953
-SinaTools-0.1.13.dist-info/WHEEL,sha256=6T3TYZE4YFi2HTS1BeZHNXAi8N52OZT4O-dJ6-ome_4,116
-SinaTools-0.1.13.dist-info/entry_points.txt,sha256=ZwZLolnWog2fjdDrfaHNHob8SE_YtMbD6ayzsOzItxs,1234
-SinaTools-0.1.13.dist-info/top_level.txt,sha256=8tNdPTeJKw3TQCaua8IJIx6N6WpgZZmVekf1OdBNJpE,10
-SinaTools-0.1.13.dist-info/RECORD,,
+SinaTools-0.1.14.dist-info/AUTHORS.rst,sha256=aTWeWlIdfLi56iLJfIUAwIrmqDcgxXKLji75_Fjzjyg,174
+SinaTools-0.1.14.dist-info/LICENSE,sha256=uwsKYG4TayHXNANWdpfMN2lVW4dimxQjA_7vuCVhD70,1088
+SinaTools-0.1.14.dist-info/METADATA,sha256=FAA_sKla3nCg0ZJWQWzpGyL3Q7SPvFviNtONqqIMbqE,953
+SinaTools-0.1.14.dist-info/WHEEL,sha256=6T3TYZE4YFi2HTS1BeZHNXAi8N52OZT4O-dJ6-ome_4,116
+SinaTools-0.1.14.dist-info/entry_points.txt,sha256=ZwZLolnWog2fjdDrfaHNHob8SE_YtMbD6ayzsOzItxs,1234
+SinaTools-0.1.14.dist-info/top_level.txt,sha256=8tNdPTeJKw3TQCaua8IJIx6N6WpgZZmVekf1OdBNJpE,10
+SinaTools-0.1.14.dist-info/RECORD,,

sinatools/CLI/ner/entity_extractor.py

@@ -42,9 +42,11 @@ from sinatools.ner.entity_extractor import extract
 from sinatools.utils.tokenizer import corpus_tokenizer
 from sinatools.utils.tokenizers_words import simple_word_tokenize
 
+def jsons_to_list_of_lists(json_list):
+    return [[d['token'], d['tags']] for d in json_list]
 
 def combine_tags(sentence):
-    output = extract(sentence)
+    output = jsons_to_list_of_lists(extract(sentence))
     return [word[1] for word in output]
 
 

sinatools/VERSION

@@ -1 +1 @@
-0.1.13
+0.1.14

sinatools/ner/entity_extractor.py

@@ -61,7 +61,7 @@ def extract(text, batch_size=32):
     for segment in segments:
         for token in segment:
             segments_list = {}
-            segments_list["word"] = token.text
+            segments_list["token"] = token.text
             list_of_tags = [t['tag'] for t in token.pred_tag]
             list_of_tags = [i for i in list_of_tags if i not in('O',' ','')]
             if list_of_tags == []: