adulau/napkin-text-analysis
1
0
Fork 0
mirror of https://github.com/adulau/napkin-text-analysis.git synced 2024-12-22 00:26:00 +00:00
Code Issues Projects Releases Packages Wiki Activity

new: [option] to disable parser and/or tagger from the standard processing pipeline of Spacy

Browse source 
If you don't need any of the syntactic information while using napkin,
you can disable parser and tagger. You can gain some memory space and time
for processing. By default, it's still active as napkin might use
of the syntactic information in the future.
This commit is contained in:
Alexandre Dulaunoy 2020-10-11 11:04:30 +02:00
parent ab728e60c6
commit 85044335f4
Signed by: adulau
GPG key ID: 09E2CD4944E6CBCD
2 changed files with 13 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

3
README.md
View file

@ -32,6 +32,7 @@ Intermediate results are stored in a Redis database to allow the analysis of mul
~~~~
usage: napkin.py [-h] [-v V] [-f F] [-t T] [-s] [-o O] [-l L] [--verbatim]
[--no-flushdb] [--binary] [--analysis ANALYSIS]
[--disable-parser] [--disable-tagger]
Extract statistical analysis of text
@ -53,6 +54,8 @@ optional arguments:
--analysis ANALYSIS Limit output to a specific analysis (verb, noun,
hashtag, mention, digit, url, oov, labels, punct).
(Default is all analysis are displayed)
--disable-parser disable parser component in Spacy
--disable-tagger disable tagger component in Spacy
~~~~
# example usage of napkin

12
bin/napkin.py
View file

@ -20,6 +20,8 @@ parser.add_argument('--verbatim', help="Don't use the lemmatized form, use verba
parser.add_argument('--no-flushdb', help="Don't flush the redisdb, useful when you want to process multiple files and aggregate the results. (by default the redis database is flushed at each run)", default=False, action='store_true')
parser.add_argument('--binary', help="set output in binary instead of UTF-8 (default)", default=False, action='store_true')
parser.add_argument('--analysis', help="Limit output to a specific analysis (verb, noun, hashtag, mention, digit, url, oov, labels, punct). (Default is all analysis are displayed)", default='all')
parser.add_argument('--disable-parser', help="disable parser component in Spacy", default=False, action='store_true')
parser.add_argument('--disable-tagger', help="disable tagger component in Spacy", default=False, action='store_true')
args = parser.parse_args()
if args.f is None:
@ -40,10 +42,16 @@ except:
if not args.no_flushdb:
redisdb.flushdb()
disable = []
if args.disable_parser:
disable.append("parser")
if args.disable_tagger:
disable.append("tagger")
if args.l == "fr":
nlp = spacy.load("fr_core_news_md")
nlp = spacy.load("fr_core_news_md", disable=disable)
elif args.l == "en":
nlp = spacy.load("en_core_web_md")
nlp = spacy.load("en_core_web_md", disable=disable)
else:
sys.exit("Language not supported")