Mercurial > hg > cc > azure
annotate master/src/wecu/sac_mapper.py @ 60:5fdca5baa4e9
refactor a bit, add support for sac with bespoke mapper
author | Henry S. Thompson <ht@markup.co.uk> |
---|---|
date | Thu, 28 May 2020 12:55:03 +0000 |
parents | a3edba8dab11 |
children | cfaf5223b071 |
rev | line source |
---|---|
57
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
1 #!/usr/bin/python3 |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
2 import os |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
3 import sys |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
4 import re |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
5 |
60
5fdca5baa4e9
refactor a bit, add support for sac with bespoke mapper
Henry S. Thompson <ht@markup.co.uk>
parents:
58
diff
changeset
|
6 print('args',sys.argv) |
5fdca5baa4e9
refactor a bit, add support for sac with bespoke mapper
Henry S. Thompson <ht@markup.co.uk>
parents:
58
diff
changeset
|
7 |
57
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
8 is_regex = sys.argv[1] == 'true' |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
9 search_terms = sys.argv[2:] |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
10 search_terms_counters = dict() |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
11 |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
12 # Initialise |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
13 for term in search_terms: |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
14 search_terms_counters[term] = 0 |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
15 |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
16 if not is_regex: |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
17 for line in sys.stdin: |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
18 for term in search_terms_counters: |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
19 search_terms_counters[term] += line.count(term) |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
20 else: |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
21 for line in sys.stdin: |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
22 for term_regex in search_terms_counters: |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
23 search_terms_counters[term_regex] += len(re.findall(term_regex, line)) |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
24 |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
25 |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
26 for term in search_terms_counters: |
ac1a20e627a9
from lukasz git repo 2020-05-26 (see ~/src/wecu), then editted,
Henry S. Thompson <ht@markup.co.uk>
parents:
diff
changeset
|
27 print("{}\t{}".format(term, search_terms_counters[term])) |