33 lines
2.3 KiB
Python
33 lines
2.3 KiB
Python
from pathlib import Path
|
||
import re,string
|
||
out=Path('outputs/investmoz-audit-current-20261012');doc=Path('docs/investmoz/v2/auditoria-portal')
|
||
known=['2e2f16c83483','a02ef2300d6c','a23156a65be2','C131','C132','C37','COUNTRY02','ClamAV','InertiaDevTools','PortalProfileAccess','TrimStrings','OpportunityContacts','APP_ENV','READ ONLY']
|
||
pattern=re.compile(r'https?://[^\s)>*]+|`[^`]*`|AUD-[A-Za-z0-9/–\-]+|[A-Za-z0-9_\-/]+\.(?:md|csv|json|php|tsx|docx)|[a-f0-9]{64}|\d+(?:\.\d+){1,3}|(?:stability_consumers_\d+_\d+_test|investemoz):\d+')
|
||
def polish(s):
|
||
original=s;keep={}
|
||
def token(value):
|
||
i=len(keep);name='§KEPT'+''.join(string.ascii_uppercase[int(c)] for c in str(i))+'§';keep[name]=value;return name
|
||
for v in known:s=s.replace(v,token(v))
|
||
s=pattern.sub(lambda m:token(m[0]),s)
|
||
s=s.replace('emdocs/','em docs/').replace('de12/','de 12/').replace('replaymesmo','replay com o mesmo ').replace('exclusivamenteQA','exclusivamente em QA')
|
||
s=re.sub(r'(?<=[A-Za-zÀ-ú])(?=\d)|(?<=\d)(?=[A-Za-zÀ-ú])',' ',s)
|
||
s=re.sub(r'(?<=[a-zà-ú])(?=[A-Z])',' ',s)
|
||
s=re.sub(r'(?<=[A-Za-zÀ-ú0-9])(?=§KEPT)',' ',s)
|
||
s=re.sub(r'(?<=[,;:])(?=[A-Za-zÀ-ú§])',' ',s)
|
||
s=re.sub(r'(?<=[.!?])(?=[A-Z0-9§])',' ',s)
|
||
s=s.replace('§27integral','§27 integral').replace('fonte exacta','fonte exata')
|
||
# Nested protection can occur when a tokenised known identifier is inside a code span.
|
||
for _ in range(4):
|
||
for k,v in keep.items():s=s.replace(k,v)
|
||
assert '§KEPT' not in s
|
||
for v in ['2e2f16c83483','a02ef2300d6c','a23156a65be2']:
|
||
assert original.count(v)==s.count(v)
|
||
return s
|
||
for name in ['relatorio-executivo.md','backlog.md','plano-ativacao.md']:
|
||
p=doc/name;s=p.read_text(encoding='utf-8');end=s.index('<!-- /audit-current-20261012 -->');p.write_text(polish(s[:end])+s[end:],encoding='utf-8')
|
||
for name in ['registo-constrangimentos.md','demonstracao-auditoria-atual.md','consumidores-restantes.md']:
|
||
p=doc/name;p.write_text(polish(p.read_text(encoding='utf-8')),encoding='utf-8')
|
||
p=doc/'casos-reproducao.md';s=p.read_text(encoding='utf-8');i=s.index('## C131 —');p.write_text(s[:i]+polish(s[i:]),encoding='utf-8')
|
||
for p in (out/'para-revisao').glob('*.md'):p.write_text(polish(p.read_text(encoding='utf-8')),encoding='utf-8')
|
||
print('Current prose normalized; historical blocks and CSV untouched.')
|