Files

68 lines
2.8 KiB
Python

#!/usr/bin/env python3
import json, os, re, time
from datetime import datetime,timezone
from pathlib import Path
from urllib.parse import urlencode, quote
from common import request,store,state_file
ROOT=Path(__file__).resolve().parents[1]
API=os.getenv('ROCKY_API','https://apollo.build.resf.org/v2/advisories').rstrip('/')
PAGE_SIZE=int(os.getenv('ROCKY_PAGE_SIZE','100'))
MAX_PAGES=int(os.getenv('ROCKY_MAX_PAGES','20'))
DELAY=float(os.getenv('ROCKY_DELAY','0.15'))
STATE=ROOT/'state/sync.json'
def get_advisories(body):
obj=json.loads(body)
if not isinstance(obj,dict) or not isinstance(obj.get('advisories'),list):
raise ValueError('Unexpected Apollo list response')
total=int(obj['total']) if 'total' in obj else None
if total is None: raise ValueError('Apollo response lacks total count')
return obj['advisories'],total
def identity(doc):
name=doc.get('name') or doc.get('id')
if not isinstance(name,str) or not re.fullmatch(r'[A-Za-z0-9:._-]+',name):
raise ValueError(f'Invalid advisory identity {name!r}')
return name
def date_filter(doc):
return identity(doc).startswith(('RLSA','RXSA'))
def main():
old=json.loads(STATE.read_text()) if STATE.exists() else {}
page=max(1,int(old.get('next_page',1)))
pages=0
while pages<MAX_PAGES:
q=urlencode({'page':page,'limit':PAGE_SIZE,'filters.type':'TYPE_SECURITY','filters.includeRpms':'false'})
_,body,_=request(API+'?'+q)
entries,total=get_advisories(body)
if not entries:
if page==1: raise RuntimeError('Apollo returned zero advisories')
page=1
break
count=0
for summary in entries:
if not date_filter(summary): continue
name=identity(summary)
# Individual API responses retain affected RPM mappings. Do not use only list summaries.
url=API+'/'+quote(name,safe='')
_,detail,_=request(url)
parsed=json.loads(detail)
advisory=parsed.get('advisory') if isinstance(parsed,dict) else None
if not isinstance(advisory,dict) or identity(advisory)!=name:
raise ValueError(f'Invalid detail for {name}')
year=re.search(r'(20\d\d)',name)
if not year: raise ValueError(f'No year for {name}')
dest=ROOT/'data/advisories'/year.group(1)/(name.replace(':','-')+'.json')
store(dest,detail)
count+=1
time.sleep(DELAY)
print(f'Rocky page {page}: {len(entries)} indexed, {count} details, total={total}',flush=True)
page+=1; pages+=1
if page>(total+PAGE_SIZE-1)//PAGE_SIZE: page=1
state_file(STATE,{'next_page':page,'last_success_utc':datetime.now(timezone.utc).isoformat(),'last_total':total})
if page==1: break
if __name__=='__main__': main()