mirror of
https://github.com/MadsLorentzen/ai-job-search.git
synced 2026-11-02 13:12:50 +00:00
30 lines
1.3 KiB
Python
30 lines
1.3 KiB
Python
|
"""Every seen_jobs.json reference in /scrape's spec names the full path.
|
||
|
|
|
||
|
|
Step 0 reads `job_scraper/seen_jobs.json`, but later steps (notably Step 4,
|
||
|
|
where the file is written) used the bare filename. An agent following Step 4
|
||
|
|
literally can create `seen_jobs.json` in the repo root. The ignore rules only
|
||
|
|
cover `**/job_scraper/seen_jobs.json`, so a root-level copy is untracked-by-rule
|
||
|
|
only by accident: it shows up in `git status`, gets swept into a commit, and
|
||
|
|
publishes the user's scrape history. Pinning the full path in the spec keeps
|
||
|
|
the read path, the write path and the ignore rule pointing at the same file.
|
||
|
|
"""
|
||
|
|
import re
|
||
|
|
import unittest
|
||
|
|
from pathlib import Path
|
||
|
|
|
||
|
|
SKILL = Path(__file__).resolve().parent.parent / ".claude" / "skills" / "job-scraper" / "SKILL.md"
|
||
|
|
BARE = re.compile(r"(?<!job_scraper/)(?<![\w/.-])seen_jobs\.json")
|
||
|
|
|
||
|
|
|
||
|
|
class ScrapeSeenJobsPathSpec(unittest.TestCase):
|
||
|
|
def test_no_bare_seen_jobs_reference(self):
|
||
|
|
offenders = [
|
||
|
|
f"line {i}: {line.strip()[:90]}"
|
||
|
|
for i, line in enumerate(SKILL.read_text(encoding="utf-8").splitlines(), 1)
|
||
|
|
if BARE.search(line)
|
||
|
|
]
|
||
|
|
self.assertEqual(offenders, [], "use job_scraper/seen_jobs.json everywhere in the spec")
|
||
|
|
|
||
|
|
|
||
|
|
if __name__ == "__main__":
|
||
|
|
unittest.main()
|