mirror of
https://github.com/kevinveenbirkenbach/computer-playbook.git
synced 2026-11-08 11:46:35 +00:00
kevinveenbirkenbachandClaude Opus 5
636caf7fab
The script read sys.argv[1] and opened a zip at module level, so importing it ran it. autodoc imports to document, was handed sphinx's own '-b' as the path, and died with FileNotFoundError. It now has main() behind the __main__ guard and holds the archive in a with block. The URL scan appeared twice verbatim, once per trace member, and is one function called in a loop. That also drops a bug of the original: its 'if len(seen) >= 50: break' sat outside the membership check, so a duplicate hit ended the scan early. Invocation is unchanged; a missing argument now prints the usage and exits 2 instead of raising IndexError. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
61 lines
1.6 KiB
Python
61 lines
1.6 KiB
Python
"""Print the files and the URLs a Playwright trace archive holds.
|
|
|
|
Usage:
|
|
python -m utils.parse_trace <trace.zip>
|
|
"""
|
|
|
|
import re
|
|
import sys
|
|
import zipfile
|
|
|
|
URL = re.compile(r'https?://[^\s"\\]+')
|
|
LIMIT = 50
|
|
TRACE_SUFFIXES = (".trace", ".network")
|
|
|
|
|
|
def urls(archive: zipfile.ZipFile, name: str) -> list[str]:
|
|
"""Return up to ``LIMIT`` distinct URLs found in one member of ``archive``.
|
|
|
|
Args:
|
|
archive: the opened trace archive.
|
|
name: the member to scan.
|
|
"""
|
|
if name not in archive.namelist():
|
|
return []
|
|
text = archive.read(name).decode(errors="replace")
|
|
seen: set[str] = set()
|
|
for match in URL.finditer(text):
|
|
seen.add(match.group(0).rstrip('",:;)'))
|
|
if len(seen) >= LIMIT:
|
|
break
|
|
return sorted(seen)
|
|
|
|
|
|
def main(argv: list[str] | None = None) -> int:
|
|
"""Print the archive's trace members and the URLs they mention.
|
|
|
|
Args:
|
|
argv: command line arguments; ``sys.argv[1:]`` when omitted.
|
|
"""
|
|
args = sys.argv[1:] if argv is None else argv
|
|
if len(args) != 1:
|
|
print(__doc__.strip(), file=sys.stderr)
|
|
return 2
|
|
|
|
with zipfile.ZipFile(args[0]) as archive:
|
|
print("==== files in trace ====")
|
|
for name in archive.namelist():
|
|
if name.endswith(TRACE_SUFFIXES):
|
|
print(name, archive.getinfo(name).file_size)
|
|
|
|
for member in ("0-trace.network", "0-trace.trace"):
|
|
print()
|
|
print(f"==== url-like strings from {member} (first {LIMIT}) ====")
|
|
for url in urls(archive, member):
|
|
print(url)
|
|
return 0
|
|
|
|
|
|
if __name__ == "__main__":
|
|
raise SystemExit(main())
|