Jason Thistlethwaite
68 lines
2.8 KiB
Python
68 lines
2.8 KiB
Python
#!/usr/bin/env python3
|
|
"""Training/export helpers for local control tags."""
|
|
|
|
from __future__ import annotations
|
|
|
|
import json
|
|
from pathlib import Path
|
|
from typing import Any, Iterable
|
|
|
|
|
|
def export_control_tag_jsonl(
|
|
*,
|
|
cache_dir: Path,
|
|
tags_payload: dict[str, Any],
|
|
issues_by_id: dict[int, dict[str, Any]],
|
|
tag: str | None = None,
|
|
out_path: Path,
|
|
) -> int:
|
|
"""Write one JSONL record per tagged ticket for later LLM refine / training.
|
|
|
|
Each line: {tag, issue_id, subject, status, assignee, contact_email,
|
|
properties, annotation, text_snippet}
|
|
"""
|
|
out_path = Path(out_path)
|
|
out_path.parent.mkdir(parents=True, exist_ok=True)
|
|
tags = tags_payload.get("tags") if isinstance(tags_payload, dict) else {}
|
|
if not isinstance(tags, dict):
|
|
tags = {}
|
|
names = [tag] if tag else sorted(tags.keys())
|
|
count = 0
|
|
with out_path.open("w", encoding="utf-8") as fh:
|
|
for name in names:
|
|
bucket = tags.get(name) or {}
|
|
tickets = bucket.get("tickets") if isinstance(bucket, dict) else {}
|
|
if not isinstance(tickets, dict):
|
|
continue
|
|
for iid_s, meta in tickets.items():
|
|
try:
|
|
iid = int(iid_s)
|
|
except (TypeError, ValueError):
|
|
continue
|
|
issue = issues_by_id.get(iid) or {}
|
|
helpdesk_raw = issue.get("helpdesk_ticket")
|
|
helpdesk: dict[str, Any] = helpdesk_raw if isinstance(helpdesk_raw, dict) else {}
|
|
contact_raw = helpdesk.get("contact")
|
|
contact: dict[str, Any] = contact_raw if isinstance(contact_raw, dict) else {}
|
|
status_raw = issue.get("status")
|
|
status_obj: dict[str, Any] = status_raw if isinstance(status_raw, dict) else {}
|
|
assignee_raw = issue.get("assigned_to")
|
|
assignee_obj: dict[str, Any] = assignee_raw if isinstance(assignee_raw, dict) else {}
|
|
meta_d = meta if isinstance(meta, dict) else {}
|
|
rec = {
|
|
"tag": name,
|
|
"issue_id": iid,
|
|
"subject": issue.get("subject"),
|
|
"status": status_obj.get("name"),
|
|
"assignee": assignee_obj.get("name"),
|
|
"contact_email": contact.get("email") or helpdesk.get("from_address"),
|
|
"properties": meta_d.get("properties") or [],
|
|
"annotation": meta_d.get("annotation") or "",
|
|
"text_snippet": (
|
|
f"{issue.get('subject') or ''}\n{(issue.get('description') or '')[:500]}"
|
|
).strip(),
|
|
}
|
|
fh.write(json.dumps(rec, sort_keys=True) + "\n")
|
|
count += 1
|
|
return count
|