"""Export group issues from git.lab to analysis/data/gitlab_issues.json.

Metadata only -- titles, labels, dates, state -- never descriptions or
comment bodies. The token is read from the path the management repo's
CLAUDE.md sanctions (~/.config/gitlab-lab/token) and is never printed,
logged or written anywhere.

All issues of the group are exported, open and closed alike: the group is
small enough that a full export beats an arbitrary "recently closed"
cutoff, and a full export is trivially reproducible. Issues of projects
the human placed out of scope are exported too and tagged
in_scope=false -- they are evidence about where work lives.

Skipped without failing the run when the token file is missing or git.lab
is unreachable; run_all.sh then proceeds without the ticket dimension.
"""

import json
import sys
import urllib.error
import urllib.request
from pathlib import Path

from common import (
    COMPONENTS, DATA_DIR, GITLAB_HOST, GROUP, OUT_OF_SCOPE_PROJECTS,
)

TOKEN_FILE = Path.home() / ".config" / "gitlab-lab" / "token"
IN_SCOPE = {slug for slug, _ in COMPONENTS} | {"management"}

# Exactly the fields the analysis needs. Anything else -- above all
# `description` -- is dropped before it reaches disk.
FIELDS = ["id", "iid", "project_id", "title", "state", "labels",
          "created_at", "updated_at", "closed_at", "due_date"]


def api(token, path):
    """Yield every page of a GitLab list endpoint."""
    url = f"{GITLAB_HOST}/api/v4/{path}"
    while url:
        req = urllib.request.Request(url, headers={"PRIVATE-TOKEN": token})
        with urllib.request.urlopen(req, timeout=60) as resp:
            yield from json.load(resp)
            url = None
            for link in resp.headers.get("Link", "").split(","):
                if 'rel="next"' in link:
                    url = link.split(";")[0].strip().strip("<>")


def slim(issue, projects):
    out = {k: issue.get(k) for k in FIELDS}
    out["labels"] = sorted(out["labels"] or [])
    out["project"] = projects.get(issue["project_id"], str(issue["project_id"]))
    out["in_scope"] = out["project"] in IN_SCOPE
    milestone = issue.get("milestone")
    out["milestone"] = milestone["title"] if milestone else None
    assignee = issue.get("assignee")
    out["assignee"] = assignee["username"] if assignee else None
    return out


def main():
    if not TOKEN_FILE.exists():
        print(f"  gitlab_issues.json: SKIPPED, no token at {TOKEN_FILE}")
        return 0
    token = TOKEN_FILE.read_text().strip()

    try:
        projects = {
            p["id"]: p["path"]
            for p in api(token, f"groups/{GROUP}/projects?per_page=100&archived=false")
        }
        issues = list(api(token, f"groups/{GROUP}/issues?per_page=100&state=all&scope=all"))
    except (urllib.error.URLError, TimeoutError) as exc:
        print(f"  gitlab_issues.json: SKIPPED, git.lab unreachable ({exc.reason})")
        return 0

    payload = {
        "source": f"{GITLAB_HOST}/api/v4/groups/{GROUP}/issues?state=all",
        "note": "metadata only; descriptions and comments deliberately not exported",
        "projects": {
            path: {"id": pid, "in_scope": path in IN_SCOPE,
                   "declared_out_of_scope": path in OUT_OF_SCOPE_PROJECTS}
            for pid, path in sorted(projects.items(), key=lambda kv: kv[1])
        },
        "issues": sorted((slim(i, projects) for i in issues), key=lambda i: i["id"]),
    }

    out = DATA_DIR / "gitlab_issues.json"
    DATA_DIR.mkdir(parents=True, exist_ok=True)
    out.write_text(json.dumps(payload, indent=2, ensure_ascii=False,
                              sort_keys=True) + "\n", encoding="utf-8")
    print(f"  gitlab_issues.json: {len(payload['issues'])} issues, "
          f"{len(projects)} projects")
    return 0


if __name__ == "__main__":
    sys.exit(main())