"""File upload endpoints (/api/upload/*)."""
from __future__ import annotations

import json
import logging
import uuid
from pathlib import Path
from typing import Optional

from fastapi import APIRouter, Cookie, File, Form, HTTPException, Request, UploadFile
from fastapi.responses import JSONResponse, StreamingResponse

from bincio.serve import deps, tasks

log = logging.getLogger("bincio.serve")

router = APIRouter()

_SUPPORTED_SUFFIXES = {".fit", ".gpx", ".tcx", ".fit.gz", ".gpx.gz", ".tcx.gz"}


def _file_suffix(name: str) -> str:
    """Return the effective suffix, including .gz double-extension."""
    p = Path(name.lower())
    if p.suffix == ".gz":
        return p.stem.rsplit(".", 1)[-1].join([".", ".gz"]) if "." in p.stem else ".gz"
    return p.suffix


def _upsert_index_summary(user_dir: Path, activity_id: str, activity: dict, geojson: Optional[dict] = None) -> None:
    """Add or update an activity summary in user_dir/index.json.

    Called after writing BAS activity files so that merge_all can include the
    activity in year shards.  Without this, uploaded activities exist on disk
    but never appear in the browser feed.
    """
    # Build preview coords from geojson if available ([lat, lng] order)
    preview: Optional[list] = None
    if geojson:
        try:
            coords = geojson.get("geometry", {}).get("coordinates", [])
            if coords:
                step = max(1, len(coords) // 9)
                preview = [[c[1], c[0]] for c in coords[::step]][:9]
        except (TypeError, IndexError, AttributeError):
            pass

    has_track = (user_dir / "activities" / f"{activity_id}.geojson").exists()
    summary = {
        "id": activity_id,
        "title": activity.get("title", activity_id),
        "sport": activity.get("sport"),
        "sub_sport": activity.get("sub_sport"),
        "started_at": activity.get("started_at"),
        "distance_m": activity.get("distance_m"),
        "duration_s": activity.get("duration_s"),
        "moving_time_s": activity.get("moving_time_s"),
        "elevation_gain_m": activity.get("elevation_gain_m"),
        "avg_speed_kmh": activity.get("avg_speed_kmh"),
        "max_speed_kmh": activity.get("max_speed_kmh"),
        "avg_hr_bpm": activity.get("avg_hr_bpm"),
        "max_hr_bpm": activity.get("max_hr_bpm"),
        "avg_cadence_rpm": activity.get("avg_cadence_rpm"),
        "avg_power_w": activity.get("avg_power_w"),
        "mmp": activity.get("mmp"),
        "best_efforts": activity.get("best_efforts"),
        "best_climb_m": activity.get("best_climb_m"),
        "source": activity.get("source"),
        "privacy": activity.get("privacy", "public"),
        "detail_url": f"activities/{activity_id}.json",
        "track_url": f"activities/{activity_id}.geojson" if has_track else None,
        "preview_coords": preview,
    }

    index_path = user_dir / "index.json"
    if index_path.exists():
        index_data = json.loads(index_path.read_text(encoding="utf-8"))
    else:
        index_data = {
            "bas_version": "1.0",
            "owner": {"handle": user_dir.name},
            "generated_at": None,
            "activities": [],
        }
    existing = {a["id"]: a for a in index_data.get("activities", [])}
    existing[activity_id] = summary
    index_data["activities"] = sorted(existing.values(), key=lambda a: a.get("started_at", ""), reverse=True)
    index_path.write_text(json.dumps(index_data, indent=2, ensure_ascii=False), encoding="utf-8")


@router.post("/api/upload/bas")
async def upload_bas_activity(
    request: Request,
    bincio_session: Optional[str] = Cookie(default=None),
) -> JSONResponse:
    """Accept a pre-extracted BAS activity JSON from the mobile app.

    Body (JSON):
      activity   – full BAS activity dict (required, must have 'id')
      timeseries – timeseries dict (optional)
      geojson    – GeoJSON dict (optional)

    Returns:
      {"ok": true, "id": "...", "status": "imported" | "duplicate"}
    """
    user = deps._require_auth(request, bincio_session)
    body = await request.json()

    activity = body.get("activity")
    if not activity or not activity.get("id"):
        raise HTTPException(400, "Missing activity.id")

    activity_id = str(activity["id"])
    deps._check_id(activity_id)

    user_dir = deps._get_data_dir() / user.handle
    acts_dir = user_dir / "activities"
    acts_dir.mkdir(parents=True, exist_ok=True)

    out = acts_dir / f"{activity_id}.json"
    if out.exists():
        return JSONResponse({"ok": True, "id": activity_id, "status": "duplicate"})

    out.write_text(json.dumps(activity, ensure_ascii=False, indent=2), encoding="utf-8")

    if body.get("timeseries"):
        ts_path = acts_dir / f"{activity_id}.timeseries.json"
        if not ts_path.exists():
            ts_path.write_text(json.dumps(body["timeseries"], ensure_ascii=False), encoding="utf-8")

    geojson_body: Optional[dict] = body.get("geojson") or None
    if geojson_body:
        gj_path = acts_dir / f"{activity_id}.geojson"
        if not gj_path.exists():
            gj_path.write_text(json.dumps(geojson_body, ensure_ascii=False), encoding="utf-8")

    _upsert_index_summary(user_dir, activity_id, activity, geojson_body)

    try:
        from bincio.render.merge import merge_one, write_combined_feed
        merge_one(user_dir, activity_id)
        write_combined_feed(deps._get_data_dir())
    except Exception as exc:
        log.warning("upload/bas[%s]: merge/feed failed (non-fatal): %s", user.handle, exc)

    log.info("upload/bas[%s]: imported %s", user.handle, activity_id)
    return JSONResponse({"ok": True, "id": activity_id, "status": "imported"})


@router.post("/api/upload/raw")
async def upload_raw_activity(
    request: Request,
    bincio_session: Optional[str] = Cookie(default=None),
) -> JSONResponse:
    """Accept a raw FIT/GPX file (base64-encoded) from the mobile app, extract it
    server-side, store it in the user's activity library, and return the full
    extracted data so the mobile can cache it locally.

    Used when the device WebView is too old to run Pyodide (e.g. Karoo / Chrome <69).

    Body (JSON):
      filename  – original filename (used only to determine file extension)
      base64    – base64-encoded raw file bytes

    Auth: Authorization: Bearer <token>

    Returns:
      {"ok": true, "id": "...", "detail": {...}, "timeseries": {...}|null,
       "geojson": {...}|null, "source_hash": "<sha256-hex>"}
    """
    import base64 as _b64
    import hashlib
    import shutil

    user = deps._require_auth(request, bincio_session)

    body = await request.json()
    filename_hint: str = body.get("filename") or "activity.fit"
    b64: str = body.get("base64") or ""
    user_title: Optional[str] = body.get("user_title") or None
    if not b64:
        raise HTTPException(400, "Missing base64 field")

    try:
        raw = _b64.b64decode(b64)
    except ValueError:
        raise HTTPException(400, "Invalid base64 encoding")

    source_hash = hashlib.sha256(raw).hexdigest()

    suffix = Path(filename_hint).suffix or ".fit"
    tmp_in = Path(f"/tmp/bincio_raw_{uuid.uuid4()}{suffix}")
    tmp_out = Path(f"/tmp/bincio_out_{uuid.uuid4()}")
    try:
        tmp_in.write_bytes(raw)
        tmp_out.mkdir()

        from bincio.extract.parsers.factory import parse_file
        from bincio.extract.metrics import compute
        from bincio.extract.writer import make_activity_id, write_activity
        from bincio.extract.timeseries import build_timeseries

        activity = parse_file(tmp_in)
        metrics = compute(activity)
        write_activity(activity, metrics, tmp_out, privacy="public", rdp_epsilon=0.0001)
        act_id = make_activity_id(activity)

        acts_tmp = tmp_out / "activities"
        detail_path  = acts_tmp / f"{act_id}.json"
        ts_path      = acts_tmp / f"{act_id}.timeseries.json"
        geojson_path = acts_tmp / f"{act_id}.geojson"

        if not ts_path.exists():
            ts_data = build_timeseries(activity.points, activity.started_at, "public")
            if ts_data.get("t"):
                ts_path.write_text(json.dumps(ts_data))

        detail     = json.loads(detail_path.read_text())
        timeseries = json.loads(ts_path.read_text()) if ts_path.exists() else None
        geojson    = json.loads(geojson_path.read_text()) if geojson_path.exists() else None

        # Also store on the server so the activity appears in the user's feed.
        user_dir = deps._get_data_dir() / user.handle
        acts_dir = user_dir / "activities"
        acts_dir.mkdir(parents=True, exist_ok=True)
        out = acts_dir / f"{act_id}.json"
        if not out.exists():
            out.write_text(json.dumps(detail, ensure_ascii=False, indent=2), encoding="utf-8")
        if timeseries and not (acts_dir / f"{act_id}.timeseries.json").exists():
            (acts_dir / f"{act_id}.timeseries.json").write_text(json.dumps(timeseries), encoding="utf-8")
        if geojson and not (acts_dir / f"{act_id}.geojson").exists():
            (acts_dir / f"{act_id}.geojson").write_text(json.dumps(geojson), encoding="utf-8")

        _upsert_index_summary(user_dir, act_id, detail, geojson)

        if user_title:
            import yaml as _yaml
            edits_dir = user_dir / "edits"
            edits_dir.mkdir(parents=True, exist_ok=True)
            (edits_dir / f"{act_id}.md").write_text(
                f"---\n{_yaml.dump({'title': user_title}, allow_unicode=True)}---\n",
                encoding="utf-8",
            )

    except Exception as exc:
        log.warning("upload/raw[%s]: extraction failed: %s", user.handle, exc)
        raise HTTPException(422, f"Could not extract activity: {exc}") from exc
    finally:
        tmp_in.unlink(missing_ok=True)
        shutil.rmtree(tmp_out, ignore_errors=True)

    # Merge and update feed — best effort; a race or transient FS error here must
    # not turn a successful extraction into a 422 (the file is on disk; the mobile
    # would retry indefinitely and the activity would never be marked synced).
    try:
        from bincio.render.merge import merge_one, write_combined_feed
        merge_one(user_dir, act_id)
        write_combined_feed(deps._get_data_dir())
    except Exception as exc:
        log.warning("upload/raw[%s]: merge/feed failed (non-fatal): %s", user.handle, exc)

    log.info("upload/raw[%s]: imported %s", user.handle, act_id)
    return JSONResponse({
        "ok": True,
        "id": act_id,
        "detail": detail,
        "timeseries": timeseries,
        "geojson": geojson,
        "source_hash": source_hash,
    })


@router.post("/api/upload")
async def upload_activity(
    files: list[UploadFile] = File(...),
    store_original: bool = Form(False),
    overwrite: bool = Form(False),
    bincio_session: Optional[str] = Cookie(default=None),
) -> StreamingResponse:
    """Accept FIT/GPX/TCX files and/or activities.csv; stream SSE progress while processing.

    activities.csv (Strava export format) can be included in the batch to:
      - Enrich activity files in the same batch (matched by filename)
      - Retroactively update sidecars for existing activities (matched by strava_id)

    SSE events:
      {"type": "progress", "n": N, "total": T, "name": "...", "status": "imported"|"overwritten"|"duplicate"|"error"}
      {"type": "csv", "updates": N}   -- only when CSV was included
      {"type": "done", "added": N, "csv_updates": N, "duplicates": N, "overwritten": N, "errors": N}
    """
    from bincio.extract.ingest import ingest_parsed
    from bincio.extract.parsers.factory import parse_file
    from bincio.extract.writer import make_activity_id
    from bincio.render.merge import merge_all

    user = deps._require_user(bincio_session)
    dd = deps._get_data_dir() / user.handle
    staging = dd / "_uploads"
    staging.mkdir(exist_ok=True)

    # Read all files into memory now (async), then process synchronously in the generator
    csv_bytes_list: list[bytes] = []
    activity_items: list[tuple[str, bytes]] = []   # (original_filename, bytes)

    for f in files:
        fname = Path(f.filename or "").name
        raw = await f.read()
        if fname.lower().endswith(".csv"):
            csv_bytes_list.append(raw)
        else:
            activity_items.append((fname, raw))

    # Build metadata from the first CSV
    metadata = None
    if csv_bytes_list:
        from bincio.extract.strava_csv import StravaMetadata
        import tempfile
        with tempfile.NamedTemporaryFile(suffix=".csv", delete=False) as tmp:
            tmp.write(csv_bytes_list[0])
            tmp_path = Path(tmp.name)
        try:
            metadata = StravaMetadata(tmp_path)
        finally:
            tmp_path.unlink(missing_ok=True)

    total_files = len(activity_items)
    job_id = tasks._job_start(user.handle, total_files) if total_files > 0 else None

    def event_stream():
        added = 0
        overwritten = 0
        duplicates = 0
        errors = 0
        any_added = False

        for n, (name, contents) in enumerate(activity_items, 1):
            if job_id:
                tasks._job_update(job_id, n - 1, name)

            suffix = _file_suffix(name)
            if suffix not in _SUPPORTED_SUFFIXES:
                errors += 1
                yield f"data: {json.dumps({'type': 'progress', 'n': n, 'total': total_files, 'name': name, 'status': 'error', 'detail': 'unsupported type'})}\n\n"
                continue

            if len(contents) > 50 * 1024 * 1024:
                errors += 1
                yield f"data: {json.dumps({'type': 'progress', 'n': n, 'total': total_files, 'name': name, 'status': 'error', 'detail': 'file too large'})}\n\n"
                continue

            staged = staging / name
            staged.write_bytes(contents)
            kept = False
            try:
                activity = parse_file(staged)
                if metadata is not None:
                    metadata.enrich(name, activity)
                activity_id = make_activity_id(activity)
                was_overwrite = False
                if (dd / "activities" / f"{activity_id}.json").exists():
                    if not overwrite:
                        duplicates += 1
                        yield f"data: {json.dumps({'type': 'progress', 'n': n, 'total': total_files, 'name': name, 'status': 'duplicate'})}\n\n"
                        continue
                    # Overwrite: delete existing files before re-ingesting.
                    for ext in (".json", ".geojson", ".timeseries.json"):
                        (dd / "activities" / f"{activity_id}{ext}").unlink(missing_ok=True)
                    # Remove stale summary from index so ingest_parsed writes a clean one
                    index_path = dd / "index.json"
                    if index_path.exists():
                        idx = json.loads(index_path.read_text(encoding="utf-8"))
                        idx["activities"] = [a for a in idx.get("activities", []) if a.get("id") != activity_id]
                        index_path.write_text(json.dumps(idx, indent=2, ensure_ascii=False))
                    # Remove from dedup hash cache so the new file isn't blocked
                    cache_path = dd / ".bincio_cache.json"
                    if cache_path.exists():
                        try:
                            cache = json.loads(cache_path.read_text(encoding="utf-8"))
                            cache.pop(activity_id, None)
                            cache_path.write_text(json.dumps(cache, ensure_ascii=False))
                        except (OSError, json.JSONDecodeError):
                            pass
                    # Remove merged copies (merge_all will regenerate them after ingest)
                    merged_acts = dd / "_merged" / "activities"
                    if merged_acts.exists():
                        for ext in (".json", ".geojson", ".timeseries.json"):
                            p = merged_acts / f"{activity_id}{ext}"
                            if p.exists() or p.is_symlink():
                                p.unlink(missing_ok=True)
                    was_overwrite = True
                ingest_parsed(activity, dd, privacy="public")
                if store_original:
                    originals_dir = dd / "originals"
                    originals_dir.mkdir(exist_ok=True)
                    staged.rename(originals_dir / name)
                    kept = True
                if was_overwrite:
                    overwritten += 1
                else:
                    added += 1
                any_added = True
                status = 'overwritten' if was_overwrite else 'imported'
                yield f"data: {json.dumps({'type': 'progress', 'n': n, 'total': total_files, 'name': name, 'status': status})}\n\n"
            except Exception as exc:
                errors += 1
                log.error("upload[%s]: failed to process %s: %s", user.handle, name, exc, exc_info=True)
                yield f"data: {json.dumps({'type': 'progress', 'n': n, 'total': total_files, 'name': name, 'status': 'error', 'detail': str(exc)})}\n\n"
            finally:
                if not kept:
                    staged.unlink(missing_ok=True)

        # Retroactively apply CSV metadata to existing activities
        csv_updates = 0
        if metadata is not None:
            from bincio.extract.strava_csv import apply_csv_to_data_dir
            csv_updates = apply_csv_to_data_dir(dd, metadata)
            if csv_updates:
                yield f"data: {json.dumps({'type': 'csv', 'updates': csv_updates})}\n\n"

        if any_added or csv_updates:
            merge_all(dd)
            if any_added:
                tasks._trigger_rebuild(user.handle)

        yield f"data: {json.dumps({'type': 'done', 'added': added, 'overwritten': overwritten, 'csv_updates': csv_updates, 'duplicates': duplicates, 'errors': errors})}\n\n"

        if job_id:
            tasks._job_finish(job_id)

    return StreamingResponse(
        event_stream(),
        media_type="text/event-stream",
        headers={"Cache-Control": "no-cache", "X-Accel-Buffering": "no"},
    )


@router.post("/api/upload/strava-zip")
async def upload_strava_zip(
    file: UploadFile = File(...),
    private: str = Form(default="false"),
    bincio_session: Optional[str] = Cookie(default=None),
) -> StreamingResponse:
    """Accept a Strava bulk export ZIP and stream SSE progress while processing.

    The ZIP is written to a temp file, processed activity-by-activity, then deleted.
    Originals are never kept — the UI informs the user of this upfront.
    """
    user = deps._require_user(bincio_session)
    if not file.filename or not file.filename.lower().endswith(".zip"):
        raise HTTPException(400, "Please upload a .zip file")

    privacy = "unlisted" if private.lower() in ("true", "1", "yes") else "public"

    dd = deps._get_data_dir() / user.handle
    import tempfile
    tmp = tempfile.NamedTemporaryFile(suffix=".zip", delete=False)
    zip_path = Path(tmp.name)
    try:
        while chunk := await file.read(1024 * 1024):  # 1 MB chunks
            tmp.write(chunk)
    finally:
        tmp.close()

    from bincio.extract.strava_zip import strava_zip_iter
    from bincio.render.merge import merge_all

    log.info("strava-zip[%s]: received %s, privacy=%s", user.handle, file.filename, privacy)

    def event_stream():
        any_imported = False
        imported_count = 0
        error_count = 0
        try:
            for event in strava_zip_iter(zip_path, dd, privacy=privacy):
                yield f"data: {json.dumps(event)}\n\n"
                if event.get("type") == "progress":
                    status = event.get("status")
                    if status == "imported":
                        any_imported = True
                        imported_count += 1
                    elif status == "error":
                        error_count += 1
                        log.warning("strava-zip[%s]: error on %s: %s",
                                    user.handle, event.get("name"), event.get("detail", ""))
                if event.get("type") == "done":
                    log.info("strava-zip[%s]: done — imported=%d errors=%d",
                             user.handle, imported_count, error_count)
                    if any_imported:
                        merge_all(dd)
                        try:
                            from bincio.explore import bake_tracks
                            bake_tracks(user.handle, deps._get_data_dir())
                        except Exception as exc:
                            log.warning("strava-zip[%s]: bake_tracks failed (non-fatal): %s", user.handle, exc)
                        tasks._trigger_rebuild(user.handle)
        except Exception as exc:
            log.error("strava-zip[%s]: fatal error: %s", user.handle, exc, exc_info=True)
            yield f"data: {json.dumps({'type': 'error', 'message': str(exc)})}\n\n"
        finally:
            zip_path.unlink(missing_ok=True)

    return StreamingResponse(
        event_stream(),
        media_type="text/event-stream",
        headers={"Cache-Control": "no-cache", "X-Accel-Buffering": "no"},
    )