daybook/artifacts/views.py

import hashlib
import html
import json
import mimetypes
import re
import secrets
from datetime import timedelta
from pathlib import Path
from urllib.parse import urlencode

import markdown
from django.conf import settings
from django.contrib import messages
from django.contrib.auth import views as auth_views
from django.contrib.auth.decorators import login_required
from django.db.models import Count, Q
from django.http import FileResponse, Http404, HttpResponse, HttpResponseForbidden, JsonResponse
from django.shortcuts import get_object_or_404, redirect, render
from django.template.defaultfilters import slugify
from django.urls import reverse
from django.utils.dateparse import parse_date
from django.utils import timezone
from django.utils.decorators import method_decorator
from django.views.decorators.clickjacking import xframe_options_sameorigin
from django.views.decorators.http import require_POST
from rest_framework.authtoken.models import Token
from rest_framework.parsers import FormParser, JSONParser, MultiPartParser
from rest_framework.response import Response
from rest_framework.views import APIView

from .demo_data import create_demo_artifacts
from .forms import DaybookLoginForm
from .models import Artifact, TagColor


TEXT_KINDS = {Artifact.HTML_FRAGMENT, Artifact.HTML_FULL, Artifact.MARKDOWN, Artifact.DECK}

REMEMBER_ME_SESSION_AGE = 60 * 60 * 24 * 30  # 30 days


def filter_by_label(queryset, label):
    matching_ids = [artifact.id for artifact in queryset.only("id", "tags") if label in artifact.tags]
    return queryset.filter(pk__in=matching_ids)


DEFAULT_THEMES = {
    Artifact.HTML_FRAGMENT: "report",
    Artifact.MARKDOWN: "report",
    Artifact.DECK: "deck",
}


@method_decorator(login_required, name="get")
class ArtifactListCreateView(APIView):
    parser_classes = [JSONParser, FormParser, MultiPartParser]

    def get(self, request):
        artifacts = Artifact.objects.filter(owner=request.user)
        if source := request.query_params.get("source"):
            artifacts = artifacts.filter(source=source)
        if label := request.query_params.get("label"):
            artifacts = filter_by_label(artifacts, label)
        if day := request.query_params.get("day"):
            artifacts = artifacts.filter(day=day)
        if kind := request.query_params.get("kind"):
            artifacts = artifacts.filter(kind=kind)
        if title_query := request.query_params.get("q"):
            artifacts = artifacts.filter(title__icontains=title_query)

        return Response([serialize_artifact(artifact, request) for artifact in artifacts])

    def post(self, request):
        user = resolve_token_user(request)
        if user is None:
            return HttpResponseForbidden("Invalid or missing API token")

        if request.content_type and request.content_type.startswith("multipart/"):
            artifact = create_binary_artifact(request.data, user)
            if isinstance(artifact, Response):
                return artifact
            apply_tag_colors(request.data)
            return Response(serialize_artifact(artifact, request), status=201)

        data = request.data
        kind = data.get("kind")
        if kind not in TEXT_KINDS:
            return Response({"detail": "kind must be html_fragment, html_full, markdown, or deck"}, status=400)

        required = ["source", "title", "content"]
        missing = [field for field in required if not data.get(field)]
        if missing:
            return Response({"detail": f"Missing required field(s): {', '.join(missing)}"}, status=400)

        source_error = validate_source_fields(data)
        if source_error:
            return source_error

        artifact = create_text_artifact(data, request, user)
        apply_tag_colors(data)
        return Response(serialize_artifact(artifact, request), status=201)


@method_decorator(login_required, name="get")
class ArtifactDetailView(APIView):
    def get(self, request, artifact_id):
        artifact = get_object_or_404(Artifact, pk=artifact_id, owner=request.user)
        return Response(serialize_artifact(artifact, request))

    def patch(self, request, artifact_id):
        user = resolve_token_user(request)
        if user is None:
            return HttpResponseForbidden("Invalid or missing API token")

        artifact = get_object_or_404(Artifact, pk=artifact_id, owner=user)
        data = request.data
        if "title" in data:
            if not data["title"]:
                return Response({"detail": "title cannot be blank"}, status=400)
            artifact.title = data["title"]
        if "tags" in data:
            artifact.tags = normalize_tags(data["tags"])
        if "source_prompt" in data:
            artifact.source_prompt = data["source_prompt"] or ""
        if "source_type" in data:
            if data["source_type"] not in {choice[0] for choice in Artifact.SOURCE_TYPE_CHOICES}:
                return Response({"detail": "source_type must be agent, person, or system"}, status=400)
            artifact.source_type = data["source_type"]
        if "source_name" in data:
            artifact.source_name = data["source_name"] or ""
        if "source_id" in data:
            artifact.source_id = data["source_id"] or ""
        if "metadata" in data:
            if not isinstance(data["metadata"], dict):
                return Response({"detail": "metadata must be an object"}, status=400)
            artifact.metadata = data["metadata"]
        if "tag_colors" in data:
            try:
                apply_tag_colors(data)
            except ValueError as exc:
                return Response({"detail": str(exc)}, status=400)
        artifact.save(
            update_fields=[
                "title",
                "tags",
                "source_prompt",
                "source_type",
                "source_name",
                "source_id",
                "metadata",
            ]
        )
        return Response(serialize_artifact(artifact, request))


class CapabilitiesView(APIView):
    """Static capability info. Reachable by a logged-in session or a valid API token,
    so agents can check connectivity before posting. Exposes no user data."""

    def get(self, request):
        if not request.user.is_authenticated and resolve_token_user(request) is None:
            return HttpResponseForbidden("Login or a valid API token is required")
        return Response(
            {
                "artifact_kinds": [choice[0] for choice in Artifact.KIND_CHOICES],
                "filters": ["source", "label", "day", "kind", "q"],
                "views": ["rendered", "raw"],
                "raw_supported_for": [Artifact.MARKDOWN, Artifact.HTML_FULL],
                "source_types": [choice[0] for choice in Artifact.SOURCE_TYPE_CHOICES],
                "auth": "bearer_token",
            }
        )


FEED_WINDOW_DAYS = 7


def apply_feed_filters(queryset, source="", label="", kind="", title_query=""):
    if source:
        queryset = queryset.filter(source=source)
    if label:
        queryset = filter_by_label(queryset, label)
    if kind:
        queryset = queryset.filter(kind=kind)
    if title_query:
        queryset = queryset.filter(title__icontains=title_query)
    return queryset


def group_artifacts_by_day(artifacts, overrides):
    """Turn a flat, reverse-chronological list into [{day, artifacts}] groups."""
    groups = []
    for artifact in artifacts:
        # Source has its own column in the row, so chips are the free-form tags only.
        artifact.label_chips = tags_with_colors(artifact.tags, overrides)
        if not groups or groups[-1]["day"] != artifact.day:
            groups.append({"day": artifact.day, "artifacts": []})
        groups[-1]["artifacts"].append(artifact)
    return groups


@login_required
def feed(request):
    selected_source = request.GET.get("source", "")
    selected_label = request.GET.get("label", "")
    selected_kind = request.GET.get("kind", "")
    title_query = request.GET.get("q", "")
    today = timezone.localdate()
    selected_artifact_id = request.GET.get("artifact", "")
    preview_none = request.GET.get("preview") == "none"
    selected_view = "raw" if request.GET.get("view") == "raw" else "rendered"

    all_artifacts = Artifact.objects.filter(owner=request.user)
    artifacts = apply_feed_filters(
        all_artifacts, selected_source, selected_label, selected_kind, title_query
    )

    window_start = today - timedelta(days=FEED_WINDOW_DAYS - 1)
    if not artifacts.filter(day__gte=window_start).exists():
        # Nothing recent for these filters: fall back to the most recent days that
        # do match, so selecting a source is never met with a blank list.
        recent_days = list(artifacts.dates("day", "day", order="DESC")[:FEED_WINDOW_DAYS])
        if recent_days:
            window_start = recent_days[-1]
    window_artifacts = list(artifacts.filter(day__gte=window_start).order_by("-created_at"))
    older_count = artifacts.filter(day__lt=window_start).count()

    overrides = tag_color_overrides()
    day_groups = group_artifacts_by_day(window_artifacts, overrides)

    selected_artifact = None
    if not preview_none:
        selected_artifact = next((a for a in window_artifacts if a.id == selected_artifact_id), None)

    toggle_query = {"view": "rendered"}
    if selected_source:
        toggle_query["source"] = selected_source
    if selected_label:
        toggle_query["label"] = selected_label
    if selected_kind:
        toggle_query["kind"] = selected_kind
    if title_query:
        toggle_query["q"] = title_query
    if selected_artifact:
        toggle_query["artifact"] = selected_artifact.id
    rendered_url = f"?{urlencode(toggle_query)}"
    toggle_query["view"] = "raw"
    raw_url = f"?{urlencode(toggle_query)}"

    # Source counts come from the other filters only, so the sidebar still shows
    # every source available under the current label/type/name filters.
    source_scoped = apply_feed_filters(
        all_artifacts, "", selected_label, selected_kind, title_query
    )
    source_counts = (
        source_scoped.values("source")
        .annotate(count=Count("id", filter=Q(read_at__isnull=True)))
        .order_by("source")
    )

    all_labels = sorted({tag for tags in all_artifacts.values_list("tags", flat=True) for tag in tags})
    return render(
        request,
        "artifacts/feed.html",
        {
            "artifacts": window_artifacts,
            "day_groups": day_groups,
            "older_count": older_count,
            "window_start": window_start,
            "source_counts": source_counts,
            "total_count": source_scoped.filter(read_at__isnull=True).count(),
            "all_labels": labels_with_colors_list(all_labels, overrides),
            "raw_url": raw_url,
            "rendered_url": rendered_url,
            "selected_artifact": selected_artifact,
            "selected_kind": selected_kind,
            "selected_label": selected_label,
            "selected_source": selected_source,
            "selected_view": selected_view,
            "preview_none": preview_none,
            "today": today,
            "title_query": title_query,
            "kind_choices": Artifact.KIND_CHOICES,
        },
    )


@login_required
@require_POST
def mark_artifact_read(request, artifact_id):
    artifact = get_object_or_404(Artifact, pk=artifact_id, owner=request.user)
    if artifact.read_at is None:
        artifact.read_at = timezone.now()
        artifact.save(update_fields=["read_at"])
    return JsonResponse({"id": artifact.id, "read_at": artifact.read_at.isoformat()})


@login_required
def feed_older(request):
    """Return the next page of older day-groups as an HTML fragment for infinite scroll."""
    before = parse_date(request.GET.get("before", "")) or timezone.localdate()
    artifacts = apply_feed_filters(
        Artifact.objects.filter(owner=request.user),
        request.GET.get("source", ""),
        request.GET.get("label", ""),
        request.GET.get("kind", ""),
        request.GET.get("q", ""),
    ).filter(day__lt=before).order_by("-created_at")

    # Take a whole page of days at a time so a day is never split across pages.
    page_days = list(artifacts.dates("day", "day", order="DESC")[:FEED_WINDOW_DAYS])
    if not page_days:
        return HttpResponse("")

    page_artifacts = list(artifacts.filter(day__gte=page_days[-1]))
    day_groups = group_artifacts_by_day(page_artifacts, tag_color_overrides())
    remaining = artifacts.filter(day__lt=page_days[-1]).count()

    return render(
        request,
        "artifacts/_day_groups.html",
        {
            "day_groups": day_groups,
            "next_before": page_days[-1].isoformat(),
            "has_more": remaining > 0,
            "selected_artifact": None,
            "selected_source": request.GET.get("source", ""),
            "selected_label": request.GET.get("label", ""),
            "selected_kind": request.GET.get("kind", ""),
            "title_query": request.GET.get("q", ""),
            "selected_view": "raw" if request.GET.get("view") == "raw" else "rendered",
            "today": timezone.localdate(),
        },
    )


@login_required
@xframe_options_sameorigin
def artifact_view(request, artifact_id):
    artifact = get_object_or_404(Artifact, pk=artifact_id, owner=request.user)
    if request.GET.get("raw") == "1":
        raw_path = raw_artifact_path(artifact)
        if raw_path and raw_path.exists():
            return FileResponse(raw_path.open("rb"), content_type="text/plain; charset=utf-8")
        return HttpResponse(raw_unavailable_html(artifact), content_type="text/html")

    path = safe_artifact_path(artifact.path)
    if not path.exists():
        raise Http404("Artifact file not found")

    content_type, _ = mimetypes.guess_type(path.name)
    return FileResponse(path.open("rb"), content_type=content_type or "application/octet-stream")


def create_text_artifact(data, request, user):
    now = timezone.localtime()
    artifact_id = generate_id()
    source = slugify(data["source"]) or "unknown"
    kind = data["kind"]
    theme = data.get("theme") or DEFAULT_THEMES.get(kind, "base")
    title = data["title"]
    content = data["content"]

    directory = Path(source) / now.date().isoformat()
    target_dir = settings.ARTIFACTS_ROOT / directory
    target_dir.mkdir(parents=True, exist_ok=True)

    if kind == Artifact.HTML_FULL:
        rendered = content
    elif kind == Artifact.MARKDOWN:
        rendered = wrap_html(markdown.markdown(content, extensions=["extra", "sane_lists"]), title, theme, request)
        (target_dir / f"{artifact_id}.md").write_text(content, encoding="utf-8")
    elif kind == Artifact.DECK:
        rendered = wrap_deck(content, title, request)
    else:
        rendered = wrap_html(content, title, theme, request)

    stored_path = directory / f"{artifact_id}.html"
    (settings.ARTIFACTS_ROOT / stored_path).write_text(rendered, encoding="utf-8")

    return Artifact.objects.create(
        id=artifact_id,
        owner=user,
        source=source,
        title=title,
        kind=kind,
        theme=theme,
        day=now.date(),
        path=stored_path.as_posix(),
        original_kind=kind,
        tags=normalize_tags(data.get("tags", [])),
        source_prompt=data.get("source_prompt", ""),
        **source_fields(data, source),
    )


def create_binary_artifact(data, user):
    uploaded = data.get("file")
    required = ["source", "title"]
    missing = [field for field in required if not data.get(field)]
    if not uploaded:
        missing.append("file")
    if missing:
        return Response({"detail": f"Missing required field(s): {', '.join(missing)}"}, status=400)
    source_error = validate_source_fields(data)
    if source_error:
        return source_error

    now = timezone.localtime()
    artifact_id = generate_id()
    source = slugify(data["source"]) or "unknown"
    title = data["title"]
    content_type = getattr(uploaded, "content_type", "") or mimetypes.guess_type(uploaded.name)[0] or ""
    kind = binary_kind(content_type)

    directory = Path(source) / now.date().isoformat()
    target_dir = settings.ARTIFACTS_ROOT / directory
    target_dir.mkdir(parents=True, exist_ok=True)

    suffix = safe_suffix(uploaded.name, content_type)
    stored_path = directory / f"{artifact_id}{suffix}"
    with (settings.ARTIFACTS_ROOT / stored_path).open("wb") as target:
        for chunk in uploaded.chunks():
            target.write(chunk)

    return Artifact.objects.create(
        id=artifact_id,
        owner=user,
        source=source,
        title=title,
        kind=kind,
        theme="",
        day=now.date(),
        path=stored_path.as_posix(),
        original_kind=kind,
        tags=normalize_tags(data.get("tags", [])),
        source_prompt=data.get("source_prompt", ""),
        **source_fields(data, source),
    )


def wrap_html(body, title, theme, request):
    stylesheet = request.build_absolute_uri(f"/static/themes/{theme}.css")
    return f"""<!doctype html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>{html.escape(title)}</title>
  <link rel="stylesheet" href="{stylesheet}">
</head>
<body class="artifact">
{body}
</body>
</html>
"""


def wrap_deck(body, title, request):
    stylesheet = request.build_absolute_uri("/static/themes/deck.css")
    return f"""<!doctype html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>{html.escape(title)}</title>
  <link rel="stylesheet" href="{stylesheet}">
</head>
<body class="deck">
  <main class="slides">{body}</main>
</body>
</html>
"""


def generate_id():
    while True:
        artifact_id = secrets.token_hex(4)
        if not Artifact.objects.filter(pk=artifact_id).exists():
            return artifact_id


def resolve_token_user(request):
    auth = request.headers.get("Authorization", "")
    if not auth.startswith("Bearer "):
        return None
    key = auth[len("Bearer "):].strip()
    if not key:
        return None
    try:
        return Token.objects.select_related("user").get(key=key).user
    except Token.DoesNotExist:
        return None


def normalize_tags(tags):
    if isinstance(tags, str):
        return [tag.strip() for tag in tags.split(",") if tag.strip()]
    if isinstance(tags, list):
        return [str(tag).strip() for tag in tags if str(tag).strip()]
    return []


HEX_COLOR_RE = re.compile(r"^#[0-9a-fA-F]{6}$")

TAG_PALETTE = [
    "#4f6df5", "#e0578c", "#2fa88f", "#d98b2b",
    "#8c5fd6", "#3fa1d8", "#c14f4f", "#5aa552",
]


def tag_color(name, overrides):
    if override := overrides.get(name):
        return override
    index = int(hashlib.sha256(name.encode("utf-8")).hexdigest(), 16) % len(TAG_PALETTE)
    return TAG_PALETTE[index]


def tag_color_overrides():
    return dict(TagColor.objects.values_list("name", "color"))


def tags_with_colors(tags, overrides=None):
    overrides = overrides if overrides is not None else tag_color_overrides()
    return [{"name": tag, "color": tag_color(tag, overrides)} for tag in tags]


def labels_with_colors_list(labels, overrides=None):
    overrides = overrides if overrides is not None else tag_color_overrides()
    return [{"name": label, "color": tag_color(label, overrides)} for label in labels]


def labels_with_colors(artifact, overrides=None):
    return labels_with_colors_list([artifact.source, *artifact.tags], overrides)


def validate_source_fields(data):
    if source_type := data.get("source_type"):
        if source_type not in {choice[0] for choice in Artifact.SOURCE_TYPE_CHOICES}:
            return Response({"detail": "source_type must be agent, person, or system"}, status=400)
    try:
        parse_metadata_value(data.get("metadata", {}))
    except ValueError as exc:
        return Response({"detail": str(exc)}, status=400)
    try:
        parse_tag_colors_value(data.get("tag_colors", {}))
    except ValueError as exc:
        return Response({"detail": str(exc)}, status=400)
    return None


def parse_tag_colors_value(value):
    if value in (None, ""):
        return {}
    if isinstance(value, str):
        try:
            value = json.loads(value)
        except json.JSONDecodeError as exc:
            raise ValueError("tag_colors must be an object") from exc
    if not isinstance(value, dict):
        raise ValueError("tag_colors must be an object")
    for tag, color in value.items():
        if not HEX_COLOR_RE.match(str(color)):
            raise ValueError(f"tag_colors[{tag}] must be a hex color like #RRGGBB")
    return value


def apply_tag_colors(data):
    tag_colors = parse_tag_colors_value(data.get("tag_colors", {}))
    for tag, color in tag_colors.items():
        TagColor.objects.update_or_create(name=tag, defaults={"color": color})


def source_fields(data, source):
    return {
        "source_type": data.get("source_type") or Artifact.AGENT,
        "source_name": data.get("source_name") or source,
        "source_id": data.get("source_id") or "",
        "metadata": parse_metadata_value(data.get("metadata", {})),
    }


def parse_metadata_value(value):
    if value in (None, ""):
        return {}
    if isinstance(value, dict):
        return value
    if isinstance(value, str):
        try:
            parsed = json.loads(value)
        except json.JSONDecodeError as exc:
            raise ValueError("metadata must be an object") from exc
        if isinstance(parsed, dict):
            return parsed
    raise ValueError("metadata must be an object")


def binary_kind(content_type):
    if content_type.startswith("image/"):
        return Artifact.IMAGE
    if content_type == "application/pdf":
        return Artifact.PDF
    return Artifact.BINARY


def safe_suffix(filename, content_type):
    suffix = Path(filename).suffix.lower()
    if suffix and suffix[1:].isalnum() and len(suffix) <= 10:
        return suffix
    return mimetypes.guess_extension(content_type) or ".bin"


def serialize_artifact(artifact, request):
    return {
        "id": artifact.id,
        "url": request.build_absolute_uri(reverse("artifact", args=[artifact.id])),
        "api_url": request.build_absolute_uri(reverse("artifact-detail-api", args=[artifact.id])),
        "raw_url": request.build_absolute_uri(f"{reverse('artifact', args=[artifact.id])}?raw=1"),
        "raw_available": raw_artifact_path(artifact) is not None,
        "source": artifact.source,
        "title": artifact.title,
        "kind": artifact.kind,
        "theme": artifact.theme,
        "day": artifact.day.isoformat(),
        "created_at": artifact.created_at.isoformat(),
        "tags": tags_with_colors(artifact.tags),
        "labels": labels_with_colors(artifact),
        "source_prompt": artifact.source_prompt,
        "source_type": artifact.source_type,
        "source_name": artifact.source_name,
        "source_id": artifact.source_id,
        "metadata": artifact.metadata,
        "read_at": artifact.read_at.isoformat() if artifact.read_at else None,
    }


def safe_artifact_path(relative_path):
    root = settings.ARTIFACTS_ROOT.resolve()
    path = (root / relative_path).resolve()
    if root not in path.parents and path != root:
        raise Http404("Invalid artifact path")
    return path


def raw_artifact_path(artifact):
    path = safe_artifact_path(artifact.path)
    if artifact.kind == Artifact.MARKDOWN:
        return path.with_suffix(".md")
    if artifact.kind == Artifact.HTML_FULL:
        return path
    return None


class DaybookLoginView(auth_views.LoginView):
    template_name = "artifacts/login.html"
    authentication_form = DaybookLoginForm

    def form_valid(self, form):
        response = super().form_valid(form)
        if form.cleaned_data.get("remember_me"):
            self.request.session.set_expiry(REMEMBER_ME_SESSION_AGE)
        else:
            self.request.session.set_expiry(0)
        return response


@login_required
def settings_home(request):
    return redirect("settings-account")


@login_required
def settings_account(request):
    token, _ = Token.objects.get_or_create(user=request.user)
    if request.method == "POST" and request.POST.get("action") == "regenerate":
        token.delete()
        token = Token.objects.create(user=request.user)
        messages.success(request, "API token regenerated.")
        return redirect("settings-account")
    return render(
        request,
        "artifacts/settings_account.html",
        {"token": token, "active_section": "account"},
    )


@login_required
def settings_demo_data(request):
    if request.method == "POST":
        count = create_demo_artifacts(request.user, request)
        messages.success(request, f"Created {count} demo artifacts.")
        return redirect("settings-demo-data")
    return render(
        request,
        "artifacts/settings_demo_data.html",
        {"active_section": "demo-data"},
    )


def raw_unavailable_html(artifact):
    return f"""<!doctype html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1">
  <title>Raw source unavailable</title>
  <style>
    body {{ background: #f4f1ea; color: #1d2528; font-family: system-ui, sans-serif; margin: 0; padding: 2rem; }}
    main {{ background: #fffaf0; border-radius: 16px; max-width: 42rem; padding: 1.4rem; }}
  </style>
</head>
<body>
  <main>
    <h1>Raw source unavailable</h1>
    <p>Daybook does not have preserved raw source for <strong>{html.escape(artifact.title)}</strong>.</p>
  </main>
</body>
</html>
"""