added uvicorn web api

2026-09-02 20:20:01 +02:00
parent cabab25861
commit 7065dfd8d2
8 changed files with 250 additions and 0 deletions
@@ -0,0 +1,55 @@
+# Fundi Scraper API
+
+A small REST API (FastAPI) that serves the events scraped by
+[`scraper/scraper.py`](../scraper/scraper.py) from a JSON file.
+
+## Setup
+
+```bash
+cd api
+pip install -r requirements.txt
+```
+
+## Run
+
+```bash
+cd api
+uvicorn main:app --reload
+```
+
+Interactive docs: http://127.0.0.1:8000/docs
+
+By default the API reads `../fundi-scraped-output.json` (the repo root).
+Point it elsewhere with an env var:
+
+```bash
+DATA_FILE=/path/to/events.json uvicorn main:app
+```
+
+## Endpoints
+
+| Method | Path             | Description |
+|--------|------------------|-------------|
+| GET    | `/health`        | Status + number of events loaded |
+| GET    | `/events`        | List events, with optional filters (see below) |
+| GET    | `/events/{index}`| Single event by its position in the file (0-based) |
+| GET    | `/artists`       | Deduplicated, sorted list of all artist names |
+| POST   | `/reload`        | Re-read the JSON file from disk (after a fresh scrape) |
+
+### `/events` query parameters
+
+| Param      | Type   | Meaning |
+|------------|--------|---------|
+| `free`     | bool   | Only free / only paid events |
+| `name`     | string | Case-insensitive substring match on the event name |
+| `artist`   | string | Case-insensitive substring match on any artist name |
+| `date`     | string | Exact match on the raw date string (`dd.mm.yy`) |
+| `upcoming` | bool   | `true` = today or later, `false` = past events |
+
+Examples:
+
+```bash
+curl 'http://127.0.0.1:8000/events?free=true'
+curl 'http://127.0.0.1:8000/events?artist=randali&upcoming=true'
+curl 'http://127.0.0.1:8000/events/0'
+```
@@ -0,0 +1,68 @@
+"""Loads the scraped events JSON file and keeps it in memory.
+
+The file is treated as read-only. Call ``reload()`` to pick up a fresh
+scrape without restarting the server.
+"""
+
+import json
+import os
+from datetime import date
+from pathlib import Path
+from threading import Lock
+from typing import List, Optional
+
+from models import Event
+
+# Default: the fundi-scraped-output.json sitting in the repo root, one
+# directory above this file. Override with the DATA_FILE env var.
+_DEFAULT_DATA_FILE = Path(__file__).resolve().parent.parent / "fundi-scraped-output.json"
+
+
+class EventStore:
+    def __init__(self, path: Optional[os.PathLike] = None):
+        self.path = Path(path or os.environ.get("DATA_FILE") or _DEFAULT_DATA_FILE)
+        self._lock = Lock()
+        self._events: List[Event] = []
+        self.reload()
+
+    def reload(self) -> int:
+        """Re-read the JSON file from disk. Returns the number of events loaded."""
+        if not self.path.exists():
+            raise FileNotFoundError(f"Data file not found: {self.path}")
+        with self.path.open(encoding="utf-8") as fh:
+            raw = json.load(fh)
+        events = [Event.model_validate(item) for item in raw]
+        with self._lock:
+            self._events = events
+        return len(events)
+
+    def all(self) -> List[Event]:
+        with self._lock:
+            return list(self._events)
+
+    def get(self, index: int) -> Optional[Event]:
+        with self._lock:
+            if 0 <= index < len(self._events):
+                return self._events[index]
+        return None
+
+
+def parse_event_date(value: str) -> Optional[date]:
+    """Parse the scraper's ``dd.mm.yy`` date strings. Returns None if unparseable."""
+    try:
+        day, month, year = (int(part) for part in value.strip().split("."))
+    except (ValueError, AttributeError):
+        return None
+    if year < 100:
+        year += 2000
+    try:
+        return date(year, month, day)
+    except ValueError:
+        return None
+
+
+def iter_artist_names(event: Event):
+    for group in event.event_artists:
+        for artist in group:
+            if artist.artist_name:
+                yield artist.artist_name
@@ -0,0 +1,102 @@
+"""REST API over the scraped Fundbureau events.
+
+Datasource is the JSON file produced by ``scraper/scraper.py``
+(``fundi-scraped-output.json`` by default; override with the DATA_FILE env var).
+
+Run with:
+    uvicorn main:app --reload    # from inside the api/ folder
+"""
+
+import datetime
+from typing import List, Optional
+
+from fastapi import FastAPI, HTTPException, Query
+
+from datasource import EventStore, iter_artist_names, parse_event_date
+from models import Event, EventList
+
+app = FastAPI(
+    title="Fundi Scraper API",
+    description="REST API over the scraped upcoming/past events at the Fundbureau Hamburg.",
+    version="1.0.0",
+)
+
+store = EventStore()
+
+
+def _matches(
+    event: Event,
+    free: Optional[bool],
+    name: Optional[str],
+    artist: Optional[str],
+    event_date: Optional[str],
+    upcoming: Optional[bool],
+    today: datetime.date,
+) -> bool:
+    if free is not None and event.event_free is not free:
+        return False
+    if name is not None and name.lower() not in event.event_name.lower():
+        return False
+    if artist is not None:
+        needle = artist.lower()
+        if not any(needle in n.lower() for n in iter_artist_names(event)):
+            return False
+    if event_date is not None and event.event_date != event_date:
+        return False
+    if upcoming is not None:
+        parsed = parse_event_date(event.event_date)
+        if parsed is None:
+            return False
+        is_upcoming = parsed >= today
+        if is_upcoming is not upcoming:
+            return False
+    return True
+
+
+@app.get("/health")
+def health():
+    return {"status": "ok", "events_loaded": len(store.all()), "data_file": str(store.path)}
+
+
+@app.post("/reload")
+def reload():
+    """Re-read the JSON file from disk (e.g. after a fresh scrape)."""
+    try:
+        count = store.reload()
+    except FileNotFoundError as exc:
+        raise HTTPException(status_code=500, detail=str(exc))
+    return {"reloaded": True, "events_loaded": count}
+
+
+@app.get("/events", response_model=EventList)
+def list_events(
+    free: Optional[bool] = Query(None, description="Filter by free admission"),
+    name: Optional[str] = Query(None, description="Case-insensitive substring match on event name"),
+    artist: Optional[str] = Query(None, description="Case-insensitive substring match on any artist name"),
+    date: Optional[str] = Query(None, description="Exact match on the raw event date (dd.mm.yy)"),
+    upcoming: Optional[bool] = Query(None, description="true = today or later, false = past events"),
+):
+    today = datetime.date.today()
+    events = [
+        e
+        for e in store.all()
+        if _matches(e, free, name, artist, date, upcoming, today)
+    ]
+    return EventList(count=len(events), events=events)
+
+
+@app.get("/artists", response_model=List[str])
+def list_artists():
+    """Deduplicated, sorted list of all artist names across all events."""
+    names = set()
+    for event in store.all():
+        names.update(iter_artist_names(event))
+    return sorted(names, key=str.lower)
+
+
+@app.get("/events/{index}", response_model=Event)
+def get_event(index: int):
+    event = store.get(index)
+    if event is None:
+        raise HTTPException(status_code=404, detail=f"No event at index {index}")
+    return event
@@ -0,0 +1,23 @@
+from typing import List, Optional
+
+from pydantic import BaseModel
+
+
+class Artist(BaseModel):
+    artist_name: str
+    artist_play_time_start: str
+    artist_play_time_end: str
+
+
+class Event(BaseModel):
+    event_date: str
+    event_name: str
+    event_artists: List[List[Artist]]
+    event_starttime: str
+    event_ticket_link: str
+    event_free: Optional[bool] = None
+
+
+class EventList(BaseModel):
+    count: int
+    events: List[Event]
@@ -0,0 +1,2 @@
+fastapi>=0.103
+uvicorn[standard]>=0.23