app.py

import io, json
from fastapi import FastAPI, Request, Query
from fastapi.responses import JSONResponse, HTMLResponse, Response
from PIL import Image

app = FastAPI()
IMG = Image.open("page.jpg").convert("RGB")
W, H = IMG.size
LINES = [l for l in json.load(open("lines.json")) if l.get("text")]
CORS = {"Access-Control-Allow-Origin": "*"}

def base(req): return str(req.base_url).rstrip("/")
def canvas(b): return f"{b}/canvas/p1"

# ---- minimal dynamic IIIF Image API 2.1 (level1) ----
def parse_region(r):
    if r in ("full", "square"): return (0, 0, W, H)
    if r.startswith("pct:"):
        x, y, w, h = [float(v) for v in r[4:].split(",")]
        return (int(x/100*W), int(y/100*H), int(w/100*W), int(h/100*H))
    x, y, w, h = [int(float(v)) for v in r.split(",")]
    return (x, y, w, h)

def parse_size(s, rw, rh):
    if s in ("full", "max", "^max"): return (rw, rh)
    if s.startswith("pct:"):
        p = float(s[4:]) / 100; return (max(1, int(rw*p)), max(1, int(rh*p)))
    fit = s.startswith("!");  s = s[1:] if fit else s
    w, h = (s.split(",") + [""])[:2]
    if w and not h: w = int(w); return (w, max(1, round(rh*w/rw)))
    if h and not w: h = int(h); return (max(1, round(rw*h/rh)), h)
    w, h = int(w), int(h)
    if fit:
        sc = min(w/rw, h/rh); return (max(1, int(rw*sc)), max(1, int(rh*sc)))
    return (w, h)

@app.get("/iiif/page/info.json")
def info(req: Request):
    b = base(req)
    return JSONResponse({"@context": "http://iiif.io/api/image/2/context.json",
        "@id": f"{b}/iiif/page", "@type": "iiif:Image",
        "protocol": "http://iiif.io/api/image", "width": W, "height": H,
        "profile": ["http://iiif.io/api/image/2/level1.json"],
        "tiles": [{"width": 1024, "scaleFactors": [1, 2, 4, 8, 16, 32]}]}, headers=CORS)

@app.get("/iiif/page/{region}/{size}/{rotation}/{quality}.{fmt}")
def tile(region: str, size: str, rotation: str, quality: str, fmt: str):
    x, y, w, h = parse_region(region)
    crop = IMG.crop((x, y, x + w, y + h))
    tw, th = parse_size(size, w, h)
    crop = crop.resize((max(1, tw), max(1, th)))
    if quality == "gray": crop = crop.convert("L").convert("RGB")
    buf = io.BytesIO(); crop.save(buf, "JPEG", quality=85)
    return Response(buf.getvalue(), media_type="image/jpeg", headers=CORS)

# ---- Presentation 3.0 manifest (image service + OCR annotations + search service) ----
@app.get("/manifest.json")
def manifest(req: Request):
    b = base(req); c = canvas(b)
    anns = []
    for i, ln in enumerate(LINES):
        x0, y0, x1, y1 = [int(round(v)) for v in ln["bbox"]]
        anns.append({"id": f"{b}/anno/l{i}", "type": "Annotation", "motivation": "supplementing",
            "body": {"type": "TextualBody", "format": "text/plain", "language": "en", "value": ln["text"]},
            "target": f"{c}#xywh={x0},{y0},{max(1,x1-x0)},{max(1,y1-y0)}"})
    m = {"@context": "http://iiif.io/api/presentation/3/context.json",
        "id": f"{b}/manifest.json", "type": "Manifest",
        "label": {"en": ["The Commoner, 1901-01-23, p.1 — Surya OCR 2 (line-level)"]},
        "service": [{"id": f"{b}/search", "type": "SearchService1",
                     "profile": "http://iiif.io/api/search/1/search"}],
        "items": [{"id": c, "type": "Canvas", "width": W, "height": H,
            "items": [{"id": f"{c}/page", "type": "AnnotationPage", "items": [
                {"id": f"{c}/painting", "type": "Annotation", "motivation": "painting",
                 "body": {"id": f"{b}/iiif/page/full/max/0/default.jpg", "type": "Image",
                          "format": "image/jpeg", "width": W, "height": H,
                          "service": [{"id": f"{b}/iiif/page", "type": "ImageService2", "profile": "level1"}]},
                 "target": c}]}],
            "annotations": [{"id": f"{c}/ocr", "type": "AnnotationPage", "items": anns}]}]}
    return JSONResponse(m, headers=CORS)

# ---- IIIF Content Search 1.0 (Mirador's search box queries this) ----
@app.get("/search")
def search(req: Request, q: str = Query("")):
    b = base(req); c = canvas(b); ql = q.lower().strip()
    res = []
    for i, ln in enumerate(LINES):
        if ql and ql in ln["text"].lower():
            x0, y0, x1, y1 = [int(round(v)) for v in ln["bbox"]]
            res.append({"@id": f"{b}/anno/s{i}", "@type": "oa:Annotation", "motivation": "sc:painting",
                "resource": {"@type": "cnt:ContentAsText", "chars": ln["text"]},
                "on": f"{c}#xywh={x0},{y0},{max(1,x1-x0)},{max(1,y1-y0)}"})
    return JSONResponse({"@context": "http://iiif.io/api/presentation/2/context.json",
        "@id": f"{b}/search?q={q}", "@type": "sc:AnnotationList",
        "within": {"@type": "sc:Layer", "total": len(res)}, "resources": res}, headers=CORS)

@app.get("/", response_class=HTMLResponse)
def index():
    return """<!DOCTYPE html><html><head><meta charset="utf-8">
<title>Surya OCR 2 — IIIF search (Mirador)</title>
<style>html,body,#m{height:100%;width:100%;margin:0}</style></head>
<body><div id="m"></div>
<script src="https://unpkg.com/mirador@3.3.0/dist/mirador.min.js"></script>
<script>
  Mirador.viewer({ id:'m',
    window:{ allowFullscreen:true, sideBarOpenByDefault:true, defaultSideBarPanel:'search' },
    windows:[{ manifestId: location.origin + '/manifest.json' }] });
</script></body></html>"""