Develop

.env.example

@@ -1,2 +1,3 @@
-APP_NAME=""
-APP_VERSION=""
+APP_NAME="OCR PDF"
+APP_VERSION="0.0.1"
+APP_ORIGINS=http://127.0.0.1:8001

.github/workflows/docker-image.yml

@@ -28,7 +28,9 @@ jobs:
 
       - name: Smoke test container
         run: |
-          docker run --rm -d --name test_app -p 8000:8000 myapp:test
+          docker run -d --name test_app -p 8000:8000 myapp:test
           sleep 10
-          docker ps | grep test_app
+          docker ps -a
           docker logs test_app
+          curl --fail http://localhost:8000/docs
+          docker rm -f test_app

Dockerfile

@@ -1,13 +1,11 @@
-# ========== Base ==========
 FROM python:3.11-slim AS base
 
-# Evita prompts e melhora logs
+WORKDIR /app
+
 ENV DEBIAN_FRONTEND=noninteractive \
     PYTHONDONTWRITEBYTECODE=1 \
     PYTHONUNBUFFERED=1
 
-# Dependências do sistema para o ocrmypdf e Tesseract
-# (inclui ghostscript, qpdf e idiomas pt/en)
 RUN apt-get update && \
     apt-get install -y --no-install-recommends \
         ocrmypdf \
@@ -17,30 +15,24 @@ RUN apt-get update && \
         ghostscript \
         qpdf \
         # utilitários úteis
-        curl ca-certificates && \
-    rm -rf /var/lib/apt/lists/*
+        curl ca-certificates \
+        redis-tools \
+        && rm -rf /var/lib/apt/lists/*
 
-# ========== App ==========
-WORKDIR /app
 
-# Copia requirements primeiro para aproveitar cache de camadas
 COPY requirements.txt /app/requirements.txt
 RUN pip install --no-cache-dir --upgrade pip && \
     pip install --no-cache-dir -r /app/requirements.txt
 
-# Copia o código da API (ajuste se tiver mais arquivos)
 COPY app.py /app/app.py
+COPY worker.py /app/worker.py
+COPY tasks.py /app/tasks.py
 COPY .env /app/.env
 
-# Exponha a porta do Uvicorn
 EXPOSE 8000
 
-# Opcional: usuário sem privilégios
-RUN useradd -m appuser
-USER appuser
-
-# Healthcheck simples (tente GET /health se você adicionou esse endpoint)
-# HEALTHCHECK --interval=30s --timeout=3s --retries=3 CMD curl -fsS http://localhost:8000/health || exit 1
+# usuário sem privilégios
+# RUN useradd -m appuser
+# USER appuser
 
-# Comando padrão (ajuste host/port conforme seu ambiente)
 CMD ["python", "-m", "uvicorn", "app:app", "--host", "0.0.0.0", "--port", "8000"]

app.py

@@ -1,119 +1,36 @@
+import uuid
 import shutil
-import subprocess
-import tempfile
-import time
 from pathlib import Path
-from typing import Literal
-
 from fastapi import FastAPI, UploadFile, File, Form, HTTPException
-from fastapi.responses import JSONResponse, StreamingResponse
 from fastapi.middleware.cors import CORSMiddleware
-from pypdf import PdfReader
 from dotenv import load_dotenv
 import os
+from redis import Redis
+from rq import Queue
+from fastapi.responses import FileResponse
+from tasks import process_ocr_job
+
+load_dotenv()
 
-load_dotenv()  # Lê as variáveis do .env
+origins_env = os.getenv("APP_ORIGINS", "")
+allowed_origins = [o.strip() for o in origins_env.split(",") if o.strip()]
 
+app = FastAPI(title=os.getenv("APP_NAME"), version=os.getenv("APP_VERSION"))
 
-app = FastAPI(title="OCR API", version="1.2.0")
 app.add_middleware(
     CORSMiddleware,
-    allow_origins=["http://127.0.0.1:8001",],  # ajuste
+    allow_origins=[allowed_origins],
     allow_credentials=True,
     allow_methods=["*"],
     allow_headers=["*"],
 )
 
+# Conexão Redis e fila
+redis_conn = Redis.from_url(os.getenv("REDIS_URL", "redis://localhost:6379"))
+q = Queue("ocr", connection=redis_conn)
 
-def extract_text_with_pypdf(pdf_path: Path) -> str:
-    reader = PdfReader(str(pdf_path))
-    texts = []
-    for page in reader.pages:
-        try:
-            t = page.extract_text() or ""
-        except Exception:
-            t = ""
-        texts.append(t.strip())
-    return "\n\n".join(texts)
-
-
-def run_ocrmypdf(src: Path, dst: Path, lang: str = "por+eng") -> None:
-    cmd = [
-        "ocrmypdf",
-        "--language", lang,
-        "--deskew",
-        "--optimize", "1",
-        "--force-ocr",
-        str(src),
-        str(dst),
-    ]
-    completed = subprocess.run(cmd, capture_output=True, text=True)
-    if completed.returncode != 0:
-        raise RuntimeError(
-            f"OCR falhou (code {completed.returncode}). "
-            f"stderr: {completed.stderr.strip() or 'sem stderr'}"
-        )
-
-
-@app.post("/ocr")
-async def ocr_endpoint(
-    file: UploadFile = File(..., description="PDF para OCR"),
-    lang: str = Form("por+eng"),
-    response_type: Literal["json", "pdf"] = Form("pdf"),
-):
-    ct = (file.content_type or "").lower()
-    if ct not in {"application/pdf", "application/x-pdf", "application/octet-stream"}:
-        raise HTTPException(status_code=400, detail=f"Tipo de arquivo não suportado: {ct or 'desconhecido'}")
-
-    try:
-        start_time = time.perf_counter()
-
-        with tempfile.TemporaryDirectory() as tmpdir:
-            tmpdir = Path(tmpdir)
-
-            # Nome do arquivo de saída
-            original_name = file.filename or "arquivo.pdf"
-            original_stem = Path(original_name).stem or "arquivo"
-            download_name = f"{original_stem}_ocr.pdf"
-
-            in_path = tmpdir / "entrada.pdf"
-            out_pdf = tmpdir / download_name
-
-            # Salva upload no disco
-            with in_path.open("wb") as f:
-                shutil.copyfileobj(file.file, f)
-
-            # Executa OCR
-            run_ocrmypdf(in_path, out_pdf, lang=lang)
-
-            elapsed = round(time.perf_counter() - start_time, 2)  # segundos
-
-            if response_type == "json":
-                text_ocr = extract_text_with_pypdf(out_pdf)
-                pages = len(PdfReader(str(out_pdf)).pages)
-                return JSONResponse(
-                    {
-                        "filename_in": original_name,
-                        "filename_out": download_name,
-                        "lang": lang,
-                        "pages": pages,
-                        "elapsed_seconds": elapsed,
-                        "text": text_ocr,
-                    }
-                )
-
-            # PDF como resposta
-            headers = {
-                "Content-Disposition": f'attachment; filename="{download_name}"',
-                "X-Original-Filename": original_name,
-                "X-Processing-Time": f"{elapsed}s",
-            }
-            return StreamingResponse(out_pdf.open("rb"), media_type="application/pdf", headers=headers)
-
-    except RuntimeError as e:
-        raise HTTPException(status_code=500, detail=str(e))
-    except Exception as e:
-        raise HTTPException(status_code=500, detail=f"Erro inesperado: {e}")
+UPLOAD_DIR = Path(os.getenv("UPLOAD_DIR", "./uploads"))
+UPLOAD_DIR.mkdir(parents=True, exist_ok=True)
 
 
 @app.get("/")
@@ -122,3 +39,56 @@ async def index():
         "status": "success",
         "message": f"{os.getenv('APP_NAME')} - v{os.getenv('APP_VERSION')}"
     }
+
+
+@app.post("/ocr")
+async def enqueue_ocr(
+    file: UploadFile = File(...),
+    lang: str = Form("por+eng"),
+    response_type: str = Form("pdf")
+):
+    # Salva o upload em disco para o worker ler
+    tmp_path = UPLOAD_DIR / f"{uuid.uuid4()}_{file.filename}"
+    with tmp_path.open("wb") as f:
+        shutil.copyfileobj(file.file, f)
+
+    job = q.enqueue(
+        process_ocr_job,
+        str(tmp_path),
+        lang,
+        response_type,
+        job_timeout=3600
+    )
+    return {"task_id": job.get_id(), "status": "queued"}
+
+
+@app.get("/status/{task_id}")
+async def job_status(task_id: str):
+    from rq.job import Job
+    try:
+        job = Job.fetch(task_id, connection=redis_conn)
+    except Exception:
+        raise HTTPException(status_code=404, detail="Job não encontrado")
+
+    if job.is_queued:
+        return {"status": "queued"}
+    if job.is_started:
+        return {"status": "processing"}
+    if job.is_finished:
+        result = job.result
+        download_url = f"/download/{result['filename_out']}"
+        return {"status": "done", "result": result, "download_url": download_url}
+    if job.is_failed:
+        return {"status": "failed", "error": str(job.exc_info)}
+
+
+@app.get("/download/{filename}")
+async def download_file(filename: str):
+    file_path = UPLOAD_DIR / filename
+    if not file_path.exists():
+        raise HTTPException(status_code=404, detail="Arquivo não encontrado")
+    return FileResponse(
+        file_path,
+        media_type="application/pdf",
+        filename=filename
+    )

requirements.txt

@@ -3,3 +3,5 @@ python-dotenv
 uvicorn
 fastapi
 python-multipart
+rq
+redis

tasks.py

@@ -0,0 +1,74 @@
+import os
+import time
+import subprocess
+import tempfile
+from pathlib import Path
+from pypdf import PdfReader
+
+
+def extract_text_with_pypdf(pdf_path: Path) -> str:
+    reader = PdfReader(str(pdf_path))
+    texts = []
+    for page in reader.pages:
+        try:
+            t = page.extract_text() or ""
+        except Exception:
+            t = ""
+        texts.append(t.strip())
+    return "\n\n".join(texts)
+
+
+def run_ocrmypdf(src: Path, dst: Path, lang: str = "por+eng") -> None:
+    cmd = [
+        "ocrmypdf",
+        "--language", lang,
+        "--deskew",
+        "--optimize", "1",
+        "--force-ocr",
+        str(src),
+        str(dst),
+    ]
+    completed = subprocess.run(cmd, capture_output=True, text=True)
+    if completed.returncode != 0:
+        raise RuntimeError(
+            f"OCR falhou (code {completed.returncode}). "
+            f"stderr: {completed.stderr.strip() or 'sem stderr'}"
+        )
+
+
+def process_ocr_job(file_path: str, lang: str, response_type: str):
+    """
+    Função executada pelo worker RQ.
+    Recebe o caminho do arquivo salvo pelo endpoint.
+    """
+    start = time.perf_counter()
+
+    shared_dir = Path(os.getenv("UPLOAD_DIR", "/shared_uploads"))
+    shared_dir.mkdir(parents=True, exist_ok=True)
+
+    original_path = Path(file_path)
+    download_name = f"{original_path.stem}_ocr.pdf"
+    out_pdf = shared_dir / download_name   # <- grava no volume compartilhado
+
+    run_ocrmypdf(original_path, out_pdf, lang=lang)
+
+    elapsed = round(time.perf_counter() - start, 2)
+
+    if response_type == "json":
+        text_ocr = extract_text_with_pypdf(out_pdf)
+        pages = len(PdfReader(str(out_pdf)).pages)
+        return {
+            "status": "done",
+            "filename_in": original_path.name,
+            "filename_out": download_name,      # apenas o nome do arquivo
+            "elapsed_seconds": elapsed,
+            "pages": pages,
+            "text": text_ocr,
+        }
+    else:
+        return {
+            "status": "done",
+            "filename_in": original_path.name,
+            "filename_out": download_name,
+            "elapsed_seconds": elapsed,
+        }

worker.py

@@ -0,0 +1,13 @@
+import os
+from redis import Redis
+from rq import Worker, Queue
+
+listen = ["ocr"]
+redis_url = os.getenv("REDIS_URL", "redis://localhost:6379/0")
+conn = Redis.from_url(redis_url)
+
+if __name__ == "__main__":
+    # Cria as filas e inicia o worker usando a conexão direta
+    queues = [Queue(name, connection=conn) for name in listen]
+    worker = Worker(queues, connection=conn)
+    worker.work()