SAFEHR-data · p-j-smith · Oct 8, 2025 · Oct 7, 2025 · Oct 7, 2025 · Oct 7, 2025
diff --git a/docker-compose-postgres.yml b/docker-compose-postgres.yml
diff --git a/docker-compose.yml b/docker-compose.yml
@@ -185,7 +185,7 @@ services:
 
   ocr-forwarding-api:
     build:
-      context: src/api
+      context: packages/api
       dockerfile: Dockerfile
       args:
         <<: *build-args-common
@@ -199,7 +199,6 @@ services:
     ports:
       - "${OCR_FORWARDING_API_PORT}:${OCR_FORWARDING_API_PORT}"
     volumes:
-      - ./src/api/app:/app
       - ${PWD}/${DATA_FOLDER}:/data
     networks:
       - pyonb_ocr_api

diff --git a/packages/api/Dockerfile b/packages/api/Dockerfile
@@ -0,0 +1,19 @@
+FROM ghcr.io/astral-sh/uv:python3.13-bookworm AS app
+
+SHELL ["/bin/bash", "-o", "pipefail", "-e", "-u", "-x", "-c"]
+
+WORKDIR /app
+ENV PYTHONDONTWRITEBYTECODE=1
+ENV PYTHONUNBUFFERED=1
+
+COPY ./pyproject.toml .
+COPY ./README.md .
+COPY ./src src/
+
+RUN uv venv
+RUN --mount=type=cache,target=/root/.cache/uv,sharing=locked uv sync --no-editable --no-dev
+
+# make uvicorn etc available
+ENV PATH="/app/.venv/bin:$PATH"
+
+CMD uvicorn pyonb_api.main:app --host 0.0.0.0 --port "$OCR_FORWARDING_API_PORT" --workers 4 --use-colors
diff --git a/packages/api/README.md b/packages/api/README.md
@@ -0,0 +1,25 @@
+# `pyonb` forwarding API
+
+The forwarding API is used to send documents to an OCR service for processing. It provides a consistent
+interface for using the various OCR tools supported by `pyonb`.
+
+## Usage
+
+You will need to define the `OCR_FORWARDING_API_PORT` in a `.env` file.
+
+Then, spin up the `ocr-forwarding-api` servicer, along with the OCR service you wish to use.
+For example, if you would like to use `kreuzberg`, run the following from the top-level `pyonb` directory:
+
+```shell
+docker-compose --profile kreuzberg up --build --detach
+```
+
+You can then use `curl` to send a PDF to the forwarding API:
+
+```shell
+curl -v -X POST http://127.0.0.1:8110/kreuzberg-ocr/inference_single \
+  -F "file_upload=@document.pdf" \
+  -H "accept: application/json"
+```
+
+Note, this assumes you have set `OCR_FORWARDING_API_PORT` to `8110`.
diff --git a/packages/api/pyproject.toml b/packages/api/pyproject.toml
@@ -0,0 +1,16 @@
+[build-system]
+build-backend = "hatchling.build"
+requires = ["hatchling"]
+
+[project]
+dependencies = [
+    "aiohttp",
+    "fastapi[standard]",
+    "requests",
+    "uvicorn",
+]
+description = "pyonb forwarding API"
+name = "pyonb-api"
+readme = "README.md"
+requires-python = ">=3.11"
+version = "0.1.0"
diff --git a/src/api/app/__init__.py → packages/api/src/pyonb_api/__init__.py b/src/api/app/__init__.py → packages/api/src/pyonb_api/__init__.py
diff --git a/src/api/app/main.py → packages/api/src/pyonb_api/main.py b/src/api/app/main.py → packages/api/src/pyonb_api/main.py
diff --git a/src/api/app/routers/__init__.py → ...ges/api/src/pyonb_api/routers/__init__.py b/src/api/app/routers/__init__.py → ...ges/api/src/pyonb_api/routers/__init__.py
diff --git a/src/api/app/routers/docling.py → ...ages/api/src/pyonb_api/routers/docling.py b/src/api/app/routers/docling.py → ...ages/api/src/pyonb_api/routers/docling.py
diff --git a/src/api/app/routers/kreuzberg.py → ...es/api/src/pyonb_api/routers/kreuzberg.py b/src/api/app/routers/kreuzberg.py → ...es/api/src/pyonb_api/routers/kreuzberg.py
diff --git a/src/api/app/routers/marker.py → packages/api/src/pyonb_api/routers/marker.py b/src/api/app/routers/marker.py → packages/api/src/pyonb_api/routers/marker.py
diff --git a/src/api/app/routers/paddleocr.py → ...es/api/src/pyonb_api/routers/paddleocr.py b/src/api/app/routers/paddleocr.py → ...es/api/src/pyonb_api/routers/paddleocr.py
diff --git a/pyproject.toml b/pyproject.toml
@@ -34,7 +34,9 @@ dynamic = [
 keywords = [
 ]
 name = "pyonb"
-optional-dependencies = {dev = [
+optional-dependencies = {api = [
+    "pyonb-api",
+], dev = [
     "build",
     "marker-pdf[full]==1.6.2",
     "mypy",
@@ -159,12 +161,14 @@ gh.python."3.12" = ["py312"]
 gh.python."3.13" = ["py313"]
 
 [tool.uv.sources]
+pyonb-api = {workspace = true}
 pyonb-docling = {workspace = true}
 pyonb-kreuzberg = {workspace = true}
 pyonb-marker = {workspace = true}
 pyonb-paddleocr = {workspace = true}
 
 [tool.uv.workspace]
 members = [
+    "packages/api",
     "packages/ocr/*",
 ]
diff --git a/src/api/Dockerfile b/src/api/Dockerfile
diff --git a/src/api/app/util.py b/src/api/app/util.py
diff --git a/src/api/requirements.txt b/src/api/requirements.txt
diff --git a/tests/api/test_app.py b/tests/api/test_app.py
@@ -1,8 +1,5 @@
 """Test functions in /src/api/app/main.py."""
 
-import os
-from pathlib import Path
-
 import pytest
 import requests
 
@@ -29,26 +26,3 @@ def test_local_start_api_and_healthy(ocr_forwarding_api_port: str) -> None:
 def test_check_services(check_container_healthy: bool) -> None:
     """Test OCR API Docker services are up."""
     assert check_container_healthy, "Docker services not running. Required for full API testing."
-
-
-def test_data_folder_contents() -> None:
-    """
-    Test documents data folder correct for testing.
-
-    For testing, env variable HOST_DATA_FOLDER should mount /tests/data/single_synthetic_doc
-    to CONTAINER_DATA_FOLDER, which should be /data inside the container.
-    There should be a single test PDF file (ms-note-one-page.pdf) inside.
-    """
-    assert "single_synthetic_doc" in str(os.getenv("HOST_DATA_FOLDER")), (
-        "For testing, HOST_DATA_FOLDER should point at /tests/data/single_synthetic_doc. "
-        "Check environment variables if this fails. "
-        f"HOST_DATA_FOLDER: {os.getenv('HOST_DATA_FOLDER')}"
-    )
-    assert "data" in str(os.getenv("CONTAINER_DATA_FOLDER")), (
-        "CONTAINER_DATA_FOLDER should point at /data."
-        "Check environment variables if this fails. "
-        f"CONTAINER_DATA_FOLDER: {os.getenv('CONTAINER_DATA_FOLDER')}"
-    )
-
-    test_pdf_file_path = Path(str(os.getenv("HOST_DATA_FOLDER"))) / "ms-note-one-page.pdf"
-    assert test_pdf_file_path.is_file(), f"File at {test_pdf_file_path} not found."