Merge pull request #6 from baloise/main

Refactor api routes

.dockerignore

@@ -0,0 +1,2 @@
+poetry.lock
+.install.stamp

.gitignore

@@ -160,3 +160,6 @@ cython_debug/
 #  and can be added to the global gitignore or merged into this file.  For a more nuclear
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
+
+/.install.stamp
+/poetry.lock

.vscode/extensions.json

@@ -1,4 +1,9 @@
 {
-    "recommendations": ["matangover.mypy"]
+    "recommendations": [
+      "matangover.mypy",
+      "pamaron.pytest-runner",
+      "tamasfe.even-better-toml",
+      "ms-vscode.makefile-tools"
+    ]
   }
 

Dockerfile

@@ -1,15 +1,41 @@
-FROM python:3.13-bookworm
+FROM alpine:latest
 
+# set metadata
 LABEL maintainer="culmat, trichie, robbizbal" \
       description="Yo-Yo-Maskr application Docker image" \
       version="0.1.0"
 
-WORKDIR /app
+# set default environment variables
+ENV OLLAMA_BASE_URL=http://localhost:11434 \
+    OLLAMA_MODEL=llama3.2:latest \
+    HTTPX_CLIENT_VERIFY=
+
+# install Python and dependencies
+RUN apk add --no-cache --update \
+    python3 \
+    py3-pip \
+    make \
+    bash \
+    && rm -rf ~/.cache/* /usr/local/share/man /tmp/*
+
+RUN python3 -m pip install pipx --break-system-packages \
+    && python3 -m pipx ensurepath \
+    && python3 -m pipx completions 
 
+# add app src
 COPY . /app/
 
-RUN "./install_dependencies.sh"
+# set workdir
+WORKDIR /app
+
+# set script permissions
+RUN chmod +x entrypoint.sh setup.sh
+
+# run app setup script
+RUN "./setup.sh"
 
+# expose port
 EXPOSE 8000
 
-CMD ["fastapi", "run", "src/api.py"]
+# run app
+ENTRYPOINT ["/app/entrypoint.sh"]

Makefile

@@ -0,0 +1,65 @@
+# Define variables
+POETRY := $(shell command -v poetry 2> /dev/null)
+INSTALL_STAMP := .install.stamp
+
+# Set the default goal
+.DEFAULT_GOAL := help
+
+# Declare phony targets
+.PHONY: help install clean lint format test cleaner run dev
+
+# Help target to display available commands
+help:
+	@echo "Please use 'make <target>' where <target> is one of:"
+	@echo ""
+	@echo " install   - Install packages and prepare environment"
+	@echo " clean     - Remove all temporary files"
+	@echo " cleaner   - Remove all temporary files and the .venv folder"
+	@echo " lint      - Run the code linters"
+	@echo " format    - Reformat code"
+	@echo " test      - Run all the tests"
+	@echo " dev       - Run in dev mode with auto reload"
+	@echo " run       - Run in prod mode with 4 workers"
+	@echo ""
+
+# Install target to set up the environment with Poetry
+install: $(INSTALL_STAMP)
+
+$(INSTALL_STAMP): pyproject.toml
+	@if [ -z $(POETRY) ]; then echo "Poetry could not be found. Please install it."; exit 2; fi
+	$(POETRY) install
+	touch $(INSTALL_STAMP)
+
+# Clean target to remove temporary files and caches
+clean:
+	find . -type d -name "__pycache__" | xargs rm -rf {}
+	rm -rf $(INSTALL_STAMP) .coverage .mypy_cache
+
+# Clean target to remove temporary files and caches
+cleaner:
+	$(MAKE) clean
+	rm -rf .venv
+
+# Lint target to run code linters
+lint: $(INSTALL_STAMP)
+	$(POETRY) run ruff check .
+	$(POETRY) run mypy .
+
+# Format target to reformat code using Black
+format: $(INSTALL_STAMP)
+	$(POETRY) run ruff format .
+
+# Test target to run tests with coverage
+test: $(INSTALL_STAMP)
+	$(POETRY) run pytest .
+
+# Run target to execute the application for production
+dev: $(INSTALL_STAMP)
+	$(POETRY) run uvicorn src.app:app --host 0.0.0.0 --port 8000 --reload
+
+# Run target to execute the application for production
+run: $(INSTALL_STAMP)
+#	$(POETRY) run gunicorn src.app:app --workers 4 --worker-class uvicorn.workers.UvicornWorker
+	$(POETRY) run uvicorn src.app:app --host 0.0.0.0 --port 8000 --reload
+
+

README.md

@@ -4,12 +4,15 @@ A reversible anonymisation service for us all
 # Setup development environment
 
 ```
-python3 -m venv .venv
-. .venv/bin/activate
-./install_dependencies.sh
+./setup.sh
 ```
 # Run
 
 ```
-fastapi run src/api.py
+make run
+```
+# Trouble shooting
+
+```
+echo -n 'db' | gnome-keyring-daemon --unlock
 ```

entrypoint.sh

@@ -0,0 +1,6 @@
+#!/bin/bash
+
+#set -x
+echo "$(cat banner.txt)"
+export PATH="/root/.local/bin:$PATH"
+make run

pyproject.toml

@@ -13,22 +13,31 @@ license = "Apache 2.0"
 authors = ["AI Circle <aicommunity@groups.baloise.com>"]
 
 [tool.poetry.dependencies]
-python = "^3"
-"fastapi[standard]"= "^0"
-langchain-core= "^0"
+python = "^3.12"
+fastapi= {extras = ["standard"], version = "^0"}
 langchain-ollama= "^0"
-ollama= "^0"
 rapidfuzz= "^3"
 regex= "^2024"
 python-dotenv = "^1"
 mypy = "^1"
+setuptools = "^0"
+gunicorn = "^0"
 
 [tool.poetry.dev-dependencies]
 pytest = "^6.0"
+ruff = "^0.7"
 
 [tool.pytest.ini_options]
 minversion = "6.0"
 addopts = "-ra -q"
+python_classes = ["Test"]
+python_functions = ["test_", "it_", "and_", "but_", "they_"]
+python_files = ["test_*.py"]
 testpaths = [
     "tests"
 ]
+filterwarnings = [
+    "error",
+    'ignore:ast.* is deprecated:DeprecationWarning',
+]
+

setup.sh

@@ -0,0 +1,4 @@
+pipx install poetry
+pipx ensurepath
+. ~/.bashrc
+make install

src/api/health.py

@@ -0,0 +1,9 @@
+import fastapi
+from fastapi.responses import JSONResponse
+from fastapi import Request, routing
+
+router = fastapi.APIRouter()
+
+@router.get("/health", response_class=JSONResponse, include_in_schema=False)
+def health(): 
+    return {"status": "ok"}

src/api/mask.py

@@ -0,0 +1,15 @@
+import fastapi
+from fastapi.responses import JSONResponse
+from fastapi import Request, routing
+from pydantic import BaseModel
+from ..utils.llm import llm_find_entities
+
+router = fastapi.APIRouter()
+
+class MaskRequest(BaseModel):
+    text: str
+
+@router.post("/mask", response_class=JSONResponse, include_in_schema=True)
+async def mask(request: MaskRequest):
+    entities = llm_find_entities(request.text)
+    return {"original_text": request.text, "llm_entities": entities}

src/app.py

@@ -0,0 +1,15 @@
+from fastapi import FastAPI
+
+from src.gui import landing
+from src.api import health
+from src.api import mask
+
+app = FastAPI()
+
+app.include_router(landing.router)
+app.include_router(health.router, prefix="/api")
+app.include_router(mask.router, prefix="/api")
+
+if __name__ == "__main__":
+    import uvicorn
+    uvicorn.run(app, host="0.0.0.0", port=8000)

src/gui/landing.py

@@ -0,0 +1,9 @@
+import fastapi
+from fastapi.responses import HTMLResponse
+from fastapi import Request, routing
+
+router = fastapi.APIRouter()
+
+@router.get("/", response_class=HTMLResponse, include_in_schema=False)
+def landing(): 
+    return "<h1>hola mundo</h1>"

src/templates/form.html

@@ -0,0 +1,18 @@
+<!DOCTYPE html>
+<html>
+<head>
+    <title>Simple Form</title>
+</head>
+<body>
+    <h1>Contact Us</h1>
+    <form method="post"> 
+        <label for="name">Name:</label><br>
+        <input type="text" id="name" name="name"><br><br>
+
+        <label for="email">Email:</label><br>
+        <input type="email" id="email" name="email"><br><br>
+
+        <input type="submit" value="Submit"> 
+    </form>
+</body>
+</html>

src/utils/env.py

@@ -0,0 +1,7 @@
+from dotenv import load_dotenv
+import os
+load_dotenv()
+
+OLLAMA_BASE_URL = os.getenv("OLLAMA_BASE_URL")
+OLLAMA_MODEL = os.getenv("OLLAMA_MODEL")
+HTTPX_CLIENT_VERIFY = os.getenv("HTTPX_CLIENT_VERIFY")

src/llm.py → src/utils/llm.py

@@ -1,7 +1,7 @@
 import json
 from langchain_core.prompts import ChatPromptTemplate
 from langchain_ollama.llms import OllamaLLM
-
+from .env import *
 
 TEMPLATE = """
 Context: You anonymize texts in a way that can be reverted and return nothing but a json dictionary that can be parsed automatically.
@@ -14,7 +14,7 @@
 Text to anonymize: {text}
 """
 
-def llm_find_entities(text, model='gemma2', temperature=0, template=TEMPLATE, raw=False):
+def llm_find_entities(text, temperature=0, template=TEMPLATE, raw=False):
     """
     :param text:
     :param model:
@@ -24,7 +24,9 @@ def llm_find_entities(text, model='gemma2', temperature=0, template=TEMPLATE, ra
     :return:
     """
     prompt = ChatPromptTemplate.from_template(template)
-    model = OllamaLLM(model=model, temperature=temperature)
+
+
+    model = OllamaLLM(model=OLLAMA_MODEL, temperature=temperature, base_url=OLLAMA_BASE_URL, client_kwargs={"verify": os.getenv("HTTPX_CLIENT_VERIFY", True)})
     chain = prompt | model
     result = chain.invoke({"text": text})
     if raw:

yoyomaskr.ipynb

@@ -1149,7 +1149,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": ".venv",
+   "display_name": "yoyomaskr",
    "language": "python",
    "name": "python3"
   },
@@ -1163,7 +1163,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.9"
+   "version": "3.12.7"
   }
  },
  "nbformat": 4,