Spaces:
Running
Running
amaye15
commited on
Commit
·
2cb9dec
1
Parent(s):
fd8f07a
Intial Deployment
Browse files- .gitignore +2 -0
- Dockerfile +51 -0
- README.md +3 -0
- docker-compose.yml +43 -0
- requirements.txt +8 -0
- src/api/database.py +596 -0
- src/api/exceptions.py +40 -0
- src/api/models/embedding_models.py +22 -0
- src/api/services/embedding_service.py +62 -0
- src/api/services/huggingface_service.py +69 -0
- src/main.py +177 -0
.gitignore
ADDED
@@ -0,0 +1,2 @@
|
|
|
|
|
|
|
1 |
+
*pycache*
|
2 |
+
*.env*
|
Dockerfile
ADDED
@@ -0,0 +1,51 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Stage 1: Build stage
|
2 |
+
FROM python:3.12-slim as builder
|
3 |
+
|
4 |
+
# Set environment variables
|
5 |
+
ENV PYTHONDONTWRITEBYTECODE=1
|
6 |
+
ENV PYTHONUNBUFFERED=1
|
7 |
+
|
8 |
+
# Create a non-root user
|
9 |
+
RUN useradd -m -u 1000 user
|
10 |
+
|
11 |
+
# Set the working directory
|
12 |
+
WORKDIR /app
|
13 |
+
|
14 |
+
# Copy only the requirements file first to leverage Docker cache
|
15 |
+
COPY --chown=user ./requirements.txt /app/requirements.txt
|
16 |
+
|
17 |
+
# Install dependencies in a virtual environment
|
18 |
+
RUN python -m venv /opt/venv
|
19 |
+
ENV PATH="/opt/venv/bin:$PATH"
|
20 |
+
RUN pip install --no-cache-dir --upgrade pip && \
|
21 |
+
pip install --no-cache-dir -r requirements.txt
|
22 |
+
|
23 |
+
# Copy the rest of the application code
|
24 |
+
COPY --chown=user . /app
|
25 |
+
|
26 |
+
# Stage 2: Runtime stage
|
27 |
+
FROM python:3.12-slim
|
28 |
+
|
29 |
+
# Create a non-root user
|
30 |
+
RUN useradd -m -u 1000 user
|
31 |
+
USER user
|
32 |
+
|
33 |
+
# Copy the virtual environment from the builder stage
|
34 |
+
COPY --from=builder /opt/venv /opt/venv
|
35 |
+
ENV PATH="/opt/venv/bin:$PATH"
|
36 |
+
|
37 |
+
# Set the working directory
|
38 |
+
WORKDIR /app
|
39 |
+
|
40 |
+
# Copy only the necessary files from the builder stage
|
41 |
+
COPY --from=builder --chown=user /app /app
|
42 |
+
|
43 |
+
# Expose the port the app runs on
|
44 |
+
EXPOSE 7860
|
45 |
+
|
46 |
+
# Health check to ensure the application is running
|
47 |
+
HEALTHCHECK --interval=30s --timeout=30s --start-period=5s --retries=3 \
|
48 |
+
CMD curl -f http://localhost:7860/health || exit 1
|
49 |
+
|
50 |
+
# Command to run the application with hot reloading
|
51 |
+
CMD ["uvicorn", "src.main:app", "--host", "0.0.0.0", "--port", "7860", "--reload"]
|
README.md
CHANGED
@@ -6,6 +6,9 @@ colorTo: gray
|
|
6 |
sdk: docker
|
7 |
pinned: false
|
8 |
license: mit
|
|
|
|
|
|
|
9 |
---
|
10 |
|
11 |
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
|
|
6 |
sdk: docker
|
7 |
pinned: false
|
8 |
license: mit
|
9 |
+
python_version: 3.12
|
10 |
+
app_port: 7860
|
11 |
+
app_file: src/main.py
|
12 |
---
|
13 |
|
14 |
Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
|
docker-compose.yml
ADDED
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
version: "3.9"
|
2 |
+
|
3 |
+
services:
|
4 |
+
app:
|
5 |
+
build:
|
6 |
+
context: .
|
7 |
+
dockerfile: Dockerfile
|
8 |
+
container_name: similarity-search-app
|
9 |
+
ports:
|
10 |
+
- "7860:7860"
|
11 |
+
volumes:
|
12 |
+
- ./src:/app/src # Mount the local src directory for hot reloading
|
13 |
+
environment:
|
14 |
+
- PYTHONUNBUFFERED=1
|
15 |
+
restart: unless-stopped
|
16 |
+
healthcheck:
|
17 |
+
test: ["CMD", "curl", "-f", "http://localhost:7860/health"]
|
18 |
+
interval: 30s
|
19 |
+
timeout: 10s
|
20 |
+
retries: 3
|
21 |
+
# depends_on:
|
22 |
+
# - db # If you have a database service, add it here
|
23 |
+
|
24 |
+
# # Example database service (optional)
|
25 |
+
# db:
|
26 |
+
# image: postgres:latest
|
27 |
+
# container_name: similarity-search-db
|
28 |
+
# environment:
|
29 |
+
# POSTGRES_USER: user
|
30 |
+
# POSTGRES_PASSWORD: password
|
31 |
+
# POSTGRES_DB: mydatabase
|
32 |
+
# ports:
|
33 |
+
# - "5432:5432"
|
34 |
+
# volumes:
|
35 |
+
# - postgres_data:/var/lib/postgresql/data
|
36 |
+
# healthcheck:
|
37 |
+
# test: ["CMD-SHELL", "pg_isready -U user -d mydatabase"]
|
38 |
+
# interval: 5s
|
39 |
+
# timeout: 5s
|
40 |
+
# retries: 5
|
41 |
+
|
42 |
+
# volumes:
|
43 |
+
# postgres_data:
|
requirements.txt
ADDED
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
pg8000
|
2 |
+
pydantic
|
3 |
+
pydantic-settings
|
4 |
+
uvicorn
|
5 |
+
fastapi
|
6 |
+
openai
|
7 |
+
pandas
|
8 |
+
datasets
|
src/api/database.py
ADDED
@@ -0,0 +1,596 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# import logging
|
2 |
+
# from typing import Dict, List, Optional, AsyncGenerator
|
3 |
+
# from pydantic import BaseSettings, PostgresDsn
|
4 |
+
# import pg8000
|
5 |
+
# from pg8000 import Connection, Cursor
|
6 |
+
# from pg8000.exceptions import DatabaseError
|
7 |
+
# import asyncio
|
8 |
+
# from contextlib import asynccontextmanager
|
9 |
+
# from dataclasses import dataclass
|
10 |
+
# from threading import Lock
|
11 |
+
|
12 |
+
# # Set up structured logging
|
13 |
+
# logging.basicConfig(
|
14 |
+
# level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
|
15 |
+
# )
|
16 |
+
# logger = logging.getLogger(__name__)
|
17 |
+
|
18 |
+
|
19 |
+
# class DatabaseSettings(BaseSettings):
|
20 |
+
# db_url: PostgresDsn
|
21 |
+
# pool_size: int = 5
|
22 |
+
|
23 |
+
# class Config:
|
24 |
+
# env_file = ".env"
|
25 |
+
|
26 |
+
|
27 |
+
# @dataclass
|
28 |
+
# class DatabaseConfig:
|
29 |
+
# username: str
|
30 |
+
# password: str
|
31 |
+
# hostname: str
|
32 |
+
# port: int
|
33 |
+
# database: str
|
34 |
+
|
35 |
+
|
36 |
+
# class DatabaseError(Exception):
|
37 |
+
# """Custom exception for database errors."""
|
38 |
+
|
39 |
+
# pass
|
40 |
+
|
41 |
+
|
42 |
+
# class Database:
|
43 |
+
# def __init__(self, db_url: str, pool_size: int):
|
44 |
+
# self.db_url = db_url
|
45 |
+
# self.pool_size = pool_size
|
46 |
+
# self.pool: List[Connection] = []
|
47 |
+
# self.lock = Lock()
|
48 |
+
# self.config = self._parse_db_url()
|
49 |
+
|
50 |
+
# def _parse_db_url(self) -> DatabaseConfig:
|
51 |
+
# """Parse the database URL into components."""
|
52 |
+
# result = urlparse(self.db_url)
|
53 |
+
# return DatabaseConfig(
|
54 |
+
# username=result.username,
|
55 |
+
# password=result.password,
|
56 |
+
# hostname=result.hostname,
|
57 |
+
# port=result.port or 5432,
|
58 |
+
# database=result.path.lstrip("/"),
|
59 |
+
# )
|
60 |
+
|
61 |
+
# async def connect(self) -> None:
|
62 |
+
# """Create a connection pool."""
|
63 |
+
# try:
|
64 |
+
# for _ in range(self.pool_size):
|
65 |
+
# conn = await self._create_connection()
|
66 |
+
# self.pool.append(conn)
|
67 |
+
# logger.info(
|
68 |
+
# f"Database connection pool created with {self.pool_size} connections."
|
69 |
+
# )
|
70 |
+
# except DatabaseError as e:
|
71 |
+
# logger.error(f"Failed to create database connection pool: {e}")
|
72 |
+
# raise
|
73 |
+
|
74 |
+
# async def _create_connection(self) -> Connection:
|
75 |
+
# """Create a single database connection."""
|
76 |
+
# try:
|
77 |
+
# conn = pg8000.connect(
|
78 |
+
# user=self.config.username,
|
79 |
+
# password=self.config.password,
|
80 |
+
# host=self.config.hostname,
|
81 |
+
# port=self.config.port,
|
82 |
+
# database=self.config.database,
|
83 |
+
# )
|
84 |
+
# return conn
|
85 |
+
# except DatabaseError as e:
|
86 |
+
# logger.error(f"Failed to create database connection: {e}")
|
87 |
+
# raise DatabaseError("Failed to create database connection.")
|
88 |
+
|
89 |
+
# async def disconnect(self) -> None:
|
90 |
+
# """Close all connections in the pool."""
|
91 |
+
# with self.lock:
|
92 |
+
# for conn in self.pool:
|
93 |
+
# conn.close()
|
94 |
+
# self.pool.clear()
|
95 |
+
# logger.info("Database connection pool closed.")
|
96 |
+
|
97 |
+
# @asynccontextmanager
|
98 |
+
# async def get_connection(self) -> AsyncGenerator[Connection, None]:
|
99 |
+
# """Acquire a connection from the pool."""
|
100 |
+
# with self.lock:
|
101 |
+
# if not self.pool:
|
102 |
+
# raise DatabaseError("Database connection pool is empty.")
|
103 |
+
# conn = self.pool.pop()
|
104 |
+
# try:
|
105 |
+
# yield conn
|
106 |
+
# finally:
|
107 |
+
# with self.lock:
|
108 |
+
# self.pool.append(conn)
|
109 |
+
|
110 |
+
# async def fetch(self, query: str, *args) -> List[Dict]:
|
111 |
+
# """
|
112 |
+
# Execute a SELECT query and return the results as a list of dictionaries.
|
113 |
+
|
114 |
+
# Args:
|
115 |
+
# query (str): The SQL query to execute.
|
116 |
+
# *args: Query parameters.
|
117 |
+
|
118 |
+
# Returns:
|
119 |
+
# List[Dict]: A list of dictionaries where keys are column names and values are column values.
|
120 |
+
# """
|
121 |
+
# try:
|
122 |
+
# async with self.get_connection() as conn:
|
123 |
+
# cursor: Cursor = conn.cursor()
|
124 |
+
# cursor.execute(query, args)
|
125 |
+
# rows = cursor.fetchall()
|
126 |
+
# columns = [desc[0] for desc in cursor.description]
|
127 |
+
# return [dict(zip(columns, row)) for row in rows]
|
128 |
+
# except DatabaseError as e:
|
129 |
+
# logger.error(f"Error executing query: {query}. Error: {e}")
|
130 |
+
# raise DatabaseError(f"Failed to execute query: {query}")
|
131 |
+
|
132 |
+
# async def execute(self, query: str, *args) -> None:
|
133 |
+
# """
|
134 |
+
# Execute an INSERT, UPDATE, or DELETE query.
|
135 |
+
|
136 |
+
# Args:
|
137 |
+
# query (str): The SQL query to execute.
|
138 |
+
# *args: Query parameters.
|
139 |
+
# """
|
140 |
+
# try:
|
141 |
+
# async with self.get_connection() as conn:
|
142 |
+
# cursor: Cursor = conn.cursor()
|
143 |
+
# cursor.execute(query, args)
|
144 |
+
# conn.commit()
|
145 |
+
# except DatabaseError as e:
|
146 |
+
# logger.error(f"Error executing query: {query}. Error: {e}")
|
147 |
+
# raise DatabaseError(f"Failed to execute query: {query}")
|
148 |
+
|
149 |
+
|
150 |
+
# # Dependency to get the database instance
|
151 |
+
# async def get_db() -> AsyncGenerator[Database, None]:
|
152 |
+
# settings = DatabaseSettings()
|
153 |
+
# db = Database(db_url=settings.db_url, pool_size=settings.pool_size)
|
154 |
+
# await db.connect()
|
155 |
+
# try:
|
156 |
+
# yield db
|
157 |
+
# finally:
|
158 |
+
# await db.disconnect()
|
159 |
+
|
160 |
+
|
161 |
+
# # Example usage
|
162 |
+
# if __name__ == "__main__":
|
163 |
+
|
164 |
+
# async def main():
|
165 |
+
# settings = DatabaseSettings()
|
166 |
+
# db = Database(db_url=settings.db_url, pool_size=settings.pool_size)
|
167 |
+
# await db.connect()
|
168 |
+
|
169 |
+
# try:
|
170 |
+
# # Example query
|
171 |
+
# query = """
|
172 |
+
# SELECT
|
173 |
+
# ppt.type AS product_type,
|
174 |
+
# pc.name AS product_category
|
175 |
+
# FROM
|
176 |
+
# product_producttype ppt
|
177 |
+
# INNER JOIN
|
178 |
+
# product_category pc
|
179 |
+
# ON
|
180 |
+
# ppt.category_id = pc.id
|
181 |
+
# """
|
182 |
+
# result = await db.fetch(query)
|
183 |
+
# print(result)
|
184 |
+
# finally:
|
185 |
+
# await db.disconnect()
|
186 |
+
|
187 |
+
# asyncio.run(main())
|
188 |
+
|
189 |
+
# import logging
|
190 |
+
# from urllib.parse import urlparse
|
191 |
+
# from typing import Dict, List, Optional, AsyncGenerator
|
192 |
+
# from pydantic_settings import BaseSettings
|
193 |
+
# from pydantic import PostgresDsn
|
194 |
+
# import pg8000
|
195 |
+
# from pg8000 import Connection, Cursor
|
196 |
+
# from pg8000.exceptions import DatabaseError
|
197 |
+
# import asyncio
|
198 |
+
# from contextlib import asynccontextmanager
|
199 |
+
# from dataclasses import dataclass
|
200 |
+
# from threading import Lock
|
201 |
+
|
202 |
+
# # Set up structured logging
|
203 |
+
# logging.basicConfig(
|
204 |
+
# level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
|
205 |
+
# )
|
206 |
+
# logger = logging.getLogger(__name__)
|
207 |
+
|
208 |
+
|
209 |
+
# class DatabaseSettings(BaseSettings):
|
210 |
+
# db_url: PostgresDsn
|
211 |
+
# pool_size: int = 5
|
212 |
+
|
213 |
+
# class Config:
|
214 |
+
# env_file = ".env"
|
215 |
+
|
216 |
+
|
217 |
+
# @dataclass
|
218 |
+
# class DatabaseConfig:
|
219 |
+
# username: str
|
220 |
+
# password: str
|
221 |
+
# hostname: str
|
222 |
+
# port: int
|
223 |
+
# database: str
|
224 |
+
|
225 |
+
|
226 |
+
# class DatabaseError(Exception):
|
227 |
+
# """Custom exception for database errors."""
|
228 |
+
|
229 |
+
# pass
|
230 |
+
|
231 |
+
|
232 |
+
# class Database:
|
233 |
+
# def __init__(self, db_url: str, pool_size: int):
|
234 |
+
# self.db_url = db_url
|
235 |
+
# self.pool_size = pool_size
|
236 |
+
# self.pool: List[Connection] = []
|
237 |
+
# self.lock = Lock()
|
238 |
+
# self.config = self._parse_db_url()
|
239 |
+
|
240 |
+
# def _parse_db_url(self) -> DatabaseConfig:
|
241 |
+
# """Parse the database URL into components."""
|
242 |
+
# # Convert PostgresDsn to a string
|
243 |
+
# db_url_str = str(self.db_url)
|
244 |
+
# result = urlparse(db_url_str)
|
245 |
+
# return DatabaseConfig(
|
246 |
+
# username=result.username,
|
247 |
+
# password=result.password,
|
248 |
+
# hostname=result.hostname,
|
249 |
+
# port=result.port or 5432,
|
250 |
+
# database=result.path.lstrip("/"),
|
251 |
+
# )
|
252 |
+
|
253 |
+
# async def connect(self) -> None:
|
254 |
+
# """Create a connection pool."""
|
255 |
+
# try:
|
256 |
+
# for _ in range(self.pool_size):
|
257 |
+
# conn = await self._create_connection()
|
258 |
+
# self.pool.append(conn)
|
259 |
+
# logger.info(
|
260 |
+
# f"Database connection pool created with {self.pool_size} connections."
|
261 |
+
# )
|
262 |
+
# except DatabaseError as e:
|
263 |
+
# logger.error(f"Failed to create database connection pool: {e}")
|
264 |
+
# raise
|
265 |
+
|
266 |
+
# async def _create_connection(self) -> Connection:
|
267 |
+
# """Create a single database connection."""
|
268 |
+
# try:
|
269 |
+
# conn = pg8000.connect(
|
270 |
+
# user=self.config.username,
|
271 |
+
# password=self.config.password,
|
272 |
+
# host=self.config.hostname,
|
273 |
+
# port=self.config.port,
|
274 |
+
# database=self.config.database,
|
275 |
+
# )
|
276 |
+
# return conn
|
277 |
+
# except DatabaseError as e:
|
278 |
+
# logger.error(f"Failed to create database connection: {e}")
|
279 |
+
# raise DatabaseError("Failed to create database connection.")
|
280 |
+
|
281 |
+
# async def disconnect(self) -> None:
|
282 |
+
# """Close all connections in the pool."""
|
283 |
+
# with self.lock:
|
284 |
+
# for conn in self.pool:
|
285 |
+
# conn.close()
|
286 |
+
# self.pool.clear()
|
287 |
+
# logger.info("Database connection pool closed.")
|
288 |
+
|
289 |
+
# @asynccontextmanager
|
290 |
+
# async def get_connection(self) -> AsyncGenerator[Connection, None]:
|
291 |
+
# """Acquire a connection from the pool."""
|
292 |
+
# with self.lock:
|
293 |
+
# if not self.pool:
|
294 |
+
# raise DatabaseError("Database connection pool is empty.")
|
295 |
+
# conn = self.pool.pop()
|
296 |
+
# try:
|
297 |
+
# yield conn
|
298 |
+
# finally:
|
299 |
+
# with self.lock:
|
300 |
+
# self.pool.append(conn)
|
301 |
+
|
302 |
+
# async def fetch(self, query: str, *args) -> List[Dict]:
|
303 |
+
# """
|
304 |
+
# Execute a SELECT query and return the results as a list of dictionaries.
|
305 |
+
|
306 |
+
# Args:
|
307 |
+
# query (str): The SQL query to execute.
|
308 |
+
# *args: Query parameters.
|
309 |
+
|
310 |
+
# Returns:
|
311 |
+
# List[Dict]: A list of dictionaries where keys are column names and values are column values.
|
312 |
+
# """
|
313 |
+
# try:
|
314 |
+
# async with self.get_connection() as conn:
|
315 |
+
# cursor: Cursor = conn.cursor()
|
316 |
+
# cursor.execute(query, args)
|
317 |
+
# rows = cursor.fetchall()
|
318 |
+
# columns = [desc[0] for desc in cursor.description]
|
319 |
+
# return [dict(zip(columns, row)) for row in rows]
|
320 |
+
# except DatabaseError as e:
|
321 |
+
# logger.error(f"Error executing query: {query}. Error: {e}")
|
322 |
+
# raise DatabaseError(f"Failed to execute query: {query}")
|
323 |
+
|
324 |
+
# async def execute(self, query: str, *args) -> None:
|
325 |
+
# """
|
326 |
+
# Execute an INSERT, UPDATE, or DELETE query.
|
327 |
+
|
328 |
+
# Args:
|
329 |
+
# query (str): The SQL query to execute.
|
330 |
+
# *args: Query parameters.
|
331 |
+
# """
|
332 |
+
# try:
|
333 |
+
# async with self.get_connection() as conn:
|
334 |
+
# cursor: Cursor = conn.cursor()
|
335 |
+
# cursor.execute(query, args)
|
336 |
+
# conn.commit()
|
337 |
+
# except DatabaseError as e:
|
338 |
+
# logger.error(f"Error executing query: {query}. Error: {e}")
|
339 |
+
# raise DatabaseError(f"Failed to execute query: {query}")
|
340 |
+
|
341 |
+
|
342 |
+
# # Dependency to get the database instance
|
343 |
+
# async def get_db() -> AsyncGenerator[Database, None]:
|
344 |
+
# settings = DatabaseSettings()
|
345 |
+
# db = Database(db_url=settings.db_url, pool_size=settings.pool_size)
|
346 |
+
# await db.connect()
|
347 |
+
# try:
|
348 |
+
# yield db
|
349 |
+
# finally:
|
350 |
+
# await db.disconnect()
|
351 |
+
|
352 |
+
|
353 |
+
# # Example usage
|
354 |
+
# if __name__ == "__main__":
|
355 |
+
|
356 |
+
# async def main():
|
357 |
+
# settings = DatabaseSettings()
|
358 |
+
# db = Database(db_url=settings.db_url, pool_size=settings.pool_size)
|
359 |
+
# await db.connect()
|
360 |
+
|
361 |
+
# try:
|
362 |
+
# # Example query
|
363 |
+
# query = "SELECT * FROM your_table LIMIT 10"
|
364 |
+
# query = """
|
365 |
+
# SELECT
|
366 |
+
# ppt.type AS product_type,
|
367 |
+
# pc.name AS product_category
|
368 |
+
# FROM
|
369 |
+
# product_producttype ppt
|
370 |
+
# INNER JOIN
|
371 |
+
# product_category pc
|
372 |
+
# ON
|
373 |
+
# ppt.category_id = pc.id
|
374 |
+
# """
|
375 |
+
# result = await db.fetch(query)
|
376 |
+
# print(result)
|
377 |
+
# finally:
|
378 |
+
# await db.disconnect()
|
379 |
+
|
380 |
+
# asyncio.run(main())
|
381 |
+
|
382 |
+
import logging
|
383 |
+
from typing import AsyncGenerator, List, Optional, Dict
|
384 |
+
from pydantic_settings import BaseSettings
|
385 |
+
from pydantic import PostgresDsn
|
386 |
+
import pg8000
|
387 |
+
from pg8000 import Connection
|
388 |
+
from pg8000.exceptions import DatabaseError as Pg8000DatabaseError
|
389 |
+
import asyncio
|
390 |
+
from contextlib import asynccontextmanager
|
391 |
+
from threading import Lock
|
392 |
+
from urllib.parse import urlparse
|
393 |
+
|
394 |
+
# Set up structured logging
|
395 |
+
logging.basicConfig(
|
396 |
+
level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
|
397 |
+
)
|
398 |
+
logger = logging.getLogger(__name__)
|
399 |
+
|
400 |
+
|
401 |
+
class DatabaseSettings(BaseSettings):
|
402 |
+
db_url: PostgresDsn
|
403 |
+
pool_size: int = 5 # Default pool size is 5
|
404 |
+
|
405 |
+
class Config:
|
406 |
+
env_file = ".env"
|
407 |
+
|
408 |
+
|
409 |
+
# Custom database errors
|
410 |
+
class DatabaseError(Exception):
|
411 |
+
"""Base exception for database errors."""
|
412 |
+
|
413 |
+
pass
|
414 |
+
|
415 |
+
|
416 |
+
class ConnectionError(DatabaseError):
|
417 |
+
"""Exception raised when a database connection fails."""
|
418 |
+
|
419 |
+
pass
|
420 |
+
|
421 |
+
|
422 |
+
class PoolExhaustedError(DatabaseError):
|
423 |
+
"""Exception raised when the connection pool is exhausted."""
|
424 |
+
|
425 |
+
pass
|
426 |
+
|
427 |
+
|
428 |
+
class QueryExecutionError(DatabaseError):
|
429 |
+
"""Exception raised when a query execution fails."""
|
430 |
+
|
431 |
+
pass
|
432 |
+
|
433 |
+
|
434 |
+
class HealthCheckError(DatabaseError):
|
435 |
+
"""Exception raised when a health check fails."""
|
436 |
+
|
437 |
+
pass
|
438 |
+
|
439 |
+
|
440 |
+
class Database:
|
441 |
+
def __init__(self, db_url: PostgresDsn, pool_size: int):
|
442 |
+
self.db_url = db_url
|
443 |
+
self.pool_size = pool_size
|
444 |
+
self.pool: List[Connection] = []
|
445 |
+
self.lock = Lock()
|
446 |
+
|
447 |
+
async def connect(self) -> None:
|
448 |
+
"""Create a connection pool."""
|
449 |
+
try:
|
450 |
+
# Convert PostgresDsn to a string
|
451 |
+
db_url_str = str(self.db_url)
|
452 |
+
result = urlparse(db_url_str)
|
453 |
+
for _ in range(self.pool_size):
|
454 |
+
conn = pg8000.connect(
|
455 |
+
user=result.username,
|
456 |
+
password=result.password,
|
457 |
+
host=result.hostname,
|
458 |
+
port=result.port or 5432,
|
459 |
+
database=result.path.lstrip("/"),
|
460 |
+
)
|
461 |
+
self.pool.append(conn)
|
462 |
+
logger.info(
|
463 |
+
f"Database connection pool created with {self.pool_size} connections."
|
464 |
+
)
|
465 |
+
except Pg8000DatabaseError as e:
|
466 |
+
logger.error(f"Failed to create database connection pool: {e}")
|
467 |
+
raise ConnectionError("Failed to create database connection pool.") from e
|
468 |
+
|
469 |
+
async def disconnect(self) -> None:
|
470 |
+
"""Close all connections in the pool."""
|
471 |
+
with self.lock:
|
472 |
+
for conn in self.pool:
|
473 |
+
conn.close()
|
474 |
+
self.pool.clear()
|
475 |
+
logger.info("Database connection pool closed.")
|
476 |
+
|
477 |
+
@asynccontextmanager
|
478 |
+
async def get_connection(self) -> AsyncGenerator[Connection, None]:
|
479 |
+
"""Acquire a connection from the pool."""
|
480 |
+
with self.lock:
|
481 |
+
if not self.pool:
|
482 |
+
logger.error("Connection pool is exhausted.")
|
483 |
+
raise PoolExhaustedError("No available connections in the pool.")
|
484 |
+
conn = self.pool.pop()
|
485 |
+
try:
|
486 |
+
yield conn
|
487 |
+
except Pg8000DatabaseError as e:
|
488 |
+
logger.error(f"Connection error: {e}")
|
489 |
+
raise ConnectionError("Failed to use database connection.") from e
|
490 |
+
finally:
|
491 |
+
with self.lock:
|
492 |
+
self.pool.append(conn)
|
493 |
+
|
494 |
+
async def fetch(self, query: str, *args) -> List[Dict]:
|
495 |
+
"""
|
496 |
+
Execute a SELECT query and return the results as a list of dictionaries.
|
497 |
+
|
498 |
+
Args:
|
499 |
+
query (str): The SQL query to execute.
|
500 |
+
*args: Query parameters.
|
501 |
+
|
502 |
+
Returns:
|
503 |
+
List[Dict]: A list of dictionaries where keys are column names and values are column values.
|
504 |
+
|
505 |
+
Raises:
|
506 |
+
QueryExecutionError: If the query execution fails.
|
507 |
+
"""
|
508 |
+
try:
|
509 |
+
async with self.get_connection() as conn:
|
510 |
+
cursor = conn.cursor()
|
511 |
+
cursor.execute(query, args)
|
512 |
+
rows = cursor.fetchall()
|
513 |
+
columns = [desc[0] for desc in cursor.description]
|
514 |
+
return [dict(zip(columns, row)) for row in rows]
|
515 |
+
except Pg8000DatabaseError as e:
|
516 |
+
logger.error(f"Query execution failed: {e}")
|
517 |
+
raise QueryExecutionError(f"Failed to execute query: {query}") from e
|
518 |
+
|
519 |
+
async def execute(self, query: str, *args) -> None:
|
520 |
+
"""
|
521 |
+
Execute an INSERT, UPDATE, or DELETE query.
|
522 |
+
|
523 |
+
Args:
|
524 |
+
query (str): The SQL query to execute.
|
525 |
+
*args: Query parameters.
|
526 |
+
|
527 |
+
Raises:
|
528 |
+
QueryExecutionError: If the query execution fails.
|
529 |
+
"""
|
530 |
+
try:
|
531 |
+
async with self.get_connection() as conn:
|
532 |
+
cursor = conn.cursor()
|
533 |
+
cursor.execute(query, args)
|
534 |
+
conn.commit()
|
535 |
+
except Pg8000DatabaseError as e:
|
536 |
+
logger.error(f"Query execution failed: {e}")
|
537 |
+
raise QueryExecutionError(f"Failed to execute query: {query}") from e
|
538 |
+
|
539 |
+
async def health_check(self) -> bool:
|
540 |
+
"""
|
541 |
+
Perform a health check by executing a simple query (e.g., SELECT 1).
|
542 |
+
|
543 |
+
Returns:
|
544 |
+
bool: True if the database is healthy, False otherwise.
|
545 |
+
|
546 |
+
Raises:
|
547 |
+
HealthCheckError: If the health check fails.
|
548 |
+
"""
|
549 |
+
try:
|
550 |
+
async with self.get_connection() as conn:
|
551 |
+
cursor = conn.cursor()
|
552 |
+
cursor.execute("SELECT 1")
|
553 |
+
result = cursor.fetchone()
|
554 |
+
cursor.close()
|
555 |
+
|
556 |
+
# Check if the result is as expected
|
557 |
+
if result and result[0] == 1:
|
558 |
+
logger.info("Database health check succeeded.")
|
559 |
+
return True
|
560 |
+
else:
|
561 |
+
logger.error("Database health check failed: Unexpected result.")
|
562 |
+
raise HealthCheckError("Unexpected result from health check query.")
|
563 |
+
except Pg8000DatabaseError as e:
|
564 |
+
logger.error(f"Health check failed: {e}")
|
565 |
+
raise HealthCheckError("Failed to perform health check.") from e
|
566 |
+
|
567 |
+
|
568 |
+
# Dependency to get the database instance
|
569 |
+
async def get_db() -> AsyncGenerator[Database, None]:
|
570 |
+
settings = DatabaseSettings()
|
571 |
+
db = Database(db_url=settings.db_url, pool_size=settings.pool_size)
|
572 |
+
await db.connect()
|
573 |
+
try:
|
574 |
+
yield db
|
575 |
+
finally:
|
576 |
+
await db.disconnect()
|
577 |
+
|
578 |
+
|
579 |
+
# Example usage
|
580 |
+
if __name__ == "__main__":
|
581 |
+
|
582 |
+
async def main():
|
583 |
+
settings = DatabaseSettings()
|
584 |
+
db = Database(db_url=settings.db_url, pool_size=settings.pool_size)
|
585 |
+
await db.connect()
|
586 |
+
|
587 |
+
try:
|
588 |
+
# Perform a health check
|
589 |
+
is_healthy = await db.health_check()
|
590 |
+
print(f"Database health check: {'Success' if is_healthy else 'Failure'}")
|
591 |
+
except HealthCheckError as e:
|
592 |
+
print(f"Health check failed: {e}")
|
593 |
+
finally:
|
594 |
+
await db.disconnect()
|
595 |
+
|
596 |
+
asyncio.run(main())
|
src/api/exceptions.py
ADDED
@@ -0,0 +1,40 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
class DatabaseError(Exception):
|
2 |
+
"""Base exception for database errors."""
|
3 |
+
|
4 |
+
pass
|
5 |
+
|
6 |
+
|
7 |
+
class QueryExecutionError(DatabaseError):
|
8 |
+
"""Exception raised when a database query fails."""
|
9 |
+
|
10 |
+
pass
|
11 |
+
|
12 |
+
|
13 |
+
class EmbeddingError(Exception):
|
14 |
+
"""Base exception for embedding-related errors."""
|
15 |
+
|
16 |
+
pass
|
17 |
+
|
18 |
+
|
19 |
+
class OpenAIError(EmbeddingError):
|
20 |
+
"""Exception raised when OpenAI API fails."""
|
21 |
+
|
22 |
+
pass
|
23 |
+
|
24 |
+
|
25 |
+
class HuggingFaceError(Exception):
|
26 |
+
"""Base exception for Hugging Face-related errors."""
|
27 |
+
|
28 |
+
pass
|
29 |
+
|
30 |
+
|
31 |
+
class DatasetNotFoundError(HuggingFaceError):
|
32 |
+
"""Exception raised when a dataset is not found."""
|
33 |
+
|
34 |
+
pass
|
35 |
+
|
36 |
+
|
37 |
+
class DatasetPushError(HuggingFaceError):
|
38 |
+
"""Exception raised when pushing a dataset to Hugging Face Hub fails."""
|
39 |
+
|
40 |
+
pass
|
src/api/models/embedding_models.py
ADDED
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from pydantic import BaseModel
|
2 |
+
from typing import List, Dict
|
3 |
+
|
4 |
+
|
5 |
+
# Pydantic models for request validation
|
6 |
+
class CreateEmbeddingRequest(BaseModel):
|
7 |
+
query: str
|
8 |
+
target_column: str = "product_type"
|
9 |
+
output_column: str = "embedding"
|
10 |
+
model: str = "text-embedding-3-small"
|
11 |
+
batch_size: int = 100
|
12 |
+
dataset_name: str = "re-mind/product_type_embedding"
|
13 |
+
|
14 |
+
|
15 |
+
class UpdateEmbeddingRequest(BaseModel):
|
16 |
+
dataset_name: str
|
17 |
+
updates: Dict[str, List] # Column name -> List of values
|
18 |
+
|
19 |
+
|
20 |
+
class DeleteEmbeddingRequest(BaseModel):
|
21 |
+
dataset_name: str
|
22 |
+
columns: List[str] # List of columns to delete
|
src/api/services/embedding_service.py
ADDED
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from openai import AsyncOpenAI
|
2 |
+
import logging
|
3 |
+
from typing import List, Dict
|
4 |
+
import pandas as pd
|
5 |
+
import asyncio
|
6 |
+
from src.api.exceptions import OpenAIError
|
7 |
+
|
8 |
+
# Set up structured logging
|
9 |
+
logging.basicConfig(
|
10 |
+
level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
|
11 |
+
)
|
12 |
+
logger = logging.getLogger(__name__)
|
13 |
+
|
14 |
+
|
15 |
+
class EmbeddingService:
|
16 |
+
def __init__(
|
17 |
+
self,
|
18 |
+
openai_api_key: str,
|
19 |
+
model: str = "text-embedding-3-small",
|
20 |
+
batch_size: int = 100,
|
21 |
+
):
|
22 |
+
self.client = AsyncOpenAI(api_key=openai_api_key)
|
23 |
+
self.model = model
|
24 |
+
self.batch_size = batch_size
|
25 |
+
|
26 |
+
async def get_embedding(self, text: str) -> List[float]:
|
27 |
+
"""Generate embeddings for the given text using OpenAI."""
|
28 |
+
text = text.replace("\n", " ")
|
29 |
+
try:
|
30 |
+
response = await self.client.embeddings.create(
|
31 |
+
input=[text], model=self.model
|
32 |
+
)
|
33 |
+
return response.data[0].embedding
|
34 |
+
except Exception as e:
|
35 |
+
logger.error(f"Failed to generate embedding: {e}")
|
36 |
+
raise OpenAIError(f"OpenAI API error: {e}")
|
37 |
+
|
38 |
+
async def create_embeddings(
|
39 |
+
self, df: pd.DataFrame, target_column: str, output_column: str
|
40 |
+
) -> pd.DataFrame:
|
41 |
+
"""Create embeddings for the target column in the dataset."""
|
42 |
+
logger.info("Generating embeddings...")
|
43 |
+
batches = [
|
44 |
+
df[i : i + self.batch_size] for i in range(0, len(df), self.batch_size)
|
45 |
+
]
|
46 |
+
processed_batches = await asyncio.gather(
|
47 |
+
*[
|
48 |
+
self._process_batch(batch, target_column, output_column)
|
49 |
+
for batch in batches
|
50 |
+
]
|
51 |
+
)
|
52 |
+
return pd.concat(processed_batches)
|
53 |
+
|
54 |
+
async def _process_batch(
|
55 |
+
self, df_batch: pd.DataFrame, target_column: str, output_column: str
|
56 |
+
) -> pd.DataFrame:
|
57 |
+
"""Process a batch of rows to generate embeddings."""
|
58 |
+
embeddings = await asyncio.gather(
|
59 |
+
*[self.get_embedding(row[target_column]) for _, row in df_batch.iterrows()]
|
60 |
+
)
|
61 |
+
df_batch[output_column] = embeddings
|
62 |
+
return df_batch
|
src/api/services/huggingface_service.py
ADDED
@@ -0,0 +1,69 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
from datasets import Dataset, load_dataset
|
2 |
+
import logging
|
3 |
+
from typing import Optional, Dict, List
|
4 |
+
import pandas as pd
|
5 |
+
from src.api.exceptions import DatasetNotFoundError, DatasetPushError
|
6 |
+
|
7 |
+
# Set up structured logging
|
8 |
+
logging.basicConfig(
|
9 |
+
level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
|
10 |
+
)
|
11 |
+
logger = logging.getLogger(__name__)
|
12 |
+
|
13 |
+
|
14 |
+
class HuggingFaceService:
|
15 |
+
async def push_to_hub(self, df: pd.DataFrame, dataset_name: str) -> None:
|
16 |
+
"""Push the dataset to Hugging Face Hub."""
|
17 |
+
try:
|
18 |
+
logger.info(f"Creating Hugging Face Dataset: {dataset_name}...")
|
19 |
+
ds = Dataset.from_pandas(df).remove_columns("__index_level_0__")
|
20 |
+
ds.push_to_hub(dataset_name)
|
21 |
+
logger.info(f"Dataset pushed to Hugging Face Hub: {dataset_name}")
|
22 |
+
except Exception as e:
|
23 |
+
logger.error(f"Failed to push dataset to Hugging Face Hub: {e}")
|
24 |
+
raise DatasetPushError(f"Failed to push dataset: {e}")
|
25 |
+
|
26 |
+
async def read_dataset(self, dataset_name: str) -> Optional[pd.DataFrame]:
|
27 |
+
"""Read a dataset from Hugging Face Hub."""
|
28 |
+
try:
|
29 |
+
logger.info(f"Loading dataset from Hugging Face Hub: {dataset_name}...")
|
30 |
+
ds = load_dataset(dataset_name)
|
31 |
+
df = ds["train"].to_pandas()
|
32 |
+
return df
|
33 |
+
except Exception as e:
|
34 |
+
logger.error(f"Failed to read dataset: {e}")
|
35 |
+
raise DatasetNotFoundError(f"Dataset not found: {e}")
|
36 |
+
|
37 |
+
async def update_dataset(
|
38 |
+
self, dataset_name: str, updates: Dict[str, List]
|
39 |
+
) -> Optional[pd.DataFrame]:
|
40 |
+
"""Update a dataset on Hugging Face Hub."""
|
41 |
+
try:
|
42 |
+
df = await self.read_dataset(dataset_name)
|
43 |
+
for column, values in updates.items():
|
44 |
+
if column in df.columns:
|
45 |
+
df[column] = values
|
46 |
+
else:
|
47 |
+
logger.warning(f"Column '{column}' not found in dataset.")
|
48 |
+
await self.push_to_hub(df, dataset_name)
|
49 |
+
return df
|
50 |
+
except Exception as e:
|
51 |
+
logger.error(f"Failed to update dataset: {e}")
|
52 |
+
raise DatasetPushError(f"Failed to update dataset: {e}")
|
53 |
+
|
54 |
+
async def delete_columns(
|
55 |
+
self, dataset_name: str, columns: List[str]
|
56 |
+
) -> Optional[pd.DataFrame]:
|
57 |
+
"""Delete columns from a dataset on Hugging Face Hub."""
|
58 |
+
try:
|
59 |
+
df = await self.read_dataset(dataset_name)
|
60 |
+
for column in columns:
|
61 |
+
if column in df.columns:
|
62 |
+
df.drop(column, axis=1, inplace=True)
|
63 |
+
else:
|
64 |
+
logger.warning(f"Column '{column}' not found in dataset.")
|
65 |
+
await self.push_to_hub(df, dataset_name)
|
66 |
+
return df
|
67 |
+
except Exception as e:
|
68 |
+
logger.error(f"Failed to delete columns: {e}")
|
69 |
+
raise DatasetPushError(f"Failed to delete columns: {e}")
|
src/main.py
ADDED
@@ -0,0 +1,177 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
import os
|
2 |
+
from fastapi import FastAPI, Depends, HTTPException
|
3 |
+
from fastapi.responses import JSONResponse, RedirectResponse
|
4 |
+
from pydantic import BaseModel
|
5 |
+
from typing import List, Dict
|
6 |
+
from src.api.models.embedding_models import (
|
7 |
+
CreateEmbeddingRequest,
|
8 |
+
UpdateEmbeddingRequest,
|
9 |
+
DeleteEmbeddingRequest,
|
10 |
+
)
|
11 |
+
from src.api.database import get_db, Database, QueryExecutionError, HealthCheckError
|
12 |
+
from src.api.services.embedding_service import EmbeddingService
|
13 |
+
from src.api.services.huggingface_service import HuggingFaceService
|
14 |
+
from src.api.exceptions import DatasetNotFoundError, DatasetPushError, OpenAIError
|
15 |
+
import pandas as pd
|
16 |
+
import logging
|
17 |
+
from dotenv import load_dotenv
|
18 |
+
|
19 |
+
# Load environment variables
|
20 |
+
load_dotenv()
|
21 |
+
|
22 |
+
# Set up structured logging
|
23 |
+
logging.basicConfig(
|
24 |
+
level=logging.INFO, format="%(asctime)s - %(name)s - %(levelname)s - %(message)s"
|
25 |
+
)
|
26 |
+
logger = logging.getLogger(__name__)
|
27 |
+
|
28 |
+
# Initialize FastAPI app
|
29 |
+
app = FastAPI(
|
30 |
+
title="Similarity Search API",
|
31 |
+
description="A FastAPI application for similarity search with PostgreSQL and OpenAI embeddings.",
|
32 |
+
version="1.0.0",
|
33 |
+
)
|
34 |
+
|
35 |
+
|
36 |
+
# Root endpoint redirects to /docs
|
37 |
+
@app.get("/")
|
38 |
+
async def root():
|
39 |
+
return RedirectResponse(url="/docs")
|
40 |
+
|
41 |
+
|
42 |
+
# Health check endpoint
|
43 |
+
@app.get("/health")
|
44 |
+
async def health_check(db: Database = Depends(get_db)):
|
45 |
+
try:
|
46 |
+
is_healthy = await db.health_check()
|
47 |
+
if not is_healthy:
|
48 |
+
raise HTTPException(status_code=500, detail="Database is unhealthy")
|
49 |
+
return {"status": "healthy"}
|
50 |
+
except HealthCheckError as e:
|
51 |
+
raise HTTPException(status_code=500, detail=str(e))
|
52 |
+
|
53 |
+
|
54 |
+
# Dependency to get EmbeddingService
|
55 |
+
def get_embedding_service() -> EmbeddingService:
|
56 |
+
return EmbeddingService(openai_api_key=os.getenv("OPENAI_API_KEY"))
|
57 |
+
|
58 |
+
|
59 |
+
# Dependency to get HuggingFaceService
|
60 |
+
def get_huggingface_service() -> HuggingFaceService:
|
61 |
+
return HuggingFaceService()
|
62 |
+
|
63 |
+
|
64 |
+
# Endpoint to create embeddings
|
65 |
+
@app.post("/create_embedding")
|
66 |
+
async def create_embedding(
|
67 |
+
request: CreateEmbeddingRequest,
|
68 |
+
db: Database = Depends(get_db),
|
69 |
+
embedding_service: EmbeddingService = Depends(get_embedding_service),
|
70 |
+
huggingface_service: HuggingFaceService = Depends(get_huggingface_service),
|
71 |
+
):
|
72 |
+
"""
|
73 |
+
Create embeddings for the target column in the dataset.
|
74 |
+
"""
|
75 |
+
try:
|
76 |
+
# Step 1: Query the database
|
77 |
+
logger.info("Fetching data from the database...")
|
78 |
+
result = await db.fetch(request.query)
|
79 |
+
df = pd.DataFrame(result)
|
80 |
+
|
81 |
+
# Step 2: Generate embeddings
|
82 |
+
df = await embedding_service.create_embeddings(
|
83 |
+
df, request.target_column, request.output_column
|
84 |
+
)
|
85 |
+
|
86 |
+
# Step 3: Push to Hugging Face Hub
|
87 |
+
await huggingface_service.push_to_hub(df, request.dataset_name)
|
88 |
+
|
89 |
+
return JSONResponse(
|
90 |
+
content={
|
91 |
+
"message": "Embeddings created and pushed to Hugging Face Hub.",
|
92 |
+
"dataset_name": request.dataset_name,
|
93 |
+
"num_rows": len(df),
|
94 |
+
}
|
95 |
+
)
|
96 |
+
except QueryExecutionError as e:
|
97 |
+
logger.error(f"Database query failed: {e}")
|
98 |
+
raise HTTPException(status_code=500, detail=f"Database query failed: {e}")
|
99 |
+
except OpenAIError as e:
|
100 |
+
logger.error(f"OpenAI API error: {e}")
|
101 |
+
raise HTTPException(status_code=500, detail=f"OpenAI API error: {e}")
|
102 |
+
except DatasetPushError as e:
|
103 |
+
logger.error(f"Failed to push dataset: {e}")
|
104 |
+
raise HTTPException(status_code=500, detail=f"Failed to push dataset: {e}")
|
105 |
+
except Exception as e:
|
106 |
+
logger.error(f"An error occurred: {e}")
|
107 |
+
raise HTTPException(status_code=500, detail=f"An error occurred: {e}")
|
108 |
+
|
109 |
+
|
110 |
+
# Endpoint to read embeddings
|
111 |
+
@app.get("/read_embeddings/{dataset_name}")
|
112 |
+
async def read_embeddings(
|
113 |
+
dataset_name: str,
|
114 |
+
huggingface_service: HuggingFaceService = Depends(get_huggingface_service),
|
115 |
+
):
|
116 |
+
"""
|
117 |
+
Read embeddings from a Hugging Face dataset.
|
118 |
+
"""
|
119 |
+
try:
|
120 |
+
df = await huggingface_service.read_dataset(dataset_name)
|
121 |
+
return df.to_dict(orient="records")
|
122 |
+
except DatasetNotFoundError as e:
|
123 |
+
logger.error(f"Dataset not found: {e}")
|
124 |
+
raise HTTPException(status_code=404, detail=f"Dataset not found: {e}")
|
125 |
+
except Exception as e:
|
126 |
+
logger.error(f"An error occurred: {e}")
|
127 |
+
raise HTTPException(status_code=500, detail=f"An error occurred: {e}")
|
128 |
+
|
129 |
+
|
130 |
+
# Endpoint to update embeddings
|
131 |
+
@app.post("/update_embeddings")
|
132 |
+
async def update_embeddings(
|
133 |
+
request: UpdateEmbeddingRequest,
|
134 |
+
huggingface_service: HuggingFaceService = Depends(get_huggingface_service),
|
135 |
+
):
|
136 |
+
"""
|
137 |
+
Update embeddings in a Hugging Face dataset.
|
138 |
+
"""
|
139 |
+
try:
|
140 |
+
df = await huggingface_service.update_dataset(
|
141 |
+
request.dataset_name, request.updates
|
142 |
+
)
|
143 |
+
return {
|
144 |
+
"message": "Embeddings updated successfully.",
|
145 |
+
"dataset_name": request.dataset_name,
|
146 |
+
}
|
147 |
+
except DatasetPushError as e:
|
148 |
+
logger.error(f"Failed to update dataset: {e}")
|
149 |
+
raise HTTPException(status_code=500, detail=f"Failed to update dataset: {e}")
|
150 |
+
except Exception as e:
|
151 |
+
logger.error(f"An error occurred: {e}")
|
152 |
+
raise HTTPException(status_code=500, detail=f"An error occurred: {e}")
|
153 |
+
|
154 |
+
|
155 |
+
# Endpoint to delete embeddings
|
156 |
+
@app.post("/delete_embeddings")
|
157 |
+
async def delete_embeddings(
|
158 |
+
request: DeleteEmbeddingRequest,
|
159 |
+
huggingface_service: HuggingFaceService = Depends(get_huggingface_service),
|
160 |
+
):
|
161 |
+
"""
|
162 |
+
Delete embeddings from a Hugging Face dataset.
|
163 |
+
"""
|
164 |
+
try:
|
165 |
+
df = await huggingface_service.delete_columns(
|
166 |
+
request.dataset_name, request.columns
|
167 |
+
)
|
168 |
+
return {
|
169 |
+
"message": "Embeddings deleted successfully.",
|
170 |
+
"dataset_name": request.dataset_name,
|
171 |
+
}
|
172 |
+
except DatasetPushError as e:
|
173 |
+
logger.error(f"Failed to delete columns: {e}")
|
174 |
+
raise HTTPException(status_code=500, detail=f"Failed to delete columns: {e}")
|
175 |
+
except Exception as e:
|
176 |
+
logger.error(f"An error occurred: {e}")
|
177 |
+
raise HTTPException(status_code=500, detail=f"An error occurred: {e}")
|