222 lines
7.7 KiB
Python
222 lines
7.7 KiB
Python
"""
|
||
Book Fetcher Service
|
||
|
||
Fetch book information from various sources (API or LLM).
|
||
"""
|
||
|
||
import json
|
||
from typing import Optional, Literal
|
||
|
||
from loguru import logger
|
||
|
||
from reelforge.services.base import BaseService
|
||
|
||
|
||
class BookFetcherService(BaseService):
|
||
"""
|
||
Book information fetcher service
|
||
|
||
Provides unified access to various book data sources:
|
||
- API: Google Books, Douban, etc. (via configured capability)
|
||
- LLM: Generate book info using LLM (flexible, works for any book)
|
||
|
||
Usage:
|
||
# Use default source (from config, usually 'google')
|
||
book_info = await reelforge.book_fetcher("原则")
|
||
|
||
# Explicitly use Google Books API
|
||
book_info = await reelforge.book_fetcher("Atomic Habits", query_source="google")
|
||
|
||
# Explicitly use LLM (good for Chinese books)
|
||
book_info = await reelforge.book_fetcher("人性的弱点", query_source="llm")
|
||
|
||
# Use Douban (if you implemented it)
|
||
book_info = await reelforge.book_fetcher(
|
||
book_name="原则",
|
||
author="瑞·达利欧",
|
||
query_source="douban"
|
||
)
|
||
"""
|
||
|
||
def __init__(self, config_manager):
|
||
super().__init__(config_manager, "book_fetcher")
|
||
self._core = None # Will be set by ReelForgeCore (for LLM query)
|
||
|
||
def set_core(self, core):
|
||
"""Set reference to ReelForgeCore (for LLM query)"""
|
||
self._core = core
|
||
|
||
async def __call__(
|
||
self,
|
||
book_name: str,
|
||
author: Optional[str] = None,
|
||
query_source: Optional[Literal["google", "douban", "llm"]] = None,
|
||
**kwargs
|
||
) -> dict:
|
||
"""
|
||
Fetch book information
|
||
|
||
Args:
|
||
book_name: Book name (required)
|
||
author: Author name (optional, improves matching accuracy)
|
||
query_source: Data source to query:
|
||
- "google": Google Books API
|
||
- "douban": Douban Books (requires implementation)
|
||
- "llm": Generate book info using LLM
|
||
- None: Use default from config (usually "google")
|
||
**kwargs: Additional provider-specific parameters
|
||
|
||
Returns:
|
||
Book information dict with fields:
|
||
- title: Book title
|
||
- author: Author name
|
||
- summary: Book summary
|
||
- genre: Book category/genre
|
||
- publication_year: Publication year (string)
|
||
- key_points: List of key points (only from LLM)
|
||
- cover_url: Cover image URL (only from API)
|
||
- isbn: ISBN code (only from API)
|
||
- source: Data source ("google", "douban", or "llm")
|
||
|
||
Examples:
|
||
>>> # Use default source (from config)
|
||
>>> book = await reelforge.book_fetcher("Atomic Habits")
|
||
|
||
>>> # Explicitly use Google Books
|
||
>>> book = await reelforge.book_fetcher("Atomic Habits", query_source="google")
|
||
|
||
>>> # Explicitly use LLM (good for Chinese books)
|
||
>>> book = await reelforge.book_fetcher("人性的弱点", query_source="llm")
|
||
|
||
>>> # Use Douban (if implemented)
|
||
>>> book = await reelforge.book_fetcher(
|
||
... "原则",
|
||
... author="瑞·达利欧",
|
||
... query_source="douban"
|
||
... )
|
||
|
||
>>> print(f"Title: {book['title']}")
|
||
>>> print(f"Source: {book['source']}")
|
||
"""
|
||
# Route to appropriate method based on query_source
|
||
if query_source == "llm":
|
||
# Use LLM to generate book info
|
||
return await self._fetch_via_llm(book_name, author)
|
||
else:
|
||
# Use API (google, douban, or default from config)
|
||
return await self._fetch_via_api(book_name, author, query_source, **kwargs)
|
||
|
||
async def _fetch_via_api(
|
||
self,
|
||
book_name: str,
|
||
author: Optional[str] = None,
|
||
query_source: Optional[str] = None,
|
||
**kwargs
|
||
) -> dict:
|
||
"""
|
||
Fetch book information via API capability
|
||
|
||
Args:
|
||
book_name: Book name
|
||
author: Author name (optional)
|
||
query_source: Specific capability to use ("google", "douban", or None for default)
|
||
**kwargs: Additional parameters
|
||
|
||
Returns:
|
||
Book information dict
|
||
|
||
Raises:
|
||
Exception: If API call fails
|
||
"""
|
||
params = {"book_name": book_name}
|
||
if author is not None:
|
||
params["author"] = author
|
||
params.update(kwargs)
|
||
|
||
# Call book_fetcher capability
|
||
# If query_source is specified (e.g., "google"), use it
|
||
# Otherwise use default from config
|
||
result_json = await self._config_manager.call(
|
||
self._capability_type,
|
||
cap_id=query_source, # None = use default from config
|
||
**params
|
||
)
|
||
result = json.loads(result_json)
|
||
result["source"] = query_source or self.active or "api"
|
||
|
||
logger.info(f"✅ Fetched book info from {result['source']}: {result.get('title', book_name)}")
|
||
return result
|
||
|
||
async def _fetch_via_llm(self, book_name: str, author: Optional[str] = None) -> dict:
|
||
"""
|
||
Generate book information using LLM
|
||
|
||
This method uses LLM to generate book information based on its knowledge.
|
||
Good for books that are not available in API databases or for Chinese books.
|
||
|
||
Args:
|
||
book_name: Book name
|
||
author: Author name (optional)
|
||
|
||
Returns:
|
||
Book information dict
|
||
|
||
Raises:
|
||
ValueError: If LLM response cannot be parsed
|
||
Exception: If LLM call fails
|
||
"""
|
||
if not self._core:
|
||
raise RuntimeError("ReelForgeCore not set. Cannot use LLM query.")
|
||
|
||
# Build prompt
|
||
author_info = f"，作者是{author}" if author else ""
|
||
prompt = f"""请为书籍《{book_name}》{author_info}生成详细的书籍信息。
|
||
|
||
要求：
|
||
1. 如果你知道这本书，请提供真实准确的信息
|
||
2. 如果不确定，请基于书名和作者推测合理的信息
|
||
3. 严格按照JSON格式输出，不要添加任何其他内容
|
||
|
||
输出格式（JSON）：
|
||
{{
|
||
"title": "书名",
|
||
"author": "作者",
|
||
"summary": "书籍简介（100-200字，概括核心内容和价值）",
|
||
"genre": "书籍类型（如：自我成长、商业管理、心理学等）",
|
||
"publication_year": "2018",
|
||
"key_points": [
|
||
"核心观点1（20-30字）",
|
||
"核心观点2（20-30字）",
|
||
"核心观点3（20-30字）"
|
||
]
|
||
}}
|
||
|
||
只输出JSON，不要其他内容。"""
|
||
|
||
# Call LLM
|
||
response = await self._core.llm(
|
||
prompt=prompt,
|
||
temperature=0.3, # Lower temperature for more factual responses
|
||
max_tokens=1000
|
||
)
|
||
|
||
# Parse JSON
|
||
try:
|
||
book_info = json.loads(response)
|
||
except json.JSONDecodeError as e:
|
||
logger.error(f"Failed to parse LLM response as JSON: {e}")
|
||
logger.error(f"Response: {response[:200]}...")
|
||
raise ValueError(f"LLM returned invalid JSON for book: {book_name}")
|
||
|
||
# Ensure required fields exist
|
||
book_info.setdefault("title", book_name)
|
||
book_info.setdefault("author", author or "Unknown")
|
||
book_info.setdefault("summary", "No summary available")
|
||
book_info.setdefault("genre", "Unknown")
|
||
book_info.setdefault("publication_year", "")
|
||
book_info["source"] = "llm"
|
||
|
||
logger.info(f"✅ Generated book info via LLM: {book_info['title']}")
|
||
return book_info
|
||
|