Metadata-Version: 2.4
Name: selectorllm
Version: 0.0.2
Summary: Add your description here
License-File: LICENSE
Keywords: html,litellm,llm,selector,web-scraping
Requires-Python: >=3.12
Requires-Dist: litellm>=1.78.7
Provides-Extra: agentic
Requires-Dist: beautifulsoup4>=4.14.2; extra == 'agentic'
Requires-Dist: bs4-token-ext>=0.0.2; extra == 'agentic'
Requires-Dist: langgraph>=1.0.1; extra == 'agentic'
Provides-Extra: all
Requires-Dist: beautifulsoup4>=4.14.2; extra == 'all'
Requires-Dist: bs4-token-ext>=0.0.2; extra == 'all'
Requires-Dist: langgraph>=1.0.1; extra == 'all'
Description-Content-Type: text/markdown

![Selector llm logo](https://raw.githubusercontent.com/youseiushida/selectorllm/refs/heads/main/images/logo-blue.png)
# selectorllm
Generate CSS/XPath selectors from natural language using LLMs.  Extract HTML data without writing selectors manually.

---

自然言語でHTMLセレクタを生成 - LLMによる自動抽出
## Usage
```python
from selectorllm.main import get_selector

html = """<html>...</html>"""
query = "Extract link texts."

selector = get_selector(
    query,
    html,
    model="openrouter/openai/gpt-5", # Any LiteLLM-compatible model name
    selector_type="css"               # "css" or "xpath"
)

print(selector)  # example: "body > div > a"
```