thordata-sdk 1.4.0__py3-none-any.whl → 1.6.0__py3-none-any.whl

thordata/_example_utils.py

@@ -1,77 +0,0 @@
-from __future__ import annotations
-
-import json
-import os
-from collections.abc import Iterable
-from pathlib import Path
-from typing import Any
-
-try:
-    from dotenv import load_dotenv
-except Exception:  # pragma: no cover
-    load_dotenv = None
-
-
-def load_env() -> None:
-    """Load .env from repo root if python-dotenv is installed."""
-    if load_dotenv is None:
-        return
-    repo_root = Path(__file__).resolve().parents[2]
-    load_dotenv(dotenv_path=repo_root / ".env")
-
-
-def env(name: str) -> str:
-    return (os.getenv(name) or "").strip()
-
-
-def skip_if_missing(required: Iterable[str], *, tip: str | None = None) -> bool:
-    missing = [k for k in required if not env(k)]
-    if not missing:
-        return False
-    print("Skipping live example: missing env:", ", ".join(missing))
-    if tip:
-        print(tip)
-    else:
-        print("Tip: copy .env.example to .env and fill values, then re-run.")
-    return True
-
-
-def parse_json_env(name: str, default: str = "{}") -> Any:
-    raw = env(name) or default
-    return json.loads(raw)
-
-
-def normalize_task_parameters(raw: Any) -> dict[str, Any]:
-    """Accept {..} or [{..}] and return a single dict for create_scraper_task(parameters=...)."""
-    if isinstance(raw, list):
-        if not raw:
-            raise ValueError("Task parameters JSON array must not be empty")
-        raw = raw[0]
-    if not isinstance(raw, dict):
-        raise ValueError("Task parameters must be a JSON object (or array of objects)")
-    return raw
-
-
-def output_dir() -> Path:
-    """Return output dir for examples; defaults to examples/output (ignored by git)."""
-    repo_root = Path(__file__).resolve().parents[2]
-    d = env("THORDATA_OUTPUT_DIR") or str(repo_root / "examples" / "output")
-    p = Path(d)
-    p.mkdir(parents=True, exist_ok=True)
-    return p
-
-
-def write_text(filename: str, content: str) -> Path:
-    p = output_dir() / filename
-    p.write_text(content, encoding="utf-8", errors="replace")
-    return p
-
-
-def write_json(filename: str, data: Any) -> Path:
-    p = output_dir() / filename
-    p.write_text(
-        json.dumps(data, ensure_ascii=False, indent=2),
-        encoding="utf-8",
-        errors="replace",
-    )
-    return p

thordata/demo.py

@@ -1,138 +0,0 @@
-"""
-Unified demo entrypoint for the Thordata Python SDK.
-
-This module runs the example scripts from the repository's `examples/` directory
-using `runpy`, so it does not require `examples/` to be an importable package.
-
-Usage:
-    python -m thordata.demo serp
-    python -m thordata.demo universal
-    python -m thordata.demo scraper
-    python -m thordata.demo concurrency
-
-Notes:
-- This entrypoint is primarily intended for repository usage (dev/demo).
-- When installed from PyPI, the `examples/` directory is typically not included.
-"""
-
-from __future__ import annotations
-
-import runpy
-import sys
-from pathlib import Path
-
-
-def _configure_stdio() -> None:
-    # Avoid UnicodeEncodeError on Windows consoles with legacy encodings.
-    if hasattr(sys.stdout, "reconfigure"):
-        sys.stdout.reconfigure(encoding="utf-8", errors="replace")  # type: ignore[attr-defined]
-    if hasattr(sys.stderr, "reconfigure"):
-        sys.stderr.reconfigure(encoding="utf-8", errors="replace")  # type: ignore[attr-defined]
-
-
-def _load_env() -> None:
-    # Optional .env support for local development
-    try:
-        from dotenv import load_dotenv
-    except ImportError:
-        return
-    load_dotenv()
-
-
-def _repo_root() -> Path:
-    """
-    Resolve repository root based on src layout:
-    <repo>/src/thordata/demo.py -> parents[2] == <repo>
-    """
-    return Path(__file__).resolve().parents[2]
-
-
-def _examples_dir() -> Path:
-    return _repo_root() / "examples"
-
-
-def _demo_map() -> dict[str, Path]:
-    ex = _examples_dir()
-    return {
-        "serp": ex / "demo_serp_api.py",
-        "universal": ex / "demo_universal.py",
-        "scraper": ex / "demo_web_scraper_api.py",
-        "concurrency": ex / "async_high_concurrency.py",
-    }
-
-
-def _usage() -> str:
-    names = ", ".join(sorted(_demo_map().keys()))
-    return f"Usage: python -m thordata.demo [{names}]"
-
-
-def _run_demo(path: Path) -> int:
-    if not path.exists():
-        print(f"Error: demo script not found: {path}")
-        return 2
-
-    # Ensure examples dir is on sys.path (helpful if demo imports local helpers).
-    examples_dir = str(path.parent.resolve())
-    if examples_dir not in sys.path:
-        sys.path.insert(0, examples_dir)
-
-    try:
-        # Load without triggering `if __name__ == "__main__": ...`
-        ns = runpy.run_path(str(path), run_name="__thordata_demo__")
-
-        main_func = ns.get("main")
-        if callable(main_func):
-            return int(main_func())  # type: ignore[arg-type]
-
-        # Fallback: run as __main__ for scripts without main()
-        runpy.run_path(str(path), run_name="__main__")
-        return 0
-
-    except KeyboardInterrupt:
-        raise
-    except SystemExit as e:
-        # In case fallback run as __main__ triggered SystemExit
-        code = e.code
-        if code is None:
-            return 0
-        if isinstance(code, int):
-            return code
-        return 1
-    except Exception as e:
-        import traceback
-
-        print()
-        print("-" * 60)
-        print("[thordata.demo] The demo script raised an exception.")
-        print(f"[thordata.demo] Script: {path.name}")
-        print(f"[thordata.demo] Error:  {type(e).__name__}: {e}")
-        print()
-        print("Note: This is a failure within the demo script itself,")
-        print("      not an issue with the thordata.demo entrypoint.")
-        print("-" * 60)
-        traceback.print_exc()
-        return 1
-
-
-def main() -> int:
-    _configure_stdio()
-    _load_env()
-
-    if len(sys.argv) < 2:
-        print(_usage())
-        return 2
-
-    name = sys.argv[1].strip().lower()
-    mapping = _demo_map()
-
-    path = mapping.get(name)
-    if path is None:
-        print(f"Unknown demo: {name}")
-        print(_usage())
-        return 2
-
-    return _run_demo(path)
-
-
-if __name__ == "__main__":
-    raise SystemExit(main())

thordata_sdk-1.4.0.dist-info/METADATA

@@ -1,208 +0,0 @@
-Metadata-Version: 2.4
-Name: thordata-sdk
-Version: 1.4.0
-Summary: The Official Python SDK for Thordata - AI Data Infrastructure & Proxy Network.
-Author-email: Thordata Developer Team <support@thordata.com>
-License: MIT
-Project-URL: Homepage, https://www.thordata.com
-Project-URL: Documentation, https://github.com/Thordata/thordata-python-sdk#readme
-Project-URL: Source, https://github.com/Thordata/thordata-python-sdk
-Project-URL: Tracker, https://github.com/Thordata/thordata-python-sdk/issues
-Project-URL: Changelog, https://github.com/Thordata/thordata-python-sdk/blob/main/CHANGELOG.md
-Keywords: web scraping,proxy,residential proxy,datacenter proxy,ai,llm,data-mining,serp,thordata,web scraper,anti-bot bypass
-Classifier: Development Status :: 4 - Beta
-Classifier: Intended Audience :: Developers
-Classifier: Topic :: Software Development :: Libraries :: Python Modules
-Classifier: Topic :: Internet :: WWW/HTTP
-Classifier: Topic :: Internet :: Proxy Servers
-Classifier: Programming Language :: Python :: 3
-Classifier: Programming Language :: Python :: 3.9
-Classifier: Programming Language :: Python :: 3.10
-Classifier: Programming Language :: Python :: 3.11
-Classifier: Programming Language :: Python :: 3.12
-Classifier: License :: OSI Approved :: MIT License
-Classifier: Operating System :: OS Independent
-Classifier: Typing :: Typed
-Requires-Python: >=3.9
-Description-Content-Type: text/markdown
-License-File: LICENSE
-Requires-Dist: requests>=2.25.0
-Requires-Dist: aiohttp>=3.9.0
-Requires-Dist: PySocks>=1.7.1
-Provides-Extra: dev
-Requires-Dist: pytest>=7.0.0; extra == "dev"
-Requires-Dist: pytest-asyncio>=0.21.0; extra == "dev"
-Requires-Dist: pytest-cov>=4.0.0; extra == "dev"
-Requires-Dist: pytest-httpserver>=1.0.0; extra == "dev"
-Requires-Dist: python-dotenv>=1.0.0; extra == "dev"
-Requires-Dist: black>=25.11.0; extra == "dev"
-Requires-Dist: ruff>=0.1.0; extra == "dev"
-Requires-Dist: mypy>=1.0.0; extra == "dev"
-Requires-Dist: types-requests>=2.28.0; extra == "dev"
-Requires-Dist: aioresponses>=0.7.6; extra == "dev"
-Dynamic: license-file
-
-# Thordata Python SDK
-
-<div align="center">
-
-<img src="https://img.shields.io/badge/Thordata-AI%20Infrastructure-blue?style=for-the-badge" alt="Thordata Logo">
-
-**The Official Python Client for Thordata APIs**
-
-*Proxy Network • SERP API • Web Unlocker • Web Scraper API*
-
-[![PyPI version](https://img.shields.io/pypi/v/thordata-sdk.svg?style=flat-square)](https://pypi.org/project/thordata-sdk/)
-[![Python Versions](https://img.shields.io/pypi/pyversions/thordata-sdk.svg?style=flat-square)](https://pypi.org/project/thordata-sdk/)
-[![License](https://img.shields.io/badge/license-MIT-green?style=flat-square)](LICENSE)
-[![CI Status](https://img.shields.io/github/actions/workflow/status/Thordata/thordata-python-sdk/ci.yml?branch=main&style=flat-square)](https://github.com/Thordata/thordata-python-sdk/actions)
-
-</div>
-
----
-
-## 📖 Introduction
-
-This SDK provides a robust, high-performance interface to Thordata's AI data infrastructure. It is designed for high-concurrency scraping, reliable proxy tunneling, and seamless data extraction.
-
-**Key Features:**
-*   **🚀 Production Ready:** Built on `urllib3` connection pooling for low-latency proxy requests.
-*   **⚡ Async Support:** Native `aiohttp` client for high-concurrency SERP/Universal scraping.
-*   **🛡️ Robust:** Handles TLS-in-TLS tunneling, retries, and error parsing automatically.
-*   **✨ Developer Experience:** Fully typed (`mypy` compatible) with intuitive IDE autocomplete.
-*   **🧩 Lazy Validation:** Only validate credentials for the features you actually use.
-
----
-
-## 📦 Installation
-
-```bash
-pip install thordata-sdk
-```
-
----
-
-## 🔐 Configuration
-
-Set environment variables to avoid hardcoding credentials. You only need to set the variables for the features you use.
-
-```bash
-# [Required for SERP & Web Unlocker]
-export THORDATA_SCRAPER_TOKEN="your_token_here"
-
-# [Required for Proxy Network]
-export THORDATA_RESIDENTIAL_USERNAME="your_username"
-export THORDATA_RESIDENTIAL_PASSWORD="your_password"
-export THORDATA_PROXY_HOST="vpnXXXX.pr.thordata.net"
-
-# [Required for Task Management]
-export THORDATA_PUBLIC_TOKEN="public_token"
-export THORDATA_PUBLIC_KEY="public_key"
-```
-
----
-
-## 🚀 Quick Start
-
-### 1. SERP Search (Google/Bing/Yandex)
-
-```python
-from thordata import ThordataClient, Engine
-
-client = ThordataClient()  # Loads THORDATA_SCRAPER_TOKEN from env
-
-# Simple Search
-print("Searching...")
-results = client.serp_search("latest AI trends", engine=Engine.GOOGLE_NEWS)
-
-for news in results.get("news_results", [])[:3]:
-    print(f"- {news['title']} ({news['source']})")
-```
-
-### 2. Universal Scrape (Web Unlocker)
-
-Bypass Cloudflare/Akamai and render JavaScript automatically.
-
-```python
-html = client.universal_scrape(
-    url="https://example.com/protected-page",
-    js_render=True,
-    wait_for=".content-loaded",
-    country="us"
-)
-print(f"Scraped {len(html)} bytes")
-```
-
-### 3. High-Performance Proxy
-
-Use Thordata's residential IPs with automatic connection pooling.
-
-```python
-from thordata import ProxyConfig, ProxyProduct
-
-# Config is optional if env vars are set, but allows granular control
-proxy = ProxyConfig(
-    product=ProxyProduct.RESIDENTIAL,
-    country="jp",
-    city="tokyo",
-    session_id="session-001",
-    session_duration=10  # Sticky IP for 10 mins
-)
-
-# Use the client to make requests (Reuses TCP connections)
-response = client.get("https://httpbin.org/ip", proxy_config=proxy)
-print(response.json())
-```
-
----
-
-## ⚙️ Advanced Usage
-
-### Async Client (High Concurrency)
-
-For building AI agents or high-throughput spiders.
-
-```python
-import asyncio
-from thordata import AsyncThordataClient
-
-async def main():
-    async with AsyncThordataClient() as client:
-        # Fire off multiple requests in parallel
-        tasks = [
-            client.serp_search(f"query {i}") 
-            for i in range(5)
-        ]
-        results = await asyncio.gather(*tasks)
-        print(f"Completed {len(results)} searches")
-
-asyncio.run(main())
-```
-
-### Web Scraper API (Task Management)
-
-Create and manage large-scale scraping tasks asynchronously.
-
-```python
-# 1. Create a task
-task_id = client.create_scraper_task(
-    file_name="daily_scrape",
-    spider_id="universal",
-    spider_name="universal",
-    parameters={"url": "https://example.com"}
-)
-
-# 2. Wait for completion (Polling)
-status = client.wait_for_task(task_id)
-
-# 3. Get results
-if status == "ready":
-    url = client.get_task_result(task_id)
-    print(f"Download Data: {url}")
-```
-
----
-
-## 📄 License
-
-MIT License. See [LICENSE](LICENSE) for details.

thordata_sdk-1.4.0.dist-info/RECORD

@@ -1,18 +0,0 @@
-thordata/__init__.py,sha256=WVs3pZALETqLZDoAWyUoYNEjTAtC9FPvBxJEvDHnu04,3195
-thordata/_example_utils.py,sha256=T9QtVq9BHhubOShgtGp2GSusYYd-ZFUJFJAw7ubIsa4,2199
-thordata/_utils.py,sha256=Acr_6sHgdZXU7SQozd6FEYTZV6iHw__nlhpBTDwb66U,4917
-thordata/async_client.py,sha256=U6VgjsaX3-dzaIwEHtPborwYVrw-K1flTLg9LP9Xozk,80378
-thordata/async_unlimited.py,sha256=kzTksFkN21rDM21Pwy3hcayjfyGYNGGyGR3fRLtZC6I,4510
-thordata/client.py,sha256=79XzN4VxyY9EgYSYi5FhicHzo7ASGLb4hoEWig0nT7w,84500
-thordata/demo.py,sha256=DojJRFqUm9XAMBkjmk03WGeiUdLCbXguMIwtMOzfN6M,3822
-thordata/enums.py,sha256=MpZnS9_8sg2vtcFqM6UicB94cKZm5R1t83L3ejNSbLs,8502
-thordata/exceptions.py,sha256=P9czrxkFhT439DxW3LE5W-koS595ObH4-mAQOfaDM18,9976
-thordata/models.py,sha256=qtB7jE0v5zNEQfSpmOqdiacB5DgM2QfVR2PaYs-DisM,38206
-thordata/retry.py,sha256=5kRwULl3X68Nx8PlSzr9benfyCL0nRSpVQXrwjWr45M,11456
-thordata/serp_engines.py,sha256=iuMWncelcGOskCHXFzpcPMMTL5qfiLkazHB1uj3zpZo,5985
-thordata/unlimited.py,sha256=U0MEHeowNe6js8aezPFD8vGOft0VA8cpL-TUsLNyv2k,3855
-thordata_sdk-1.4.0.dist-info/licenses/LICENSE,sha256=bAxpWgQIzb-5jl3nhLdOwOJ_vlbHLtSG7yev2B7vioY,1088
-thordata_sdk-1.4.0.dist-info/METADATA,sha256=RJvTjs-N-ZaI7X9KfBI1ST5lcNTCeNpTZ92Cz7Ky9wk,6601
-thordata_sdk-1.4.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-thordata_sdk-1.4.0.dist-info/top_level.txt,sha256=Z8R_07m0lXCCSb1hapL9_nxMtyO3rf_9wOvq4n9u2Hg,9
-thordata_sdk-1.4.0.dist-info/RECORD,,