kb-skills/tests/test_crawler/test_proxy_config.py

"""Tests for Crawler Proxy Configuration"""

import os
import pytest
from unittest.mock import patch

from kb_workflow.crawler.config import (
    CrawlerConfig,
    ProxyConfig,
    get_crawler_config,
    reset_crawler_config,
)


class TestProxyConfig:
    """Tests for ProxyConfig model."""

    def test_parse_proxy_from_https_env(self):
        """Parse proxy from HTTPS_PROXY environment variable."""
        reset_crawler_config()
        with patch.dict(os.environ, {"HTTPS_PROXY": "http://127.0.0.1:7890"}, clear=False):
            os.environ.pop("HTTP_PROXY", None)
            config = CrawlerConfig.from_env()
            assert config.proxy is not None
            assert config.proxy.server == "http://127.0.0.1:7890"

    def test_parse_proxy_from_http_env(self):
        """Parse proxy from HTTP_PROXY (HTTPS_PROXY takes priority)."""
        reset_crawler_config()
        with patch.dict(os.environ, {
            "HTTP_PROXY": "http://127.0.0.1:8080",
            "HTTPS_PROXY": "http://127.0.0.1:7890"
        }, clear=False):
            config = CrawlerConfig.from_env()
            assert config.proxy.server == "http://127.0.0.1:7890"

    def test_parse_proxy_with_auth(self):
        """Parse proxy with authentication credentials."""
        reset_crawler_config()
        with patch.dict(os.environ, {
            "HTTPS_PROXY": "http://user:pass@proxy.example.com:8080"
        }, clear=False):
            os.environ.pop("HTTP_PROXY", None)
            config = CrawlerConfig.from_env()
            assert config.proxy.username == "user"
            assert config.proxy.password == "pass"
            assert config.proxy.server == "http://proxy.example.com:8080"

    def test_no_proxy_configured(self):
        """Handle case when no proxy is configured."""
        reset_crawler_config()
        # Save current values
        old_http = os.environ.get("HTTP_PROXY")
        old_https = os.environ.get("HTTPS_PROXY")

        try:
            os.environ.pop("HTTP_PROXY", None)
            os.environ.pop("HTTPS_PROXY", None)
            config = CrawlerConfig.from_env()
            assert config.proxy is None
        finally:
            # Restore
            if old_http:
                os.environ["HTTP_PROXY"] = old_http
            if old_https:
                os.environ["HTTPS_PROXY"] = old_https

    def test_proxy_required_validation_fails(self):
        """Raise ValueError when proxy required but not configured."""
        config = CrawlerConfig(proxy=None, proxy_required=True)
        with pytest.raises(ValueError, match="Proxy configuration missing"):
            config.validate_proxy()

    def test_proxy_required_validation_passes(self):
        """Pass validation when proxy is configured and required."""
        config = CrawlerConfig(
            proxy=ProxyConfig(server="http://127.0.0.1:7890"),
            proxy_required=True
        )
        # Should not raise exception
        config.validate_proxy()

    def test_proxy_not_required(self):
        """Pass validation when proxy is not required."""
        config = CrawlerConfig(proxy=None, proxy_required=False)
        # Should not raise exception
        config.validate_proxy()

    def test_to_playwright_format(self):
        """Convert to Playwright proxy format."""
        proxy = ProxyConfig(
            server="http://127.0.0.1:7890",
            username="user",
            password="pass"
        )
        result = proxy.to_playwright_format()
        assert result == {
            "server": "http://127.0.0.1:7890",
            "username": "user",
            "password": "pass"
        }

    def test_to_playwright_format_no_auth(self):
        """Convert to Playwright format without authentication."""
        proxy = ProxyConfig(server="http://127.0.0.1:7890")
        result = proxy.to_playwright_format()
        assert result == {"server": "http://127.0.0.1:7890"}

    def test_get_crawler_config_singleton(self):
        """Test that get_crawler_config returns cached instance."""
        reset_crawler_config()
        with patch.dict(os.environ, {"HTTPS_PROXY": "http://127.0.0.1:7890"}, clear=False):
            os.environ.pop("HTTP_PROXY", None)
            config1 = get_crawler_config()
            config2 = get_crawler_config()
            assert config1 is config2

    def test_default_values(self):
        """Test default configuration values."""
        config = CrawlerConfig(proxy=None, proxy_required=False)
        assert config.min_delay == 1.0
        assert config.max_delay == 3.0
        assert config.max_retries == 3
        assert config.retry_min_wait == 2.0
        assert config.retry_max_wait == 10.0