assertEquals("http://example.com:8080/path/file", HcHttpClient.constructRedirectLocation("http://example.com:8080/path/", "file"));
        assertEquals("http://example.com/%E3%83%86%E3%82%B9%E3%83%88",
                HcHttpClient.constructRedirectLocation("http://example.com/", "テスト"));
        assertEquals("http://example.com/hello%20world", HcHttpClient.constructRedirectLocation("http://example.com/", "hello world"));

sitemap2.xml http://www.example.com/sitemap1.xml.gz 2004-10-01T18:23:17+00:00 http://www.example.com/sitemap2.xml.gz 2005-01-01...

        responseData.setUrl("http://example.com/sitemap.xml");
        File file = ResourceUtil.getResourceAsFile("sitemaps/sitemap1.xml");
        responseData.setResponseBody(file, false);
        return responseData;
    }

    private ResponseData getTestData2_OK() {
        final ResponseData responseData = new ResponseData();
        responseData.setUrl("http://example.com/sitemap.xml.gz");

        url = "http://test.com/hoge1.pdf";
        params.put(ExtractData.URL, url);
        assertEquals("password", pdfExtractor.getPassword(params));

        url = "http://test.com/hoge1.pdf";
        params.put(ExtractData.URL, url);
        assertEquals("password", pdfExtractor.getPassword(params));

        url = "http://test.com/fuga.pdf";
        params.put(ExtractData.URL, url);

https://twitter.com/Mint_Lee/status/626564933350600704

@withnewsjp
デビューは23年前、アイドル冬の時代から変わらぬスタイル

>制服向上委員会とは何者か？「自民倒そう」　民主時代は「野田ダメ」 - ウィズニュース http://t.co/qtVCWlrmy8 #withnews #制服向上委員会 http://t.co/6R1CL9LwfT
https://twitter.com/withnewsjp/status/626512633483522049

@lwnish
今宵 http://t.co/ca7DKqPWVU
https://twitter.com/lwnish/status/626695622242926593

@Newsweek_JAPAN

   <url>
      <loc>http://www.example.com/</loc>
      <lastmod>2005-01-01</lastmod>
      <changefreq>monthly</changefreq>
      <priority>0.8</priority>
   </url>
   <url>
      <loc>http://www.example.com/catalog?item=12&amp;desc=vacation_hawaii</loc>
      <changefreq>weekly</changefreq>
   </url>
   <url>
      <loc>http://www.example.com/catalog?item=73&amp;desc=vacation_new_zealand</loc>

        rule.addCondition("url", "https?://.*\\.example\\.com/.*");

        ResponseData responseData1 = new ResponseData();
        responseData1.setUrl("http://www.example.com/page");
        assertTrue(rule.match(responseData1));

        ResponseData responseData2 = new ResponseData();
        responseData2.setUrl("https://api.example.com/v1/users");
        assertTrue(rule.match(responseData2));

        accessResult1.setParentUrl("http://www.parent.com/");
        accessResult1.setRuleId("htmlRule");
        accessResult1.setSessionId("id1");
        accessResult1.setStatus(200);
        accessResult1.setUrl("http://www.id1.com/");

        dataService.store(accessResult1);

        final AccessResult accessResult2 = dataService.getAccessResult("id1", "http://www.id1.com/");
        assertNotNull(accessResult2);

    public void test_checkMaxContentLength() {
        AbstractCrawlerClient client = new AbstractCrawlerClient() {
        };
        ResponseData responseData = new ResponseData();
        responseData.setUrl("http://test.com/");

        client.checkMaxContentLength(responseData);

        responseData.setContentLength(-1);
        client.checkMaxContentLength(responseData);

        responseData.setContentLength(1000L);

        url = "http://www.example.com/";
        assertEquals("http_CLN_/www.example.com/index.html", fileTransformer.getFilePath(url));

        url = "http://www.example.com/action?a=1";
        assertEquals("http_CLN_/www.example.com/action_QUEST_a=1", fileTransformer.getFilePath(url));

        url = "http://www.example.com/action?a=1&b=2";