Package net.yacy.cora.protocol

Examples of net.yacy.cora.protocol.RequestHeader

58
59
60
61
62
63
64
65
66
67
68
69
70
71
   
    public Response load(final Request request, boolean acceptOnlyParseable) throws IOException {
        DigestURI url = request.url();
        if (!url.getProtocol().equals("file")) throw new IOException("wrong loader for FileLoader: " + url.getProtocol());

        RequestHeader requestHeader = new RequestHeader();
        if (request.referrerhash() != null) {
            DigestURI ur = sb.getURL(Segments.Process.LOCALCRAWLING, request.referrerhash());
            if (ur != null) requestHeader.put(RequestHeader.REFERER, ur.toNormalform(true, false));
        }
       
        // process directories: transform them to html with meta robots=noindex (using the ftpc lib)
        String[] l = null;
        try {l = url.list();} catch (IOException e) {}
View Full Code Here
71
72
73
74
75
76
77
78
79
80
81
82
83
84
   
    public Response load(final Request request, boolean acceptOnlyParseable) throws IOException {
        DigestURI url = request.url();
        if (!url.getProtocol().equals("smb")) throw new IOException("wrong loader for SMBLoader: " + url.getProtocol());

        RequestHeader requestHeader = new RequestHeader();
        if (request.referrerhash() != null) {
            DigestURI ur = sb.getURL(Segments.Process.LOCALCRAWLING, request.referrerhash());
            if (ur != null) requestHeader.put(RequestHeader.REFERER, ur.toNormalform(true, false));
        }
       
        // process directories: transform them to html with meta robots=noindex (using the ftpc lib)
        String[] l = null;
        try {l = url.list();} catch (IOException e) {}
View Full Code Here
111
112
113
114
115
116
117
118
119
120
121
122
123
124
                }
            }

            if (file.length() == 0) {
                // directory -> get list of files
                final RequestHeader requestHeader = new RequestHeader();
                if (request.referrerhash() != null) {
                    final DigestURI u = this.sb.getURL(Segments.Process.LOCALCRAWLING, request.referrerhash());
                    if (u != null) requestHeader.put(RequestHeader.REFERER, u.toNormalform(true, false));
                }

                final StringBuilder dirList = ftpClient.dirhtml(path);

                if (dirList == null) {
View Full Code Here
218
219
220
221
222
223
224
225
226
227
228
229
230
231
        // determine the file date
        final Date fileDate = ftpClient.entryDate(path);

        // create response header
        final RequestHeader requestHeader = new RequestHeader();
        if (request.referrerhash() != null) {
            final DigestURI refurl = this.sb.getURL(Segments.Process.LOCALCRAWLING, request.referrerhash());
            if (refurl != null) requestHeader.put(RequestHeader.REFERER, refurl.toNormalform(true, false));
        }
        final ResponseHeader responseHeader = new ResponseHeader();
        responseHeader.put(HeaderFramework.LAST_MODIFIED, HeaderFramework.formatRFC1123(fileDate));
        responseHeader.put(HeaderFramework.CONTENT_TYPE, mime);
View Full Code Here
567
568
569
570
571
572
573
574
575
576
577
578
579
580
      if (uri.startsWith("http://") || uri.startsWith("https://")) {
            final String[] uris = uri.split(",");
            for (String netdef: uris) {
                netdef = netdef.trim();
                try {
                    final RequestHeader reqHeader = new RequestHeader();
                    reqHeader.put(HeaderFramework.USER_AGENT, ClientIdentification.getUserAgent());
                    final HTTPClient client = new HTTPClient();
                    client.setHeader(reqHeader.entrySet());
                    byte[] data = client.GETbytes(uri);
                    if (data == null || data.length == 0) continue;
                    // save locally in case next fetch fails
                    if (file != null) {
                      FileOutputStream f = new FileOutputStream(file);
View Full Code Here
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
        String eTag=null, oldEtag = null;
        Date lastMod=null;
        downloadStart = System.currentTimeMillis();

        // if we previously have downloaded this robots.txt then we can set the if-modified-since header
        RequestHeader reqHeaders = new RequestHeader();

        // add yacybot user agent
        reqHeaders.put(HeaderFramework.USER_AGENT, ClientIdentification.getUserAgent());

        // adding referer
        reqHeaders.put(RequestHeader.REFERER, (MultiProtocolURI.newURL(robotsURL,"/")).toNormalform(true, true));
        reqHeaders.put(RequestHeader.ACCEPT, HTTPLoader.DEFAULT_ACCEPT);
        if (entry != null) {
            oldEtag = entry.getETag();
            reqHeaders = new RequestHeader();
            final Date modDate = entry.getModDate();
            if (modDate != null) reqHeaders.put(RequestHeader.IF_MODIFIED_SINCE, HeaderFramework.formatRFC1123(entry.getModDate()));

        }

        // setup http-client
        //TODO: adding Traffic statistic for robots download?
        final HTTPClient client = new HTTPClient();
        client.setHeader(reqHeaders.entrySet());
        try {
            // check for interruption
            if (Thread.currentThread().isInterrupted()) throw new InterruptedException("Shutdown in progress.");

            // sending the get request
View Full Code Here
162
163
164
165
166
167
168
169
170
171
172
    }

    public Response(final Request request, final CrawlProfile profile) {
        this.request = request;
        // request and response headers may be zero in case that we process surrogates
        this.requestHeader = new RequestHeader();
        this.responseHeader = new ResponseHeader();
        if (request.size() > 0) this.responseHeader.put(HeaderFramework.CONTENT_LENGTH, Long.toString(request.size()));
        this.responseStatus = "200";
        this.profile = profile;
        this.status = QUEUE_STATE_FRESH;
View Full Code Here
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
        // take a file from the net
        Response response = null;

        // create a request header
        final RequestHeader requestHeader = new RequestHeader();
        requestHeader.put(HeaderFramework.USER_AGENT, ClientIdentification.getUserAgent());
        DigestURI refererURL = null;
        if (request.referrerhash() != null) refererURL = this.sb.getURL(Segments.Process.LOCALCRAWLING, request.referrerhash());
        if (refererURL != null) requestHeader.put(RequestHeader.REFERER, refererURL.toNormalform(true, true));
        requestHeader.put(HeaderFramework.ACCEPT, this.sb.getConfig("crawler.http.accept", DEFAULT_ACCEPT));
        requestHeader.put(HeaderFramework.ACCEPT_LANGUAGE, this.sb.getConfig("crawler.http.acceptLanguage", DEFAULT_LANGUAGE));
        requestHeader.put(HeaderFramework.ACCEPT_CHARSET, this.sb.getConfig("crawler.http.acceptCharset", DEFAULT_CHARSET));
        requestHeader.put(HeaderFramework.ACCEPT_ENCODING, this.sb.getConfig("crawler.http.acceptEncoding", DEFAULT_ENCODING));

        // HTTP-Client
        final HTTPClient client = new HTTPClient();
        client.setRedirecting(false); // we want to handle redirection ourselves, so we don't index pages twice
        client.setTimout(this.socketTimeout);
        client.setHeader(requestHeader.entrySet());
            // send request
          final byte[] responseBody = client.GETbytes(url, maxFileSize);
          final ResponseHeader header = new ResponseHeader(client.getHttpResponse().getAllHeaders());
          final int code = client.getHttpResponse().getStatusLine().getStatusCode();
View Full Code Here
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
        // take a file from the net
        Response response = null;

        // create a request header
        final RequestHeader requestHeader = new RequestHeader();
        requestHeader.put(HeaderFramework.USER_AGENT, ClientIdentification.getUserAgent());
        requestHeader.put(HeaderFramework.ACCEPT_LANGUAGE, DEFAULT_LANGUAGE);
        requestHeader.put(HeaderFramework.ACCEPT_CHARSET, DEFAULT_CHARSET);
        requestHeader.put(HeaderFramework.ACCEPT_ENCODING, DEFAULT_ENCODING);

        final HTTPClient client = new HTTPClient();
        client.setTimout(20000);
        client.setHeader(requestHeader.entrySet());
          final byte[] responseBody = client.GETbytes(request.url());
          final ResponseHeader header = new ResponseHeader(client.getHttpResponse().getAllHeaders());
          final int code = client.getHttpResponse().getStatusLine().getStatusCode();
            // FIXME: 30*-handling (bottom) is never reached
            // we always get the final content because httpClient.followRedirects = true
View Full Code Here
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
     */
    public File downloadRelease() {
        final File storagePath = Switchboard.getSwitchboard().releasePath;
        File download = null;
        // setup httpClient
        final RequestHeader reqHeader = new RequestHeader();
        reqHeader.put(HeaderFramework.USER_AGENT, ClientIdentification.getUserAgent());

        final String name = getUrl().getFileName();
        byte[] signatureBytes = null;

        final HTTPClient client = new HTTPClient();
        client.setTimout(6000);
        client.setHeader(reqHeader.entrySet());

        // download signature first, if public key is available
        try {
            if (this.publicKey != null) {
              final byte[] signatureData = client.GETbytes(getUrl().toString() + ".sig");
View Full Code Here
TOP

Related Classes of net.yacy.cora.protocol.RequestHeader

  • de.anomic.crawler.retrieval.FileLoader
  • de.anomic.crawler.retrieval.FTPLoader
  • de.anomic.crawler.retrieval.HTTPLoader
  • de.anomic.crawler.retrieval.Response
  • de.anomic.crawler.retrieval.SMBLoader
  • de.anomic.crawler.RobotsTxt
  • de.anomic.http.server.HTTPDemon
  • de.anomic.http.server.ServerSideIncludes
  • de.anomic.search.Switchboard
  • de.anomic.server.serverSwitch
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.