Package cc.twittertools.corpus.data

Examples of cc.twittertools.corpus.data.Status

102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
    config.setOpenMode(IndexWriterConfig.OpenMode.CREATE); // Overwrite existing.

    IndexWriter writer = new IndexWriter(FSDirectory.open(indexLocation), config);

    int cnt = 0;
    Status status;
    try {
      while ((status = stream.next()) != null) {
        if (status.getText() == null) {
          continue;
        }

        cnt++;
        String createdAt = status.getCreatedAt();
        Document doc = new Document();
        doc.add(new Field(StatusField.ID.name, status.getId() + "",
            Store.YES, Index.NOT_ANALYZED_NO_NORMS));
        doc.add(new Field(StatusField.SCREEN_NAME.name, status.getScreenname(),
            Store.YES, Index.NOT_ANALYZED_NO_NORMS));
        doc.add(new Field(StatusField.CREATED_AT.name, createdAt, Store.YES, Index.NO));
        doc.add(new Field(StatusField.TEXT.name, status.getText(), Store.YES, Index.ANALYZED));

        String[] arr = createdAt.split(" ");
        String createDay = new StringBuffer().append(arr[1]).append("_").append(arr[2]).toString();
        doc.add(new Field(StatusField.DAY.name, createDay, Store.YES, Index.NOT_ANALYZED_NO_NORMS));
View Full Code Here
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
    } else {
      stream = new JsonStatusBlockReader(file);
    }

    int cnt = 0;
    Status status;
    while ((status = stream.next()) != null) {
      if (cmdline.hasOption(DUMP_OPTION)) {
        String text = status.getText();
        if (text != null) {
          text = text.replaceAll("\\s+", " ");
          text = text.replaceAll("\0", "");
        }
        out.println(String.format("%d\t%s\t%s\t%s", status.getId(), status.getScreenname(),
            status.getCreatedAt(), text));
      }
      cnt++;
      if ( cnt % 10000 == 0 && cmdline.hasOption(VERBOSE_OPTION)) {
        LOG.info(cnt + " statuses read");
      }
View Full Code Here
94
95
96
97
98
99
100
101
102
103
104
105
106
107
      System.exit(-1);
    }

    PrintStream out = new PrintStream(System.out, true, "UTF-8");
    StatusStream stream = new JsonStatusCorpusReader(file);
    Status status;
    while ((status = stream.next()) != null) {
      if (tweetids.contains(status.getId())) {
        out.println(status.getJsonObject().toString());
      }
    }
    stream.close();
    out.close();
  }
View Full Code Here
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
    LongOpenHashSet seen = new LongOpenHashSet();
    TreeMap<Long, String> tweets = Maps.newTreeMap();

    PrintStream out = new PrintStream(System.out, true, "UTF-8");
    StatusStream stream = new JsonStatusCorpusReader(file);
    Status status;
    int cnt = 0;
    while ((status = stream.next()) != null) {
      if (!tweetids.contains(status.getId())) {
        LOG.error("tweetid " + status.getId() + " doesn't belong in collection");
        continue;
      }
      if (seen.contains(status.getId())) {
        LOG.error("tweetid " + status.getId() + " already seen!");
        continue;
      }

      tweets.put(status.getId(), status.getJsonObject().toString());
      seen.add(status.getId());
      cnt++;
    }
    LOG.info("total of " + cnt + " tweets in subcollection.");

    for ( Map.Entry<Long, String> entry : tweets.entrySet()){
View Full Code Here
39
40
41
42
43
44
45
46
47
48
49
50
    String json = html.substring(jsonStart + AsyncEmbeddedJsonStatusBlockCrawler.JSON_START.length(), jsonEnd);
    json = StringEscapeUtils.unescapeHtml(json);
    JsonObject page = (JsonObject) JSON_PARSER.parse(json);
    JsonObject statusJson = page.getAsJsonObject("embedData").getAsJsonObject("status");

    Status status = Status.fromJson(statusJson.toString());
    assertEquals(1121915133L, status.getId());
    assertEquals("jkrums", status.getScreenname());
    assertEquals("http://twitpic.com/135xa - There's a plane in the Hudson. I'm on the ferry going to pick up the people. Crazy.", status.getText());

    asyncHttpClient.close();
  }
View Full Code Here
53
54
55
56
57
58
59
60
61
62
63
64
    String json = html.substring(jsonStart + AsyncEmbeddedJsonStatusBlockCrawler.JSON_START.length(), jsonEnd);
    json = StringEscapeUtils.unescapeHtml(json);
    JsonObject page = (JsonObject) JSON_PARSER.parse(json);
    JsonObject statusJson = page.getAsJsonObject("embedData").getAsJsonObject("status");

    Status status = Status.fromJson(statusJson.toString());
    assertEquals(1121915133L, status.getId());
    assertEquals("jkrums", status.getScreenname());
    assertEquals("http://twitpic.com/135xa - There's a plane in the Hudson. I'm on the ferry going to pick up the people. Crazy.", status.getText());

    asyncHttpClient.close();
  }
View Full Code Here
77
78
79
80
81
82
83
84
85
86
87
88
89
    stream = new JsonStatusBlockReader(statuses);

    Map<Long, String> ids = new HashMap<Long, String>();

    int cnt = 0;
    Status status;
    while ((status = stream.next()) != null) {
      ids.put(status.getId(), status.getJsonString());
      cnt++;
    }
    LOG.info(String.format("Total of %d statuses read.", cnt));

    BufferedReader in = new BufferedReader(new InputStreamReader(new FileInputStream(data)));
View Full Code Here
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
    LOG.info("collection: " + collectionPath);
    LOG.info("index: " + indexPath);

    IndexWriter writer = new IndexWriter(dir, config);
    int cnt = 0;
    Status status;
    try {
      while ((status = stream.next()) != null) {
        if (status.getText() == null) {
          continue;
        }

        cnt++;
        Document doc = new Document();
        doc.add(new LongField(StatusField.ID.name, status.getId(), Field.Store.YES));
        doc.add(new LongField(StatusField.EPOCH.name, status.getEpoch(), Field.Store.YES));
        doc.add(new TextField(StatusField.SCREEN_NAME.name, status.getScreenname(), Store.YES));

        doc.add(new Field(StatusField.TEXT.name, status.getText(), textOptions));

        doc.add(new IntField(StatusField.FRIENDS_COUNT.name, status.getFollowersCount(), Store.YES));
        doc.add(new IntField(StatusField.FOLLOWERS_COUNT.name, status.getFriendsCount(), Store.YES));
        doc.add(new IntField(StatusField.STATUSES_COUNT.name, status.getStatusesCount(), Store.YES));

        long inReplyToStatusId = status.getInReplyToStatusId();
        if (inReplyToStatusId > 0) {
          doc.add(new LongField(StatusField.IN_REPLY_TO_STATUS_ID.name, inReplyToStatusId, Field.Store.YES));
          doc.add(new LongField(StatusField.IN_REPLY_TO_USER_ID.name, status.getInReplyToUserId(), Field.Store.YES));
        }
       
        String lang = status.getLang();
        if (!lang.equals("unknown")) {
          doc.add(new TextField(StatusField.LANG.name, status.getLang(), Store.YES));
        }
       
        long retweetStatusId = status.getRetweetedStatusId();
        if (retweetStatusId > 0) {
          doc.add(new LongField(StatusField.RETWEETED_STATUS_ID.name, retweetStatusId, Field.Store.YES));
          doc.add(new LongField(StatusField.RETWEETED_USER_ID.name, status.getRetweetedUserId(), Field.Store.YES));
          doc.add(new IntField(StatusField.RETWEET_COUNT.name, status.getRetweetCount(), Store.YES));
          if ( status.getRetweetCount() < 0 || status.getRetweetedStatusId() < 0) {
            LOG.warn("Error parsing retweet fields of " + status.getId());
          }
        }
       
        writer.addDocument(doc);
        if (cnt % 100000 == 0) {
View Full Code Here
47
48
49
50
51
52
53
54
55
56
      System.exit(-1);
    }

    StatusStream stream = new JsonStatusCorpusReader(file);

    Status status;
    while ((status = stream.next()) != null) {
      System.out.println(status.getId() + "\t" + status.getScreenname());
    }
  }
View Full Code Here
63
64
65
66
67
68
69
70
71
72
      System.exit(-1);
    }

    StatusStream stream = new JsonStatusCorpusReader(file);

    Status status;
    while ((status = stream.next()) != null) {
      System.out.println(status.getId() + "\t" + status.getScreenname());
    }
  }
View Full Code Here
TOP

Related Classes of cc.twittertools.corpus.data.Status

  • cc.twittertools.corpus.demo.IndexStatuses
  • cc.twittertools.corpus.demo.ReadStatuses
  • cc.twittertools.download.FetchStatusTest
  • cc.twittertools.download.VerifyJsonStatusBlockCrawl
  • cc.twittertools.index.ExtractTweetidsFromCollection
  • cc.twittertools.index.IndexStatuses
  • cc.twittertools.search.indexing.IndexStatuses
  • cc.twittertools.util.ExtractSubcollection
  • cc.twittertools.util.VerifySubcollection
  • com.google.gson.JsonObject
    • 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.