private Tweet getTweetObjectFromStatus(Status status) {
Tweet tweet = new Tweet();
tweet.setId(Long.toString(status.getId()));
tweet.setText(status.getText());
tweet.setCreatedAt(status.getCreatedAt());
tweet.setFavCount(status.getFavoriteCount());
User user = status.getUser();
tweet.setUserId(user.getId());
tweet.setUserName(user.getName());
tweet.setUserScreenName(user.getScreenName());
HashtagEntity[] hashtagEntities = status.getHashtagEntities();
List<String> hashtags = new ArrayList<String>();
for (HashtagEntity hashtagEntity : hashtagEntities) {
hashtags.add(hashtagEntity.getText());
}
tweet.setHashTags(hashtags.toArray(new String[hashtags.size()]));
GeoLocation geoLocation = status.getGeoLocation();
if (geoLocation != null) {
double[] coordinates = {geoLocation.getLongitude(), geoLocation.getLatitude()};
tweet.setCoordinates(coordinates);
}
return tweet;
}
public void indexTweet(Tweet tweet) {
this.connect();
DBObject dbObject =
BasicDBObjectBuilder.start()
// .add("_id", status.getId())
.add("id", tweet.getId())
.add("text", tweet.getText())
.add("created_at", tweet.getCreatedAt())
.add("hashtags", tweet.getHashTags())
.add("userId", tweet.getUserId())
.add("userName", tweet.getUserName())
.add("userScreenName", tweet.getUserScreenName())
.add("coordinates", tweet.getCoordinates())
.get();
// WriteConcern wc = WriteConcern.UNACKNOWLEDGED;
WriteConcern wc = WriteConcern.ACKNOWLEDGED;
wc = wc.continueOnError(true);
WriteResult writeResult = this.dbCollection.insert(dbObject, wc);
}
public List<Tweet> readFile(String fileLocation, int limit, String fromTweetId) {
long startTime = System.currentTimeMillis();
int errors = 0;
int added = 0;
int ignored = 0;
int skipped = 0;
List<Tweet> tweets = new ArrayList<Tweet>();
try {
// Read gzFile
InputStream inputStream = new GZIPInputStream(new FileInputStream(fileLocation));
Reader decoder = new InputStreamReader(inputStream);
BufferedReader br = new BufferedReader(decoder);
String status;
while ((status = br.readLine()) != null) {
// Ignore garbage and log output lines, all statuses start with JSON bracket
if (!status.equals("") && status.charAt(0) == '{') {
try {
JSONObject jsonObject = new JSONObject(status);
// We use created_at as an indicator that this is a Tweet.
if (jsonObject.has("created_at")) {
Status statusObject = TwitterObjectFactory.createStatus(status);
Tweet tweet = this.getTweetObjectFromStatus(statusObject);
if (fromTweetId != null) {
if (fromTweetId.equals(tweet.getId())) {
this.log.write("StatusFileReader - Scanner pickup from " + fromTweetId);
fromTweetId = null;
} else {
skipped++;
}
continue;
}
added++;
tweets.add(tweet);
if (limit > 0 && added >= limit) {
break;
}
} else {
ignored++;
}
} catch (JSONException e) {
this.log.write(
"Exception in StatusFileReader: Json Parse Failure on: "
+ status
+ ", "
+ e.getMessage());
}
} else {
ignored++;
}
}
br.close();
decoder.close();
inputStream.close();
} catch (Exception e) {
this.log.write(
"Exception in StatusFileReader: Error Reading File: "
+ e.getClass().getName()
+ ": "
+ e.getMessage());
}
long runTimeSeconds = TimeUnit.MILLISECONDS.toSeconds(System.currentTimeMillis() - startTime);
double ops = ((double) added / (double) Math.max(runTimeSeconds, 1));
this.log.write(
"StatusFileReader - "
+ fileLocation
+ " processed in "
+ runTimeSeconds
+ "s. "
+ added
+ " ok / "
+ errors
+ " errors / "
+ ignored
+ " ignored / "
+ skipped
+ " skipped. "
+ ops
+ " ops. Limit: "
+ limit
+ ", Fetch: "
+ tweets.size());
return tweets;
}
