You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@hbase.apache.org by mi...@apache.org on 2016/01/14 18:24:27 UTC

[23/51] [partial] hbase-site git commit: Published site at c8b9754a5e0372f93d0e9cf1d9ce788c3d0dcda1.

http://git-wip-us.apache.org/repos/asf/hbase-site/blob/9577bd5e/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TextSortReducer.html
----------------------------------------------------------------------
diff --git a/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TextSortReducer.html b/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TextSortReducer.html
index 2542539..f3aa9a6 100644
--- a/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TextSortReducer.html
+++ b/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TextSortReducer.html
@@ -33,195 +33,196 @@
 <span class="sourceLineNo">025</span>import java.util.TreeSet;<a name="line.25"></a>
 <span class="sourceLineNo">026</span><a name="line.26"></a>
 <span class="sourceLineNo">027</span>import org.apache.hadoop.conf.Configuration;<a name="line.27"></a>
-<span class="sourceLineNo">028</span>import org.apache.hadoop.hbase.Cell;<a name="line.28"></a>
-<span class="sourceLineNo">029</span>import org.apache.hadoop.hbase.CellComparator;<a name="line.29"></a>
-<span class="sourceLineNo">030</span>import org.apache.hadoop.hbase.KeyValue;<a name="line.30"></a>
-<span class="sourceLineNo">031</span>import org.apache.hadoop.hbase.KeyValueUtil;<a name="line.31"></a>
-<span class="sourceLineNo">032</span>import org.apache.hadoop.hbase.Tag;<a name="line.32"></a>
-<span class="sourceLineNo">033</span>import org.apache.hadoop.hbase.TagType;<a name="line.33"></a>
-<span class="sourceLineNo">034</span>import org.apache.hadoop.hbase.classification.InterfaceAudience;<a name="line.34"></a>
-<span class="sourceLineNo">035</span>import org.apache.hadoop.hbase.classification.InterfaceStability;<a name="line.35"></a>
-<span class="sourceLineNo">036</span>import org.apache.hadoop.hbase.io.ImmutableBytesWritable;<a name="line.36"></a>
-<span class="sourceLineNo">037</span>import org.apache.hadoop.hbase.security.visibility.InvalidLabelException;<a name="line.37"></a>
-<span class="sourceLineNo">038</span>import org.apache.hadoop.hbase.util.Base64;<a name="line.38"></a>
-<span class="sourceLineNo">039</span>import org.apache.hadoop.hbase.util.Bytes;<a name="line.39"></a>
-<span class="sourceLineNo">040</span>import org.apache.hadoop.io.Text;<a name="line.40"></a>
-<span class="sourceLineNo">041</span>import org.apache.hadoop.mapreduce.Counter;<a name="line.41"></a>
-<span class="sourceLineNo">042</span>import org.apache.hadoop.mapreduce.Reducer;<a name="line.42"></a>
-<span class="sourceLineNo">043</span>import org.apache.hadoop.util.StringUtils;<a name="line.43"></a>
-<span class="sourceLineNo">044</span><a name="line.44"></a>
-<span class="sourceLineNo">045</span>/**<a name="line.45"></a>
-<span class="sourceLineNo">046</span> * Emits Sorted KeyValues. Parse the passed text and creates KeyValues. Sorts them before emit.<a name="line.46"></a>
-<span class="sourceLineNo">047</span> * @see HFileOutputFormat2<a name="line.47"></a>
-<span class="sourceLineNo">048</span> * @see KeyValueSortReducer<a name="line.48"></a>
-<span class="sourceLineNo">049</span> * @see PutSortReducer<a name="line.49"></a>
-<span class="sourceLineNo">050</span> */<a name="line.50"></a>
-<span class="sourceLineNo">051</span>@InterfaceAudience.Public<a name="line.51"></a>
-<span class="sourceLineNo">052</span>@InterfaceStability.Evolving<a name="line.52"></a>
-<span class="sourceLineNo">053</span>public class TextSortReducer extends<a name="line.53"></a>
-<span class="sourceLineNo">054</span>    Reducer&lt;ImmutableBytesWritable, Text, ImmutableBytesWritable, KeyValue&gt; {<a name="line.54"></a>
-<span class="sourceLineNo">055</span>  <a name="line.55"></a>
-<span class="sourceLineNo">056</span>  /** Timestamp for all inserted rows */<a name="line.56"></a>
-<span class="sourceLineNo">057</span>  private long ts;<a name="line.57"></a>
-<span class="sourceLineNo">058</span><a name="line.58"></a>
-<span class="sourceLineNo">059</span>  /** Column seperator */<a name="line.59"></a>
-<span class="sourceLineNo">060</span>  private String separator;<a name="line.60"></a>
-<span class="sourceLineNo">061</span><a name="line.61"></a>
-<span class="sourceLineNo">062</span>  /** Should skip bad lines */<a name="line.62"></a>
-<span class="sourceLineNo">063</span>  private boolean skipBadLines;<a name="line.63"></a>
-<span class="sourceLineNo">064</span>  <a name="line.64"></a>
-<span class="sourceLineNo">065</span>  private Counter badLineCount;<a name="line.65"></a>
-<span class="sourceLineNo">066</span><a name="line.66"></a>
-<span class="sourceLineNo">067</span>  private ImportTsv.TsvParser parser;<a name="line.67"></a>
-<span class="sourceLineNo">068</span><a name="line.68"></a>
-<span class="sourceLineNo">069</span>  /** Cell visibility expr **/<a name="line.69"></a>
-<span class="sourceLineNo">070</span>  private String cellVisibilityExpr;<a name="line.70"></a>
-<span class="sourceLineNo">071</span><a name="line.71"></a>
-<span class="sourceLineNo">072</span>  /** Cell TTL */<a name="line.72"></a>
-<span class="sourceLineNo">073</span>  private long ttl;<a name="line.73"></a>
-<span class="sourceLineNo">074</span><a name="line.74"></a>
-<span class="sourceLineNo">075</span>  private CellCreator kvCreator;<a name="line.75"></a>
-<span class="sourceLineNo">076</span><a name="line.76"></a>
-<span class="sourceLineNo">077</span>  public long getTs() {<a name="line.77"></a>
-<span class="sourceLineNo">078</span>    return ts;<a name="line.78"></a>
-<span class="sourceLineNo">079</span>  }<a name="line.79"></a>
-<span class="sourceLineNo">080</span><a name="line.80"></a>
-<span class="sourceLineNo">081</span>  public boolean getSkipBadLines() {<a name="line.81"></a>
-<span class="sourceLineNo">082</span>    return skipBadLines;<a name="line.82"></a>
-<span class="sourceLineNo">083</span>  }<a name="line.83"></a>
-<span class="sourceLineNo">084</span><a name="line.84"></a>
-<span class="sourceLineNo">085</span>  public Counter getBadLineCount() {<a name="line.85"></a>
-<span class="sourceLineNo">086</span>    return badLineCount;<a name="line.86"></a>
-<span class="sourceLineNo">087</span>  }<a name="line.87"></a>
-<span class="sourceLineNo">088</span><a name="line.88"></a>
-<span class="sourceLineNo">089</span>  public void incrementBadLineCount(int count) {<a name="line.89"></a>
-<span class="sourceLineNo">090</span>    this.badLineCount.increment(count);<a name="line.90"></a>
-<span class="sourceLineNo">091</span>  }<a name="line.91"></a>
-<span class="sourceLineNo">092</span><a name="line.92"></a>
-<span class="sourceLineNo">093</span>  /**<a name="line.93"></a>
-<span class="sourceLineNo">094</span>   * Handles initializing this class with objects specific to it (i.e., the parser).<a name="line.94"></a>
-<span class="sourceLineNo">095</span>   * Common initialization that might be leveraged by a subsclass is done in<a name="line.95"></a>
-<span class="sourceLineNo">096</span>   * &lt;code&gt;doSetup&lt;/code&gt;. Hence a subclass may choose to override this method<a name="line.96"></a>
-<span class="sourceLineNo">097</span>   * and call &lt;code&gt;doSetup&lt;/code&gt; as well before handling it's own custom params.<a name="line.97"></a>
-<span class="sourceLineNo">098</span>   *<a name="line.98"></a>
-<span class="sourceLineNo">099</span>   * @param context<a name="line.99"></a>
-<span class="sourceLineNo">100</span>   */<a name="line.100"></a>
-<span class="sourceLineNo">101</span>  @Override<a name="line.101"></a>
-<span class="sourceLineNo">102</span>  protected void setup(Context context) {<a name="line.102"></a>
-<span class="sourceLineNo">103</span>    doSetup(context);<a name="line.103"></a>
-<span class="sourceLineNo">104</span><a name="line.104"></a>
-<span class="sourceLineNo">105</span>    Configuration conf = context.getConfiguration();<a name="line.105"></a>
-<span class="sourceLineNo">106</span><a name="line.106"></a>
-<span class="sourceLineNo">107</span>    parser = new ImportTsv.TsvParser(conf.get(ImportTsv.COLUMNS_CONF_KEY), separator);<a name="line.107"></a>
-<span class="sourceLineNo">108</span>    if (parser.getRowKeyColumnIndex() == -1) {<a name="line.108"></a>
-<span class="sourceLineNo">109</span>      throw new RuntimeException("No row key column specified");<a name="line.109"></a>
-<span class="sourceLineNo">110</span>    }<a name="line.110"></a>
-<span class="sourceLineNo">111</span>    this.kvCreator = new CellCreator(conf);<a name="line.111"></a>
-<span class="sourceLineNo">112</span>  }<a name="line.112"></a>
-<span class="sourceLineNo">113</span><a name="line.113"></a>
-<span class="sourceLineNo">114</span>  /**<a name="line.114"></a>
-<span class="sourceLineNo">115</span>   * Handles common parameter initialization that a subclass might want to leverage.<a name="line.115"></a>
-<span class="sourceLineNo">116</span>   * @param context<a name="line.116"></a>
-<span class="sourceLineNo">117</span>   */<a name="line.117"></a>
-<span class="sourceLineNo">118</span>  protected void doSetup(Context context) {<a name="line.118"></a>
-<span class="sourceLineNo">119</span>    Configuration conf = context.getConfiguration();<a name="line.119"></a>
-<span class="sourceLineNo">120</span><a name="line.120"></a>
-<span class="sourceLineNo">121</span>    // If a custom separator has been used,<a name="line.121"></a>
-<span class="sourceLineNo">122</span>    // decode it back from Base64 encoding.<a name="line.122"></a>
-<span class="sourceLineNo">123</span>    separator = conf.get(ImportTsv.SEPARATOR_CONF_KEY);<a name="line.123"></a>
-<span class="sourceLineNo">124</span>    if (separator == null) {<a name="line.124"></a>
-<span class="sourceLineNo">125</span>      separator = ImportTsv.DEFAULT_SEPARATOR;<a name="line.125"></a>
-<span class="sourceLineNo">126</span>    } else {<a name="line.126"></a>
-<span class="sourceLineNo">127</span>      separator = new String(Base64.decode(separator));<a name="line.127"></a>
-<span class="sourceLineNo">128</span>    }<a name="line.128"></a>
-<span class="sourceLineNo">129</span><a name="line.129"></a>
-<span class="sourceLineNo">130</span>    // Should never get 0 as we are setting this to a valid value in job configuration.<a name="line.130"></a>
-<span class="sourceLineNo">131</span>    ts = conf.getLong(ImportTsv.TIMESTAMP_CONF_KEY, 0);<a name="line.131"></a>
-<span class="sourceLineNo">132</span><a name="line.132"></a>
-<span class="sourceLineNo">133</span>    skipBadLines = context.getConfiguration().getBoolean(ImportTsv.SKIP_LINES_CONF_KEY, true);<a name="line.133"></a>
-<span class="sourceLineNo">134</span>    badLineCount = context.getCounter("ImportTsv", "Bad Lines");<a name="line.134"></a>
-<span class="sourceLineNo">135</span>  }<a name="line.135"></a>
-<span class="sourceLineNo">136</span>  <a name="line.136"></a>
-<span class="sourceLineNo">137</span>  @Override<a name="line.137"></a>
-<span class="sourceLineNo">138</span>  protected void reduce(<a name="line.138"></a>
-<span class="sourceLineNo">139</span>      ImmutableBytesWritable rowKey,<a name="line.139"></a>
-<span class="sourceLineNo">140</span>      java.lang.Iterable&lt;Text&gt; lines,<a name="line.140"></a>
-<span class="sourceLineNo">141</span>      Reducer&lt;ImmutableBytesWritable, Text,<a name="line.141"></a>
-<span class="sourceLineNo">142</span>              ImmutableBytesWritable, KeyValue&gt;.Context context)<a name="line.142"></a>
-<span class="sourceLineNo">143</span>      throws java.io.IOException, InterruptedException<a name="line.143"></a>
-<span class="sourceLineNo">144</span>  {<a name="line.144"></a>
-<span class="sourceLineNo">145</span>    // although reduce() is called per-row, handle pathological case<a name="line.145"></a>
-<span class="sourceLineNo">146</span>    long threshold = context.getConfiguration().getLong(<a name="line.146"></a>
-<span class="sourceLineNo">147</span>        "reducer.row.threshold", 1L * (1&lt;&lt;30));<a name="line.147"></a>
-<span class="sourceLineNo">148</span>    Iterator&lt;Text&gt; iter = lines.iterator();<a name="line.148"></a>
-<span class="sourceLineNo">149</span>    while (iter.hasNext()) {<a name="line.149"></a>
-<span class="sourceLineNo">150</span>      Set&lt;KeyValue&gt; kvs = new TreeSet&lt;KeyValue&gt;(CellComparator.COMPARATOR);<a name="line.150"></a>
-<span class="sourceLineNo">151</span>      long curSize = 0;<a name="line.151"></a>
-<span class="sourceLineNo">152</span>      // stop at the end or the RAM threshold<a name="line.152"></a>
-<span class="sourceLineNo">153</span>      while (iter.hasNext() &amp;&amp; curSize &lt; threshold) {<a name="line.153"></a>
-<span class="sourceLineNo">154</span>        Text line = iter.next();<a name="line.154"></a>
-<span class="sourceLineNo">155</span>        byte[] lineBytes = line.getBytes();<a name="line.155"></a>
-<span class="sourceLineNo">156</span>        try {<a name="line.156"></a>
-<span class="sourceLineNo">157</span>          ImportTsv.TsvParser.ParsedLine parsed = parser.parse(lineBytes, line.getLength());<a name="line.157"></a>
-<span class="sourceLineNo">158</span>          // Retrieve timestamp if exists<a name="line.158"></a>
-<span class="sourceLineNo">159</span>          ts = parsed.getTimestamp(ts);<a name="line.159"></a>
-<span class="sourceLineNo">160</span>          cellVisibilityExpr = parsed.getCellVisibility();<a name="line.160"></a>
-<span class="sourceLineNo">161</span>          ttl = parsed.getCellTTL();<a name="line.161"></a>
-<span class="sourceLineNo">162</span>          <a name="line.162"></a>
-<span class="sourceLineNo">163</span>          // create tags for the parsed line<a name="line.163"></a>
-<span class="sourceLineNo">164</span>          List&lt;Tag&gt; tags = new ArrayList&lt;Tag&gt;();<a name="line.164"></a>
-<span class="sourceLineNo">165</span>          if (cellVisibilityExpr != null) {<a name="line.165"></a>
-<span class="sourceLineNo">166</span>            tags.addAll(kvCreator.getVisibilityExpressionResolver().createVisibilityExpTags(<a name="line.166"></a>
-<span class="sourceLineNo">167</span>              cellVisibilityExpr));<a name="line.167"></a>
-<span class="sourceLineNo">168</span>          }<a name="line.168"></a>
-<span class="sourceLineNo">169</span>          // Add TTL directly to the KV so we can vary them when packing more than one KV<a name="line.169"></a>
-<span class="sourceLineNo">170</span>          // into puts<a name="line.170"></a>
-<span class="sourceLineNo">171</span>          if (ttl &gt; 0) {<a name="line.171"></a>
-<span class="sourceLineNo">172</span>            tags.add(new Tag(TagType.TTL_TAG_TYPE, Bytes.toBytes(ttl)));<a name="line.172"></a>
-<span class="sourceLineNo">173</span>          }<a name="line.173"></a>
-<span class="sourceLineNo">174</span>          for (int i = 0; i &lt; parsed.getColumnCount(); i++) {<a name="line.174"></a>
-<span class="sourceLineNo">175</span>            if (i == parser.getRowKeyColumnIndex() || i == parser.getTimestampKeyColumnIndex()<a name="line.175"></a>
-<span class="sourceLineNo">176</span>                || i == parser.getAttributesKeyColumnIndex() || i == parser.getCellVisibilityColumnIndex()<a name="line.176"></a>
-<span class="sourceLineNo">177</span>                || i == parser.getCellTTLColumnIndex()) {<a name="line.177"></a>
-<span class="sourceLineNo">178</span>              continue;<a name="line.178"></a>
-<span class="sourceLineNo">179</span>            }<a name="line.179"></a>
-<span class="sourceLineNo">180</span>            // Creating the KV which needs to be directly written to HFiles. Using the Facade<a name="line.180"></a>
-<span class="sourceLineNo">181</span>            // KVCreator for creation of kvs.<a name="line.181"></a>
-<span class="sourceLineNo">182</span>            Cell cell = this.kvCreator.create(lineBytes, parsed.getRowKeyOffset(),<a name="line.182"></a>
-<span class="sourceLineNo">183</span>                parsed.getRowKeyLength(), parser.getFamily(i), 0, parser.getFamily(i).length,<a name="line.183"></a>
-<span class="sourceLineNo">184</span>                parser.getQualifier(i), 0, parser.getQualifier(i).length, ts, lineBytes,<a name="line.184"></a>
-<span class="sourceLineNo">185</span>                parsed.getColumnOffset(i), parsed.getColumnLength(i), tags);<a name="line.185"></a>
-<span class="sourceLineNo">186</span>            KeyValue kv = KeyValueUtil.ensureKeyValue(cell);<a name="line.186"></a>
-<span class="sourceLineNo">187</span>            kvs.add(kv);<a name="line.187"></a>
-<span class="sourceLineNo">188</span>            curSize += kv.heapSize();<a name="line.188"></a>
-<span class="sourceLineNo">189</span>          }<a name="line.189"></a>
-<span class="sourceLineNo">190</span>        } catch (ImportTsv.TsvParser.BadTsvLineException | IllegalArgumentException<a name="line.190"></a>
-<span class="sourceLineNo">191</span>            | InvalidLabelException badLine) {<a name="line.191"></a>
-<span class="sourceLineNo">192</span>          if (skipBadLines) {<a name="line.192"></a>
-<span class="sourceLineNo">193</span>            System.err.println("Bad line." + badLine.getMessage());<a name="line.193"></a>
-<span class="sourceLineNo">194</span>            incrementBadLineCount(1);<a name="line.194"></a>
-<span class="sourceLineNo">195</span>            continue;<a name="line.195"></a>
-<span class="sourceLineNo">196</span>          }<a name="line.196"></a>
-<span class="sourceLineNo">197</span>          throw new IOException(badLine);<a name="line.197"></a>
-<span class="sourceLineNo">198</span>        }<a name="line.198"></a>
-<span class="sourceLineNo">199</span>      }<a name="line.199"></a>
-<span class="sourceLineNo">200</span>      context.setStatus("Read " + kvs.size() + " entries of " + kvs.getClass()<a name="line.200"></a>
-<span class="sourceLineNo">201</span>          + "(" + StringUtils.humanReadableInt(curSize) + ")");<a name="line.201"></a>
-<span class="sourceLineNo">202</span>      int index = 0;<a name="line.202"></a>
-<span class="sourceLineNo">203</span>      for (KeyValue kv : kvs) {<a name="line.203"></a>
-<span class="sourceLineNo">204</span>        context.write(rowKey, kv);<a name="line.204"></a>
-<span class="sourceLineNo">205</span>        if (++index &gt; 0 &amp;&amp; index % 100 == 0)<a name="line.205"></a>
-<span class="sourceLineNo">206</span>          context.setStatus("Wrote " + index + " key values.");<a name="line.206"></a>
-<span class="sourceLineNo">207</span>      }<a name="line.207"></a>
-<span class="sourceLineNo">208</span><a name="line.208"></a>
-<span class="sourceLineNo">209</span>      // if we have more entries to process<a name="line.209"></a>
-<span class="sourceLineNo">210</span>      if (iter.hasNext()) {<a name="line.210"></a>
-<span class="sourceLineNo">211</span>        // force flush because we cannot guarantee intra-row sorted order<a name="line.211"></a>
-<span class="sourceLineNo">212</span>        context.write(null, null);<a name="line.212"></a>
-<span class="sourceLineNo">213</span>      }<a name="line.213"></a>
-<span class="sourceLineNo">214</span>    }<a name="line.214"></a>
-<span class="sourceLineNo">215</span>  }<a name="line.215"></a>
-<span class="sourceLineNo">216</span>}<a name="line.216"></a>
+<span class="sourceLineNo">028</span>import org.apache.hadoop.hbase.ArrayBackedTag;<a name="line.28"></a>
+<span class="sourceLineNo">029</span>import org.apache.hadoop.hbase.Cell;<a name="line.29"></a>
+<span class="sourceLineNo">030</span>import org.apache.hadoop.hbase.CellComparator;<a name="line.30"></a>
+<span class="sourceLineNo">031</span>import org.apache.hadoop.hbase.KeyValue;<a name="line.31"></a>
+<span class="sourceLineNo">032</span>import org.apache.hadoop.hbase.KeyValueUtil;<a name="line.32"></a>
+<span class="sourceLineNo">033</span>import org.apache.hadoop.hbase.Tag;<a name="line.33"></a>
+<span class="sourceLineNo">034</span>import org.apache.hadoop.hbase.TagType;<a name="line.34"></a>
+<span class="sourceLineNo">035</span>import org.apache.hadoop.hbase.classification.InterfaceAudience;<a name="line.35"></a>
+<span class="sourceLineNo">036</span>import org.apache.hadoop.hbase.classification.InterfaceStability;<a name="line.36"></a>
+<span class="sourceLineNo">037</span>import org.apache.hadoop.hbase.io.ImmutableBytesWritable;<a name="line.37"></a>
+<span class="sourceLineNo">038</span>import org.apache.hadoop.hbase.security.visibility.InvalidLabelException;<a name="line.38"></a>
+<span class="sourceLineNo">039</span>import org.apache.hadoop.hbase.util.Base64;<a name="line.39"></a>
+<span class="sourceLineNo">040</span>import org.apache.hadoop.hbase.util.Bytes;<a name="line.40"></a>
+<span class="sourceLineNo">041</span>import org.apache.hadoop.io.Text;<a name="line.41"></a>
+<span class="sourceLineNo">042</span>import org.apache.hadoop.mapreduce.Counter;<a name="line.42"></a>
+<span class="sourceLineNo">043</span>import org.apache.hadoop.mapreduce.Reducer;<a name="line.43"></a>
+<span class="sourceLineNo">044</span>import org.apache.hadoop.util.StringUtils;<a name="line.44"></a>
+<span class="sourceLineNo">045</span><a name="line.45"></a>
+<span class="sourceLineNo">046</span>/**<a name="line.46"></a>
+<span class="sourceLineNo">047</span> * Emits Sorted KeyValues. Parse the passed text and creates KeyValues. Sorts them before emit.<a name="line.47"></a>
+<span class="sourceLineNo">048</span> * @see HFileOutputFormat2<a name="line.48"></a>
+<span class="sourceLineNo">049</span> * @see KeyValueSortReducer<a name="line.49"></a>
+<span class="sourceLineNo">050</span> * @see PutSortReducer<a name="line.50"></a>
+<span class="sourceLineNo">051</span> */<a name="line.51"></a>
+<span class="sourceLineNo">052</span>@InterfaceAudience.Public<a name="line.52"></a>
+<span class="sourceLineNo">053</span>@InterfaceStability.Evolving<a name="line.53"></a>
+<span class="sourceLineNo">054</span>public class TextSortReducer extends<a name="line.54"></a>
+<span class="sourceLineNo">055</span>    Reducer&lt;ImmutableBytesWritable, Text, ImmutableBytesWritable, KeyValue&gt; {<a name="line.55"></a>
+<span class="sourceLineNo">056</span>  <a name="line.56"></a>
+<span class="sourceLineNo">057</span>  /** Timestamp for all inserted rows */<a name="line.57"></a>
+<span class="sourceLineNo">058</span>  private long ts;<a name="line.58"></a>
+<span class="sourceLineNo">059</span><a name="line.59"></a>
+<span class="sourceLineNo">060</span>  /** Column seperator */<a name="line.60"></a>
+<span class="sourceLineNo">061</span>  private String separator;<a name="line.61"></a>
+<span class="sourceLineNo">062</span><a name="line.62"></a>
+<span class="sourceLineNo">063</span>  /** Should skip bad lines */<a name="line.63"></a>
+<span class="sourceLineNo">064</span>  private boolean skipBadLines;<a name="line.64"></a>
+<span class="sourceLineNo">065</span>  <a name="line.65"></a>
+<span class="sourceLineNo">066</span>  private Counter badLineCount;<a name="line.66"></a>
+<span class="sourceLineNo">067</span><a name="line.67"></a>
+<span class="sourceLineNo">068</span>  private ImportTsv.TsvParser parser;<a name="line.68"></a>
+<span class="sourceLineNo">069</span><a name="line.69"></a>
+<span class="sourceLineNo">070</span>  /** Cell visibility expr **/<a name="line.70"></a>
+<span class="sourceLineNo">071</span>  private String cellVisibilityExpr;<a name="line.71"></a>
+<span class="sourceLineNo">072</span><a name="line.72"></a>
+<span class="sourceLineNo">073</span>  /** Cell TTL */<a name="line.73"></a>
+<span class="sourceLineNo">074</span>  private long ttl;<a name="line.74"></a>
+<span class="sourceLineNo">075</span><a name="line.75"></a>
+<span class="sourceLineNo">076</span>  private CellCreator kvCreator;<a name="line.76"></a>
+<span class="sourceLineNo">077</span><a name="line.77"></a>
+<span class="sourceLineNo">078</span>  public long getTs() {<a name="line.78"></a>
+<span class="sourceLineNo">079</span>    return ts;<a name="line.79"></a>
+<span class="sourceLineNo">080</span>  }<a name="line.80"></a>
+<span class="sourceLineNo">081</span><a name="line.81"></a>
+<span class="sourceLineNo">082</span>  public boolean getSkipBadLines() {<a name="line.82"></a>
+<span class="sourceLineNo">083</span>    return skipBadLines;<a name="line.83"></a>
+<span class="sourceLineNo">084</span>  }<a name="line.84"></a>
+<span class="sourceLineNo">085</span><a name="line.85"></a>
+<span class="sourceLineNo">086</span>  public Counter getBadLineCount() {<a name="line.86"></a>
+<span class="sourceLineNo">087</span>    return badLineCount;<a name="line.87"></a>
+<span class="sourceLineNo">088</span>  }<a name="line.88"></a>
+<span class="sourceLineNo">089</span><a name="line.89"></a>
+<span class="sourceLineNo">090</span>  public void incrementBadLineCount(int count) {<a name="line.90"></a>
+<span class="sourceLineNo">091</span>    this.badLineCount.increment(count);<a name="line.91"></a>
+<span class="sourceLineNo">092</span>  }<a name="line.92"></a>
+<span class="sourceLineNo">093</span><a name="line.93"></a>
+<span class="sourceLineNo">094</span>  /**<a name="line.94"></a>
+<span class="sourceLineNo">095</span>   * Handles initializing this class with objects specific to it (i.e., the parser).<a name="line.95"></a>
+<span class="sourceLineNo">096</span>   * Common initialization that might be leveraged by a subsclass is done in<a name="line.96"></a>
+<span class="sourceLineNo">097</span>   * &lt;code&gt;doSetup&lt;/code&gt;. Hence a subclass may choose to override this method<a name="line.97"></a>
+<span class="sourceLineNo">098</span>   * and call &lt;code&gt;doSetup&lt;/code&gt; as well before handling it's own custom params.<a name="line.98"></a>
+<span class="sourceLineNo">099</span>   *<a name="line.99"></a>
+<span class="sourceLineNo">100</span>   * @param context<a name="line.100"></a>
+<span class="sourceLineNo">101</span>   */<a name="line.101"></a>
+<span class="sourceLineNo">102</span>  @Override<a name="line.102"></a>
+<span class="sourceLineNo">103</span>  protected void setup(Context context) {<a name="line.103"></a>
+<span class="sourceLineNo">104</span>    doSetup(context);<a name="line.104"></a>
+<span class="sourceLineNo">105</span><a name="line.105"></a>
+<span class="sourceLineNo">106</span>    Configuration conf = context.getConfiguration();<a name="line.106"></a>
+<span class="sourceLineNo">107</span><a name="line.107"></a>
+<span class="sourceLineNo">108</span>    parser = new ImportTsv.TsvParser(conf.get(ImportTsv.COLUMNS_CONF_KEY), separator);<a name="line.108"></a>
+<span class="sourceLineNo">109</span>    if (parser.getRowKeyColumnIndex() == -1) {<a name="line.109"></a>
+<span class="sourceLineNo">110</span>      throw new RuntimeException("No row key column specified");<a name="line.110"></a>
+<span class="sourceLineNo">111</span>    }<a name="line.111"></a>
+<span class="sourceLineNo">112</span>    this.kvCreator = new CellCreator(conf);<a name="line.112"></a>
+<span class="sourceLineNo">113</span>  }<a name="line.113"></a>
+<span class="sourceLineNo">114</span><a name="line.114"></a>
+<span class="sourceLineNo">115</span>  /**<a name="line.115"></a>
+<span class="sourceLineNo">116</span>   * Handles common parameter initialization that a subclass might want to leverage.<a name="line.116"></a>
+<span class="sourceLineNo">117</span>   * @param context<a name="line.117"></a>
+<span class="sourceLineNo">118</span>   */<a name="line.118"></a>
+<span class="sourceLineNo">119</span>  protected void doSetup(Context context) {<a name="line.119"></a>
+<span class="sourceLineNo">120</span>    Configuration conf = context.getConfiguration();<a name="line.120"></a>
+<span class="sourceLineNo">121</span><a name="line.121"></a>
+<span class="sourceLineNo">122</span>    // If a custom separator has been used,<a name="line.122"></a>
+<span class="sourceLineNo">123</span>    // decode it back from Base64 encoding.<a name="line.123"></a>
+<span class="sourceLineNo">124</span>    separator = conf.get(ImportTsv.SEPARATOR_CONF_KEY);<a name="line.124"></a>
+<span class="sourceLineNo">125</span>    if (separator == null) {<a name="line.125"></a>
+<span class="sourceLineNo">126</span>      separator = ImportTsv.DEFAULT_SEPARATOR;<a name="line.126"></a>
+<span class="sourceLineNo">127</span>    } else {<a name="line.127"></a>
+<span class="sourceLineNo">128</span>      separator = new String(Base64.decode(separator));<a name="line.128"></a>
+<span class="sourceLineNo">129</span>    }<a name="line.129"></a>
+<span class="sourceLineNo">130</span><a name="line.130"></a>
+<span class="sourceLineNo">131</span>    // Should never get 0 as we are setting this to a valid value in job configuration.<a name="line.131"></a>
+<span class="sourceLineNo">132</span>    ts = conf.getLong(ImportTsv.TIMESTAMP_CONF_KEY, 0);<a name="line.132"></a>
+<span class="sourceLineNo">133</span><a name="line.133"></a>
+<span class="sourceLineNo">134</span>    skipBadLines = context.getConfiguration().getBoolean(ImportTsv.SKIP_LINES_CONF_KEY, true);<a name="line.134"></a>
+<span class="sourceLineNo">135</span>    badLineCount = context.getCounter("ImportTsv", "Bad Lines");<a name="line.135"></a>
+<span class="sourceLineNo">136</span>  }<a name="line.136"></a>
+<span class="sourceLineNo">137</span>  <a name="line.137"></a>
+<span class="sourceLineNo">138</span>  @Override<a name="line.138"></a>
+<span class="sourceLineNo">139</span>  protected void reduce(<a name="line.139"></a>
+<span class="sourceLineNo">140</span>      ImmutableBytesWritable rowKey,<a name="line.140"></a>
+<span class="sourceLineNo">141</span>      java.lang.Iterable&lt;Text&gt; lines,<a name="line.141"></a>
+<span class="sourceLineNo">142</span>      Reducer&lt;ImmutableBytesWritable, Text,<a name="line.142"></a>
+<span class="sourceLineNo">143</span>              ImmutableBytesWritable, KeyValue&gt;.Context context)<a name="line.143"></a>
+<span class="sourceLineNo">144</span>      throws java.io.IOException, InterruptedException<a name="line.144"></a>
+<span class="sourceLineNo">145</span>  {<a name="line.145"></a>
+<span class="sourceLineNo">146</span>    // although reduce() is called per-row, handle pathological case<a name="line.146"></a>
+<span class="sourceLineNo">147</span>    long threshold = context.getConfiguration().getLong(<a name="line.147"></a>
+<span class="sourceLineNo">148</span>        "reducer.row.threshold", 1L * (1&lt;&lt;30));<a name="line.148"></a>
+<span class="sourceLineNo">149</span>    Iterator&lt;Text&gt; iter = lines.iterator();<a name="line.149"></a>
+<span class="sourceLineNo">150</span>    while (iter.hasNext()) {<a name="line.150"></a>
+<span class="sourceLineNo">151</span>      Set&lt;KeyValue&gt; kvs = new TreeSet&lt;KeyValue&gt;(CellComparator.COMPARATOR);<a name="line.151"></a>
+<span class="sourceLineNo">152</span>      long curSize = 0;<a name="line.152"></a>
+<span class="sourceLineNo">153</span>      // stop at the end or the RAM threshold<a name="line.153"></a>
+<span class="sourceLineNo">154</span>      while (iter.hasNext() &amp;&amp; curSize &lt; threshold) {<a name="line.154"></a>
+<span class="sourceLineNo">155</span>        Text line = iter.next();<a name="line.155"></a>
+<span class="sourceLineNo">156</span>        byte[] lineBytes = line.getBytes();<a name="line.156"></a>
+<span class="sourceLineNo">157</span>        try {<a name="line.157"></a>
+<span class="sourceLineNo">158</span>          ImportTsv.TsvParser.ParsedLine parsed = parser.parse(lineBytes, line.getLength());<a name="line.158"></a>
+<span class="sourceLineNo">159</span>          // Retrieve timestamp if exists<a name="line.159"></a>
+<span class="sourceLineNo">160</span>          ts = parsed.getTimestamp(ts);<a name="line.160"></a>
+<span class="sourceLineNo">161</span>          cellVisibilityExpr = parsed.getCellVisibility();<a name="line.161"></a>
+<span class="sourceLineNo">162</span>          ttl = parsed.getCellTTL();<a name="line.162"></a>
+<span class="sourceLineNo">163</span>          <a name="line.163"></a>
+<span class="sourceLineNo">164</span>          // create tags for the parsed line<a name="line.164"></a>
+<span class="sourceLineNo">165</span>          List&lt;Tag&gt; tags = new ArrayList&lt;Tag&gt;();<a name="line.165"></a>
+<span class="sourceLineNo">166</span>          if (cellVisibilityExpr != null) {<a name="line.166"></a>
+<span class="sourceLineNo">167</span>            tags.addAll(kvCreator.getVisibilityExpressionResolver().createVisibilityExpTags(<a name="line.167"></a>
+<span class="sourceLineNo">168</span>              cellVisibilityExpr));<a name="line.168"></a>
+<span class="sourceLineNo">169</span>          }<a name="line.169"></a>
+<span class="sourceLineNo">170</span>          // Add TTL directly to the KV so we can vary them when packing more than one KV<a name="line.170"></a>
+<span class="sourceLineNo">171</span>          // into puts<a name="line.171"></a>
+<span class="sourceLineNo">172</span>          if (ttl &gt; 0) {<a name="line.172"></a>
+<span class="sourceLineNo">173</span>            tags.add(new ArrayBackedTag(TagType.TTL_TAG_TYPE, Bytes.toBytes(ttl)));<a name="line.173"></a>
+<span class="sourceLineNo">174</span>          }<a name="line.174"></a>
+<span class="sourceLineNo">175</span>          for (int i = 0; i &lt; parsed.getColumnCount(); i++) {<a name="line.175"></a>
+<span class="sourceLineNo">176</span>            if (i == parser.getRowKeyColumnIndex() || i == parser.getTimestampKeyColumnIndex()<a name="line.176"></a>
+<span class="sourceLineNo">177</span>                || i == parser.getAttributesKeyColumnIndex() || i == parser.getCellVisibilityColumnIndex()<a name="line.177"></a>
+<span class="sourceLineNo">178</span>                || i == parser.getCellTTLColumnIndex()) {<a name="line.178"></a>
+<span class="sourceLineNo">179</span>              continue;<a name="line.179"></a>
+<span class="sourceLineNo">180</span>            }<a name="line.180"></a>
+<span class="sourceLineNo">181</span>            // Creating the KV which needs to be directly written to HFiles. Using the Facade<a name="line.181"></a>
+<span class="sourceLineNo">182</span>            // KVCreator for creation of kvs.<a name="line.182"></a>
+<span class="sourceLineNo">183</span>            Cell cell = this.kvCreator.create(lineBytes, parsed.getRowKeyOffset(),<a name="line.183"></a>
+<span class="sourceLineNo">184</span>                parsed.getRowKeyLength(), parser.getFamily(i), 0, parser.getFamily(i).length,<a name="line.184"></a>
+<span class="sourceLineNo">185</span>                parser.getQualifier(i), 0, parser.getQualifier(i).length, ts, lineBytes,<a name="line.185"></a>
+<span class="sourceLineNo">186</span>                parsed.getColumnOffset(i), parsed.getColumnLength(i), tags);<a name="line.186"></a>
+<span class="sourceLineNo">187</span>            KeyValue kv = KeyValueUtil.ensureKeyValue(cell);<a name="line.187"></a>
+<span class="sourceLineNo">188</span>            kvs.add(kv);<a name="line.188"></a>
+<span class="sourceLineNo">189</span>            curSize += kv.heapSize();<a name="line.189"></a>
+<span class="sourceLineNo">190</span>          }<a name="line.190"></a>
+<span class="sourceLineNo">191</span>        } catch (ImportTsv.TsvParser.BadTsvLineException | IllegalArgumentException<a name="line.191"></a>
+<span class="sourceLineNo">192</span>            | InvalidLabelException badLine) {<a name="line.192"></a>
+<span class="sourceLineNo">193</span>          if (skipBadLines) {<a name="line.193"></a>
+<span class="sourceLineNo">194</span>            System.err.println("Bad line." + badLine.getMessage());<a name="line.194"></a>
+<span class="sourceLineNo">195</span>            incrementBadLineCount(1);<a name="line.195"></a>
+<span class="sourceLineNo">196</span>            continue;<a name="line.196"></a>
+<span class="sourceLineNo">197</span>          }<a name="line.197"></a>
+<span class="sourceLineNo">198</span>          throw new IOException(badLine);<a name="line.198"></a>
+<span class="sourceLineNo">199</span>        }<a name="line.199"></a>
+<span class="sourceLineNo">200</span>      }<a name="line.200"></a>
+<span class="sourceLineNo">201</span>      context.setStatus("Read " + kvs.size() + " entries of " + kvs.getClass()<a name="line.201"></a>
+<span class="sourceLineNo">202</span>          + "(" + StringUtils.humanReadableInt(curSize) + ")");<a name="line.202"></a>
+<span class="sourceLineNo">203</span>      int index = 0;<a name="line.203"></a>
+<span class="sourceLineNo">204</span>      for (KeyValue kv : kvs) {<a name="line.204"></a>
+<span class="sourceLineNo">205</span>        context.write(rowKey, kv);<a name="line.205"></a>
+<span class="sourceLineNo">206</span>        if (++index &gt; 0 &amp;&amp; index % 100 == 0)<a name="line.206"></a>
+<span class="sourceLineNo">207</span>          context.setStatus("Wrote " + index + " key values.");<a name="line.207"></a>
+<span class="sourceLineNo">208</span>      }<a name="line.208"></a>
+<span class="sourceLineNo">209</span><a name="line.209"></a>
+<span class="sourceLineNo">210</span>      // if we have more entries to process<a name="line.210"></a>
+<span class="sourceLineNo">211</span>      if (iter.hasNext()) {<a name="line.211"></a>
+<span class="sourceLineNo">212</span>        // force flush because we cannot guarantee intra-row sorted order<a name="line.212"></a>
+<span class="sourceLineNo">213</span>        context.write(null, null);<a name="line.213"></a>
+<span class="sourceLineNo">214</span>      }<a name="line.214"></a>
+<span class="sourceLineNo">215</span>    }<a name="line.215"></a>
+<span class="sourceLineNo">216</span>  }<a name="line.216"></a>
+<span class="sourceLineNo">217</span>}<a name="line.217"></a>
 
 
 

http://git-wip-us.apache.org/repos/asf/hbase-site/blob/9577bd5e/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TsvImporterMapper.html
----------------------------------------------------------------------
diff --git a/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TsvImporterMapper.html b/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TsvImporterMapper.html
index 95af32f..de127b0 100644
--- a/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TsvImporterMapper.html
+++ b/apidocs/src-html/org/apache/hadoop/hbase/mapreduce/TsvImporterMapper.html
@@ -30,210 +30,211 @@
 <span class="sourceLineNo">022</span>import java.util.List;<a name="line.22"></a>
 <span class="sourceLineNo">023</span><a name="line.23"></a>
 <span class="sourceLineNo">024</span>import org.apache.hadoop.conf.Configuration;<a name="line.24"></a>
-<span class="sourceLineNo">025</span>import org.apache.hadoop.hbase.Cell;<a name="line.25"></a>
-<span class="sourceLineNo">026</span>import org.apache.hadoop.hbase.KeyValue;<a name="line.26"></a>
-<span class="sourceLineNo">027</span>import org.apache.hadoop.hbase.Tag;<a name="line.27"></a>
-<span class="sourceLineNo">028</span>import org.apache.hadoop.hbase.TagType;<a name="line.28"></a>
-<span class="sourceLineNo">029</span>import org.apache.hadoop.hbase.classification.InterfaceAudience;<a name="line.29"></a>
-<span class="sourceLineNo">030</span>import org.apache.hadoop.hbase.classification.InterfaceStability;<a name="line.30"></a>
-<span class="sourceLineNo">031</span>import org.apache.hadoop.hbase.client.Put;<a name="line.31"></a>
-<span class="sourceLineNo">032</span>import org.apache.hadoop.hbase.io.ImmutableBytesWritable;<a name="line.32"></a>
-<span class="sourceLineNo">033</span>import org.apache.hadoop.hbase.mapreduce.ImportTsv.TsvParser.BadTsvLineException;<a name="line.33"></a>
-<span class="sourceLineNo">034</span>import org.apache.hadoop.hbase.security.visibility.CellVisibility;<a name="line.34"></a>
-<span class="sourceLineNo">035</span>import org.apache.hadoop.hbase.security.visibility.InvalidLabelException;<a name="line.35"></a>
-<span class="sourceLineNo">036</span>import org.apache.hadoop.hbase.util.Base64;<a name="line.36"></a>
-<span class="sourceLineNo">037</span>import org.apache.hadoop.hbase.util.Bytes;<a name="line.37"></a>
-<span class="sourceLineNo">038</span>import org.apache.hadoop.io.LongWritable;<a name="line.38"></a>
-<span class="sourceLineNo">039</span>import org.apache.hadoop.io.Text;<a name="line.39"></a>
-<span class="sourceLineNo">040</span>import org.apache.hadoop.mapreduce.Counter;<a name="line.40"></a>
-<span class="sourceLineNo">041</span>import org.apache.hadoop.mapreduce.Mapper;<a name="line.41"></a>
-<span class="sourceLineNo">042</span><a name="line.42"></a>
-<span class="sourceLineNo">043</span>/**<a name="line.43"></a>
-<span class="sourceLineNo">044</span> * Write table content out to files in hdfs.<a name="line.44"></a>
-<span class="sourceLineNo">045</span> */<a name="line.45"></a>
-<span class="sourceLineNo">046</span>@InterfaceAudience.Public<a name="line.46"></a>
-<span class="sourceLineNo">047</span>@InterfaceStability.Stable<a name="line.47"></a>
-<span class="sourceLineNo">048</span>public class TsvImporterMapper<a name="line.48"></a>
-<span class="sourceLineNo">049</span>extends Mapper&lt;LongWritable, Text, ImmutableBytesWritable, Put&gt;<a name="line.49"></a>
-<span class="sourceLineNo">050</span>{<a name="line.50"></a>
-<span class="sourceLineNo">051</span><a name="line.51"></a>
-<span class="sourceLineNo">052</span>  /** Timestamp for all inserted rows */<a name="line.52"></a>
-<span class="sourceLineNo">053</span>  protected long ts;<a name="line.53"></a>
-<span class="sourceLineNo">054</span><a name="line.54"></a>
-<span class="sourceLineNo">055</span>  /** Column seperator */<a name="line.55"></a>
-<span class="sourceLineNo">056</span>  private String separator;<a name="line.56"></a>
-<span class="sourceLineNo">057</span><a name="line.57"></a>
-<span class="sourceLineNo">058</span>  /** Should skip bad lines */<a name="line.58"></a>
-<span class="sourceLineNo">059</span>  private boolean skipBadLines;<a name="line.59"></a>
-<span class="sourceLineNo">060</span>  private Counter badLineCount;<a name="line.60"></a>
-<span class="sourceLineNo">061</span>  private boolean logBadLines;<a name="line.61"></a>
-<span class="sourceLineNo">062</span><a name="line.62"></a>
-<span class="sourceLineNo">063</span>  protected ImportTsv.TsvParser parser;<a name="line.63"></a>
-<span class="sourceLineNo">064</span><a name="line.64"></a>
-<span class="sourceLineNo">065</span>  protected Configuration conf;<a name="line.65"></a>
-<span class="sourceLineNo">066</span><a name="line.66"></a>
-<span class="sourceLineNo">067</span>  protected String cellVisibilityExpr;<a name="line.67"></a>
-<span class="sourceLineNo">068</span><a name="line.68"></a>
-<span class="sourceLineNo">069</span>  protected long ttl;<a name="line.69"></a>
-<span class="sourceLineNo">070</span><a name="line.70"></a>
-<span class="sourceLineNo">071</span>  protected CellCreator kvCreator;<a name="line.71"></a>
-<span class="sourceLineNo">072</span><a name="line.72"></a>
-<span class="sourceLineNo">073</span>  private String hfileOutPath;<a name="line.73"></a>
-<span class="sourceLineNo">074</span><a name="line.74"></a>
-<span class="sourceLineNo">075</span>  /** List of cell tags */<a name="line.75"></a>
-<span class="sourceLineNo">076</span>  private List&lt;Tag&gt; tags;<a name="line.76"></a>
-<span class="sourceLineNo">077</span>  <a name="line.77"></a>
-<span class="sourceLineNo">078</span>  public long getTs() {<a name="line.78"></a>
-<span class="sourceLineNo">079</span>    return ts;<a name="line.79"></a>
-<span class="sourceLineNo">080</span>  }<a name="line.80"></a>
-<span class="sourceLineNo">081</span><a name="line.81"></a>
-<span class="sourceLineNo">082</span>  public boolean getSkipBadLines() {<a name="line.82"></a>
-<span class="sourceLineNo">083</span>    return skipBadLines;<a name="line.83"></a>
-<span class="sourceLineNo">084</span>  }<a name="line.84"></a>
-<span class="sourceLineNo">085</span><a name="line.85"></a>
-<span class="sourceLineNo">086</span>  public Counter getBadLineCount() {<a name="line.86"></a>
-<span class="sourceLineNo">087</span>    return badLineCount;<a name="line.87"></a>
-<span class="sourceLineNo">088</span>  }<a name="line.88"></a>
-<span class="sourceLineNo">089</span><a name="line.89"></a>
-<span class="sourceLineNo">090</span>  public void incrementBadLineCount(int count) {<a name="line.90"></a>
-<span class="sourceLineNo">091</span>    this.badLineCount.increment(count);<a name="line.91"></a>
-<span class="sourceLineNo">092</span>  }<a name="line.92"></a>
-<span class="sourceLineNo">093</span><a name="line.93"></a>
-<span class="sourceLineNo">094</span>  /**<a name="line.94"></a>
-<span class="sourceLineNo">095</span>   * Handles initializing this class with objects specific to it (i.e., the parser).<a name="line.95"></a>
-<span class="sourceLineNo">096</span>   * Common initialization that might be leveraged by a subsclass is done in<a name="line.96"></a>
-<span class="sourceLineNo">097</span>   * &lt;code&gt;doSetup&lt;/code&gt;. Hence a subclass may choose to override this method<a name="line.97"></a>
-<span class="sourceLineNo">098</span>   * and call &lt;code&gt;doSetup&lt;/code&gt; as well before handling it's own custom params.<a name="line.98"></a>
-<span class="sourceLineNo">099</span>   *<a name="line.99"></a>
-<span class="sourceLineNo">100</span>   * @param context<a name="line.100"></a>
-<span class="sourceLineNo">101</span>   */<a name="line.101"></a>
-<span class="sourceLineNo">102</span>  @Override<a name="line.102"></a>
-<span class="sourceLineNo">103</span>  protected void setup(Context context) {<a name="line.103"></a>
-<span class="sourceLineNo">104</span>    doSetup(context);<a name="line.104"></a>
-<span class="sourceLineNo">105</span><a name="line.105"></a>
-<span class="sourceLineNo">106</span>    conf = context.getConfiguration();<a name="line.106"></a>
-<span class="sourceLineNo">107</span>    parser = new ImportTsv.TsvParser(conf.get(ImportTsv.COLUMNS_CONF_KEY),<a name="line.107"></a>
-<span class="sourceLineNo">108</span>                           separator);<a name="line.108"></a>
-<span class="sourceLineNo">109</span>    if (parser.getRowKeyColumnIndex() == -1) {<a name="line.109"></a>
-<span class="sourceLineNo">110</span>      throw new RuntimeException("No row key column specified");<a name="line.110"></a>
-<span class="sourceLineNo">111</span>    }<a name="line.111"></a>
-<span class="sourceLineNo">112</span>    this.kvCreator = new CellCreator(conf);<a name="line.112"></a>
-<span class="sourceLineNo">113</span>    tags = new ArrayList&lt;Tag&gt;();<a name="line.113"></a>
-<span class="sourceLineNo">114</span>  }<a name="line.114"></a>
-<span class="sourceLineNo">115</span><a name="line.115"></a>
-<span class="sourceLineNo">116</span>  /**<a name="line.116"></a>
-<span class="sourceLineNo">117</span>   * Handles common parameter initialization that a subclass might want to leverage.<a name="line.117"></a>
-<span class="sourceLineNo">118</span>   * @param context<a name="line.118"></a>
-<span class="sourceLineNo">119</span>   */<a name="line.119"></a>
-<span class="sourceLineNo">120</span>  protected void doSetup(Context context) {<a name="line.120"></a>
-<span class="sourceLineNo">121</span>    Configuration conf = context.getConfiguration();<a name="line.121"></a>
-<span class="sourceLineNo">122</span><a name="line.122"></a>
-<span class="sourceLineNo">123</span>    // If a custom separator has been used,<a name="line.123"></a>
-<span class="sourceLineNo">124</span>    // decode it back from Base64 encoding.<a name="line.124"></a>
-<span class="sourceLineNo">125</span>    separator = conf.get(ImportTsv.SEPARATOR_CONF_KEY);<a name="line.125"></a>
-<span class="sourceLineNo">126</span>    if (separator == null) {<a name="line.126"></a>
-<span class="sourceLineNo">127</span>      separator = ImportTsv.DEFAULT_SEPARATOR;<a name="line.127"></a>
-<span class="sourceLineNo">128</span>    } else {<a name="line.128"></a>
-<span class="sourceLineNo">129</span>      separator = new String(Base64.decode(separator));<a name="line.129"></a>
-<span class="sourceLineNo">130</span>    }<a name="line.130"></a>
-<span class="sourceLineNo">131</span>    // Should never get 0 as we are setting this to a valid value in job<a name="line.131"></a>
-<span class="sourceLineNo">132</span>    // configuration.<a name="line.132"></a>
-<span class="sourceLineNo">133</span>    ts = conf.getLong(ImportTsv.TIMESTAMP_CONF_KEY, 0);<a name="line.133"></a>
-<span class="sourceLineNo">134</span><a name="line.134"></a>
-<span class="sourceLineNo">135</span>    skipBadLines = context.getConfiguration().getBoolean(<a name="line.135"></a>
-<span class="sourceLineNo">136</span>        ImportTsv.SKIP_LINES_CONF_KEY, true);<a name="line.136"></a>
-<span class="sourceLineNo">137</span>    badLineCount = context.getCounter("ImportTsv", "Bad Lines");<a name="line.137"></a>
-<span class="sourceLineNo">138</span>    logBadLines = context.getConfiguration().getBoolean(ImportTsv.LOG_BAD_LINES_CONF_KEY, false);<a name="line.138"></a>
-<span class="sourceLineNo">139</span>    hfileOutPath = conf.get(ImportTsv.BULK_OUTPUT_CONF_KEY);<a name="line.139"></a>
-<span class="sourceLineNo">140</span>  }<a name="line.140"></a>
-<span class="sourceLineNo">141</span><a name="line.141"></a>
-<span class="sourceLineNo">142</span>  /**<a name="line.142"></a>
-<span class="sourceLineNo">143</span>   * Convert a line of TSV text into an HBase table row.<a name="line.143"></a>
-<span class="sourceLineNo">144</span>   */<a name="line.144"></a>
-<span class="sourceLineNo">145</span>  @Override<a name="line.145"></a>
-<span class="sourceLineNo">146</span>  public void map(LongWritable offset, Text value,<a name="line.146"></a>
-<span class="sourceLineNo">147</span>    Context context)<a name="line.147"></a>
-<span class="sourceLineNo">148</span>  throws IOException {<a name="line.148"></a>
-<span class="sourceLineNo">149</span>    byte[] lineBytes = value.getBytes();<a name="line.149"></a>
-<span class="sourceLineNo">150</span><a name="line.150"></a>
-<span class="sourceLineNo">151</span>    try {<a name="line.151"></a>
-<span class="sourceLineNo">152</span>      ImportTsv.TsvParser.ParsedLine parsed = parser.parse(<a name="line.152"></a>
-<span class="sourceLineNo">153</span>          lineBytes, value.getLength());<a name="line.153"></a>
-<span class="sourceLineNo">154</span>      ImmutableBytesWritable rowKey =<a name="line.154"></a>
-<span class="sourceLineNo">155</span>        new ImmutableBytesWritable(lineBytes,<a name="line.155"></a>
-<span class="sourceLineNo">156</span>            parsed.getRowKeyOffset(),<a name="line.156"></a>
-<span class="sourceLineNo">157</span>            parsed.getRowKeyLength());<a name="line.157"></a>
-<span class="sourceLineNo">158</span>      // Retrieve timestamp if exists<a name="line.158"></a>
-<span class="sourceLineNo">159</span>      ts = parsed.getTimestamp(ts);<a name="line.159"></a>
-<span class="sourceLineNo">160</span>      cellVisibilityExpr = parsed.getCellVisibility();<a name="line.160"></a>
-<span class="sourceLineNo">161</span>      ttl = parsed.getCellTTL();<a name="line.161"></a>
-<span class="sourceLineNo">162</span><a name="line.162"></a>
-<span class="sourceLineNo">163</span>      // create tags for the parsed line<a name="line.163"></a>
-<span class="sourceLineNo">164</span>      if (hfileOutPath != null) {<a name="line.164"></a>
-<span class="sourceLineNo">165</span>        tags.clear();<a name="line.165"></a>
-<span class="sourceLineNo">166</span>        if (cellVisibilityExpr != null) {<a name="line.166"></a>
-<span class="sourceLineNo">167</span>          tags.addAll(kvCreator.getVisibilityExpressionResolver().createVisibilityExpTags(<a name="line.167"></a>
-<span class="sourceLineNo">168</span>            cellVisibilityExpr));<a name="line.168"></a>
-<span class="sourceLineNo">169</span>        }<a name="line.169"></a>
-<span class="sourceLineNo">170</span>        // Add TTL directly to the KV so we can vary them when packing more than one KV<a name="line.170"></a>
-<span class="sourceLineNo">171</span>        // into puts<a name="line.171"></a>
-<span class="sourceLineNo">172</span>        if (ttl &gt; 0) {<a name="line.172"></a>
-<span class="sourceLineNo">173</span>          tags.add(new Tag(TagType.TTL_TAG_TYPE, Bytes.toBytes(ttl)));<a name="line.173"></a>
-<span class="sourceLineNo">174</span>        }<a name="line.174"></a>
-<span class="sourceLineNo">175</span>      }<a name="line.175"></a>
-<span class="sourceLineNo">176</span>      Put put = new Put(rowKey.copyBytes());<a name="line.176"></a>
-<span class="sourceLineNo">177</span>      for (int i = 0; i &lt; parsed.getColumnCount(); i++) {<a name="line.177"></a>
-<span class="sourceLineNo">178</span>        if (i == parser.getRowKeyColumnIndex() || i == parser.getTimestampKeyColumnIndex()<a name="line.178"></a>
-<span class="sourceLineNo">179</span>            || i == parser.getAttributesKeyColumnIndex() || i == parser.getCellVisibilityColumnIndex()<a name="line.179"></a>
-<span class="sourceLineNo">180</span>            || i == parser.getCellTTLColumnIndex()) {<a name="line.180"></a>
-<span class="sourceLineNo">181</span>          continue;<a name="line.181"></a>
-<span class="sourceLineNo">182</span>        }<a name="line.182"></a>
-<span class="sourceLineNo">183</span>        populatePut(lineBytes, parsed, put, i);<a name="line.183"></a>
-<span class="sourceLineNo">184</span>      }<a name="line.184"></a>
-<span class="sourceLineNo">185</span>      context.write(rowKey, put);<a name="line.185"></a>
-<span class="sourceLineNo">186</span>    } catch (ImportTsv.TsvParser.BadTsvLineException | IllegalArgumentException<a name="line.186"></a>
-<span class="sourceLineNo">187</span>        | InvalidLabelException badLine) {<a name="line.187"></a>
-<span class="sourceLineNo">188</span>      if (logBadLines) {<a name="line.188"></a>
-<span class="sourceLineNo">189</span>        System.err.println(value);<a name="line.189"></a>
-<span class="sourceLineNo">190</span>      }<a name="line.190"></a>
-<span class="sourceLineNo">191</span>      System.err.println("Bad line at offset: " + offset.get() + ":\n" + badLine.getMessage());<a name="line.191"></a>
-<span class="sourceLineNo">192</span>      if (skipBadLines) {<a name="line.192"></a>
-<span class="sourceLineNo">193</span>        incrementBadLineCount(1);<a name="line.193"></a>
-<span class="sourceLineNo">194</span>        return;<a name="line.194"></a>
-<span class="sourceLineNo">195</span>      }<a name="line.195"></a>
-<span class="sourceLineNo">196</span>      throw new IOException(badLine);<a name="line.196"></a>
-<span class="sourceLineNo">197</span>    } catch (InterruptedException e) {<a name="line.197"></a>
-<span class="sourceLineNo">198</span>      e.printStackTrace();<a name="line.198"></a>
-<span class="sourceLineNo">199</span>    }<a name="line.199"></a>
-<span class="sourceLineNo">200</span>  }<a name="line.200"></a>
-<span class="sourceLineNo">201</span><a name="line.201"></a>
-<span class="sourceLineNo">202</span>  protected void populatePut(byte[] lineBytes, ImportTsv.TsvParser.ParsedLine parsed, Put put,<a name="line.202"></a>
-<span class="sourceLineNo">203</span>      int i) throws BadTsvLineException, IOException {<a name="line.203"></a>
-<span class="sourceLineNo">204</span>    Cell cell = null;<a name="line.204"></a>
-<span class="sourceLineNo">205</span>    if (hfileOutPath == null) {<a name="line.205"></a>
-<span class="sourceLineNo">206</span>      cell = new KeyValue(lineBytes, parsed.getRowKeyOffset(), parsed.getRowKeyLength(),<a name="line.206"></a>
-<span class="sourceLineNo">207</span>          parser.getFamily(i), 0, parser.getFamily(i).length, parser.getQualifier(i), 0,<a name="line.207"></a>
-<span class="sourceLineNo">208</span>          parser.getQualifier(i).length, ts, KeyValue.Type.Put, lineBytes,<a name="line.208"></a>
-<span class="sourceLineNo">209</span>          parsed.getColumnOffset(i), parsed.getColumnLength(i));<a name="line.209"></a>
-<span class="sourceLineNo">210</span>      if (cellVisibilityExpr != null) {<a name="line.210"></a>
-<span class="sourceLineNo">211</span>        // We won't be validating the expression here. The Visibility CP will do<a name="line.211"></a>
-<span class="sourceLineNo">212</span>        // the validation<a name="line.212"></a>
-<span class="sourceLineNo">213</span>        put.setCellVisibility(new CellVisibility(cellVisibilityExpr));<a name="line.213"></a>
-<span class="sourceLineNo">214</span>      }<a name="line.214"></a>
-<span class="sourceLineNo">215</span>      if (ttl &gt; 0) {<a name="line.215"></a>
-<span class="sourceLineNo">216</span>        put.setTTL(ttl);<a name="line.216"></a>
-<span class="sourceLineNo">217</span>      }<a name="line.217"></a>
-<span class="sourceLineNo">218</span>    } else {<a name="line.218"></a>
-<span class="sourceLineNo">219</span>      // Creating the KV which needs to be directly written to HFiles. Using the Facade<a name="line.219"></a>
-<span class="sourceLineNo">220</span>      // KVCreator for creation of kvs.<a name="line.220"></a>
-<span class="sourceLineNo">221</span>      cell = this.kvCreator.create(lineBytes, parsed.getRowKeyOffset(), parsed.getRowKeyLength(),<a name="line.221"></a>
-<span class="sourceLineNo">222</span>          parser.getFamily(i), 0, parser.getFamily(i).length, parser.getQualifier(i), 0,<a name="line.222"></a>
-<span class="sourceLineNo">223</span>          parser.getQualifier(i).length, ts, lineBytes, parsed.getColumnOffset(i),<a name="line.223"></a>
-<span class="sourceLineNo">224</span>          parsed.getColumnLength(i), tags);<a name="line.224"></a>
-<span class="sourceLineNo">225</span>    }<a name="line.225"></a>
-<span class="sourceLineNo">226</span>    put.add(cell);<a name="line.226"></a>
-<span class="sourceLineNo">227</span>  }<a name="line.227"></a>
-<span class="sourceLineNo">228</span>}<a name="line.228"></a>
+<span class="sourceLineNo">025</span>import org.apache.hadoop.hbase.ArrayBackedTag;<a name="line.25"></a>
+<span class="sourceLineNo">026</span>import org.apache.hadoop.hbase.Cell;<a name="line.26"></a>
+<span class="sourceLineNo">027</span>import org.apache.hadoop.hbase.KeyValue;<a name="line.27"></a>
+<span class="sourceLineNo">028</span>import org.apache.hadoop.hbase.Tag;<a name="line.28"></a>
+<span class="sourceLineNo">029</span>import org.apache.hadoop.hbase.TagType;<a name="line.29"></a>
+<span class="sourceLineNo">030</span>import org.apache.hadoop.hbase.classification.InterfaceAudience;<a name="line.30"></a>
+<span class="sourceLineNo">031</span>import org.apache.hadoop.hbase.classification.InterfaceStability;<a name="line.31"></a>
+<span class="sourceLineNo">032</span>import org.apache.hadoop.hbase.client.Put;<a name="line.32"></a>
+<span class="sourceLineNo">033</span>import org.apache.hadoop.hbase.io.ImmutableBytesWritable;<a name="line.33"></a>
+<span class="sourceLineNo">034</span>import org.apache.hadoop.hbase.mapreduce.ImportTsv.TsvParser.BadTsvLineException;<a name="line.34"></a>
+<span class="sourceLineNo">035</span>import org.apache.hadoop.hbase.security.visibility.CellVisibility;<a name="line.35"></a>
+<span class="sourceLineNo">036</span>import org.apache.hadoop.hbase.security.visibility.InvalidLabelException;<a name="line.36"></a>
+<span class="sourceLineNo">037</span>import org.apache.hadoop.hbase.util.Base64;<a name="line.37"></a>
+<span class="sourceLineNo">038</span>import org.apache.hadoop.hbase.util.Bytes;<a name="line.38"></a>
+<span class="sourceLineNo">039</span>import org.apache.hadoop.io.LongWritable;<a name="line.39"></a>
+<span class="sourceLineNo">040</span>import org.apache.hadoop.io.Text;<a name="line.40"></a>
+<span class="sourceLineNo">041</span>import org.apache.hadoop.mapreduce.Counter;<a name="line.41"></a>
+<span class="sourceLineNo">042</span>import org.apache.hadoop.mapreduce.Mapper;<a name="line.42"></a>
+<span class="sourceLineNo">043</span><a name="line.43"></a>
+<span class="sourceLineNo">044</span>/**<a name="line.44"></a>
+<span class="sourceLineNo">045</span> * Write table content out to files in hdfs.<a name="line.45"></a>
+<span class="sourceLineNo">046</span> */<a name="line.46"></a>
+<span class="sourceLineNo">047</span>@InterfaceAudience.Public<a name="line.47"></a>
+<span class="sourceLineNo">048</span>@InterfaceStability.Stable<a name="line.48"></a>
+<span class="sourceLineNo">049</span>public class TsvImporterMapper<a name="line.49"></a>
+<span class="sourceLineNo">050</span>extends Mapper&lt;LongWritable, Text, ImmutableBytesWritable, Put&gt;<a name="line.50"></a>
+<span class="sourceLineNo">051</span>{<a name="line.51"></a>
+<span class="sourceLineNo">052</span><a name="line.52"></a>
+<span class="sourceLineNo">053</span>  /** Timestamp for all inserted rows */<a name="line.53"></a>
+<span class="sourceLineNo">054</span>  protected long ts;<a name="line.54"></a>
+<span class="sourceLineNo">055</span><a name="line.55"></a>
+<span class="sourceLineNo">056</span>  /** Column seperator */<a name="line.56"></a>
+<span class="sourceLineNo">057</span>  private String separator;<a name="line.57"></a>
+<span class="sourceLineNo">058</span><a name="line.58"></a>
+<span class="sourceLineNo">059</span>  /** Should skip bad lines */<a name="line.59"></a>
+<span class="sourceLineNo">060</span>  private boolean skipBadLines;<a name="line.60"></a>
+<span class="sourceLineNo">061</span>  private Counter badLineCount;<a name="line.61"></a>
+<span class="sourceLineNo">062</span>  private boolean logBadLines;<a name="line.62"></a>
+<span class="sourceLineNo">063</span><a name="line.63"></a>
+<span class="sourceLineNo">064</span>  protected ImportTsv.TsvParser parser;<a name="line.64"></a>
+<span class="sourceLineNo">065</span><a name="line.65"></a>
+<span class="sourceLineNo">066</span>  protected Configuration conf;<a name="line.66"></a>
+<span class="sourceLineNo">067</span><a name="line.67"></a>
+<span class="sourceLineNo">068</span>  protected String cellVisibilityExpr;<a name="line.68"></a>
+<span class="sourceLineNo">069</span><a name="line.69"></a>
+<span class="sourceLineNo">070</span>  protected long ttl;<a name="line.70"></a>
+<span class="sourceLineNo">071</span><a name="line.71"></a>
+<span class="sourceLineNo">072</span>  protected CellCreator kvCreator;<a name="line.72"></a>
+<span class="sourceLineNo">073</span><a name="line.73"></a>
+<span class="sourceLineNo">074</span>  private String hfileOutPath;<a name="line.74"></a>
+<span class="sourceLineNo">075</span><a name="line.75"></a>
+<span class="sourceLineNo">076</span>  /** List of cell tags */<a name="line.76"></a>
+<span class="sourceLineNo">077</span>  private List&lt;Tag&gt; tags;<a name="line.77"></a>
+<span class="sourceLineNo">078</span>  <a name="line.78"></a>
+<span class="sourceLineNo">079</span>  public long getTs() {<a name="line.79"></a>
+<span class="sourceLineNo">080</span>    return ts;<a name="line.80"></a>
+<span class="sourceLineNo">081</span>  }<a name="line.81"></a>
+<span class="sourceLineNo">082</span><a name="line.82"></a>
+<span class="sourceLineNo">083</span>  public boolean getSkipBadLines() {<a name="line.83"></a>
+<span class="sourceLineNo">084</span>    return skipBadLines;<a name="line.84"></a>
+<span class="sourceLineNo">085</span>  }<a name="line.85"></a>
+<span class="sourceLineNo">086</span><a name="line.86"></a>
+<span class="sourceLineNo">087</span>  public Counter getBadLineCount() {<a name="line.87"></a>
+<span class="sourceLineNo">088</span>    return badLineCount;<a name="line.88"></a>
+<span class="sourceLineNo">089</span>  }<a name="line.89"></a>
+<span class="sourceLineNo">090</span><a name="line.90"></a>
+<span class="sourceLineNo">091</span>  public void incrementBadLineCount(int count) {<a name="line.91"></a>
+<span class="sourceLineNo">092</span>    this.badLineCount.increment(count);<a name="line.92"></a>
+<span class="sourceLineNo">093</span>  }<a name="line.93"></a>
+<span class="sourceLineNo">094</span><a name="line.94"></a>
+<span class="sourceLineNo">095</span>  /**<a name="line.95"></a>
+<span class="sourceLineNo">096</span>   * Handles initializing this class with objects specific to it (i.e., the parser).<a name="line.96"></a>
+<span class="sourceLineNo">097</span>   * Common initialization that might be leveraged by a subsclass is done in<a name="line.97"></a>
+<span class="sourceLineNo">098</span>   * &lt;code&gt;doSetup&lt;/code&gt;. Hence a subclass may choose to override this method<a name="line.98"></a>
+<span class="sourceLineNo">099</span>   * and call &lt;code&gt;doSetup&lt;/code&gt; as well before handling it's own custom params.<a name="line.99"></a>
+<span class="sourceLineNo">100</span>   *<a name="line.100"></a>
+<span class="sourceLineNo">101</span>   * @param context<a name="line.101"></a>
+<span class="sourceLineNo">102</span>   */<a name="line.102"></a>
+<span class="sourceLineNo">103</span>  @Override<a name="line.103"></a>
+<span class="sourceLineNo">104</span>  protected void setup(Context context) {<a name="line.104"></a>
+<span class="sourceLineNo">105</span>    doSetup(context);<a name="line.105"></a>
+<span class="sourceLineNo">106</span><a name="line.106"></a>
+<span class="sourceLineNo">107</span>    conf = context.getConfiguration();<a name="line.107"></a>
+<span class="sourceLineNo">108</span>    parser = new ImportTsv.TsvParser(conf.get(ImportTsv.COLUMNS_CONF_KEY),<a name="line.108"></a>
+<span class="sourceLineNo">109</span>                           separator);<a name="line.109"></a>
+<span class="sourceLineNo">110</span>    if (parser.getRowKeyColumnIndex() == -1) {<a name="line.110"></a>
+<span class="sourceLineNo">111</span>      throw new RuntimeException("No row key column specified");<a name="line.111"></a>
+<span class="sourceLineNo">112</span>    }<a name="line.112"></a>
+<span class="sourceLineNo">113</span>    this.kvCreator = new CellCreator(conf);<a name="line.113"></a>
+<span class="sourceLineNo">114</span>    tags = new ArrayList&lt;Tag&gt;();<a name="line.114"></a>
+<span class="sourceLineNo">115</span>  }<a name="line.115"></a>
+<span class="sourceLineNo">116</span><a name="line.116"></a>
+<span class="sourceLineNo">117</span>  /**<a name="line.117"></a>
+<span class="sourceLineNo">118</span>   * Handles common parameter initialization that a subclass might want to leverage.<a name="line.118"></a>
+<span class="sourceLineNo">119</span>   * @param context<a name="line.119"></a>
+<span class="sourceLineNo">120</span>   */<a name="line.120"></a>
+<span class="sourceLineNo">121</span>  protected void doSetup(Context context) {<a name="line.121"></a>
+<span class="sourceLineNo">122</span>    Configuration conf = context.getConfiguration();<a name="line.122"></a>
+<span class="sourceLineNo">123</span><a name="line.123"></a>
+<span class="sourceLineNo">124</span>    // If a custom separator has been used,<a name="line.124"></a>
+<span class="sourceLineNo">125</span>    // decode it back from Base64 encoding.<a name="line.125"></a>
+<span class="sourceLineNo">126</span>    separator = conf.get(ImportTsv.SEPARATOR_CONF_KEY);<a name="line.126"></a>
+<span class="sourceLineNo">127</span>    if (separator == null) {<a name="line.127"></a>
+<span class="sourceLineNo">128</span>      separator = ImportTsv.DEFAULT_SEPARATOR;<a name="line.128"></a>
+<span class="sourceLineNo">129</span>    } else {<a name="line.129"></a>
+<span class="sourceLineNo">130</span>      separator = new String(Base64.decode(separator));<a name="line.130"></a>
+<span class="sourceLineNo">131</span>    }<a name="line.131"></a>
+<span class="sourceLineNo">132</span>    // Should never get 0 as we are setting this to a valid value in job<a name="line.132"></a>
+<span class="sourceLineNo">133</span>    // configuration.<a name="line.133"></a>
+<span class="sourceLineNo">134</span>    ts = conf.getLong(ImportTsv.TIMESTAMP_CONF_KEY, 0);<a name="line.134"></a>
+<span class="sourceLineNo">135</span><a name="line.135"></a>
+<span class="sourceLineNo">136</span>    skipBadLines = context.getConfiguration().getBoolean(<a name="line.136"></a>
+<span class="sourceLineNo">137</span>        ImportTsv.SKIP_LINES_CONF_KEY, true);<a name="line.137"></a>
+<span class="sourceLineNo">138</span>    badLineCount = context.getCounter("ImportTsv", "Bad Lines");<a name="line.138"></a>
+<span class="sourceLineNo">139</span>    logBadLines = context.getConfiguration().getBoolean(ImportTsv.LOG_BAD_LINES_CONF_KEY, false);<a name="line.139"></a>
+<span class="sourceLineNo">140</span>    hfileOutPath = conf.get(ImportTsv.BULK_OUTPUT_CONF_KEY);<a name="line.140"></a>
+<span class="sourceLineNo">141</span>  }<a name="line.141"></a>
+<span class="sourceLineNo">142</span><a name="line.142"></a>
+<span class="sourceLineNo">143</span>  /**<a name="line.143"></a>
+<span class="sourceLineNo">144</span>   * Convert a line of TSV text into an HBase table row.<a name="line.144"></a>
+<span class="sourceLineNo">145</span>   */<a name="line.145"></a>
+<span class="sourceLineNo">146</span>  @Override<a name="line.146"></a>
+<span class="sourceLineNo">147</span>  public void map(LongWritable offset, Text value,<a name="line.147"></a>
+<span class="sourceLineNo">148</span>    Context context)<a name="line.148"></a>
+<span class="sourceLineNo">149</span>  throws IOException {<a name="line.149"></a>
+<span class="sourceLineNo">150</span>    byte[] lineBytes = value.getBytes();<a name="line.150"></a>
+<span class="sourceLineNo">151</span><a name="line.151"></a>
+<span class="sourceLineNo">152</span>    try {<a name="line.152"></a>
+<span class="sourceLineNo">153</span>      ImportTsv.TsvParser.ParsedLine parsed = parser.parse(<a name="line.153"></a>
+<span class="sourceLineNo">154</span>          lineBytes, value.getLength());<a name="line.154"></a>
+<span class="sourceLineNo">155</span>      ImmutableBytesWritable rowKey =<a name="line.155"></a>
+<span class="sourceLineNo">156</span>        new ImmutableBytesWritable(lineBytes,<a name="line.156"></a>
+<span class="sourceLineNo">157</span>            parsed.getRowKeyOffset(),<a name="line.157"></a>
+<span class="sourceLineNo">158</span>            parsed.getRowKeyLength());<a name="line.158"></a>
+<span class="sourceLineNo">159</span>      // Retrieve timestamp if exists<a name="line.159"></a>
+<span class="sourceLineNo">160</span>      ts = parsed.getTimestamp(ts);<a name="line.160"></a>
+<span class="sourceLineNo">161</span>      cellVisibilityExpr = parsed.getCellVisibility();<a name="line.161"></a>
+<span class="sourceLineNo">162</span>      ttl = parsed.getCellTTL();<a name="line.162"></a>
+<span class="sourceLineNo">163</span><a name="line.163"></a>
+<span class="sourceLineNo">164</span>      // create tags for the parsed line<a name="line.164"></a>
+<span class="sourceLineNo">165</span>      if (hfileOutPath != null) {<a name="line.165"></a>
+<span class="sourceLineNo">166</span>        tags.clear();<a name="line.166"></a>
+<span class="sourceLineNo">167</span>        if (cellVisibilityExpr != null) {<a name="line.167"></a>
+<span class="sourceLineNo">168</span>          tags.addAll(kvCreator.getVisibilityExpressionResolver().createVisibilityExpTags(<a name="line.168"></a>
+<span class="sourceLineNo">169</span>            cellVisibilityExpr));<a name="line.169"></a>
+<span class="sourceLineNo">170</span>        }<a name="line.170"></a>
+<span class="sourceLineNo">171</span>        // Add TTL directly to the KV so we can vary them when packing more than one KV<a name="line.171"></a>
+<span class="sourceLineNo">172</span>        // into puts<a name="line.172"></a>
+<span class="sourceLineNo">173</span>        if (ttl &gt; 0) {<a name="line.173"></a>
+<span class="sourceLineNo">174</span>          tags.add(new ArrayBackedTag(TagType.TTL_TAG_TYPE, Bytes.toBytes(ttl)));<a name="line.174"></a>
+<span class="sourceLineNo">175</span>        }<a name="line.175"></a>
+<span class="sourceLineNo">176</span>      }<a name="line.176"></a>
+<span class="sourceLineNo">177</span>      Put put = new Put(rowKey.copyBytes());<a name="line.177"></a>
+<span class="sourceLineNo">178</span>      for (int i = 0; i &lt; parsed.getColumnCount(); i++) {<a name="line.178"></a>
+<span class="sourceLineNo">179</span>        if (i == parser.getRowKeyColumnIndex() || i == parser.getTimestampKeyColumnIndex()<a name="line.179"></a>
+<span class="sourceLineNo">180</span>            || i == parser.getAttributesKeyColumnIndex() || i == parser.getCellVisibilityColumnIndex()<a name="line.180"></a>
+<span class="sourceLineNo">181</span>            || i == parser.getCellTTLColumnIndex()) {<a name="line.181"></a>
+<span class="sourceLineNo">182</span>          continue;<a name="line.182"></a>
+<span class="sourceLineNo">183</span>        }<a name="line.183"></a>
+<span class="sourceLineNo">184</span>        populatePut(lineBytes, parsed, put, i);<a name="line.184"></a>
+<span class="sourceLineNo">185</span>      }<a name="line.185"></a>
+<span class="sourceLineNo">186</span>      context.write(rowKey, put);<a name="line.186"></a>
+<span class="sourceLineNo">187</span>    } catch (ImportTsv.TsvParser.BadTsvLineException | IllegalArgumentException<a name="line.187"></a>
+<span class="sourceLineNo">188</span>        | InvalidLabelException badLine) {<a name="line.188"></a>
+<span class="sourceLineNo">189</span>      if (logBadLines) {<a name="line.189"></a>
+<span class="sourceLineNo">190</span>        System.err.println(value);<a name="line.190"></a>
+<span class="sourceLineNo">191</span>      }<a name="line.191"></a>
+<span class="sourceLineNo">192</span>      System.err.println("Bad line at offset: " + offset.get() + ":\n" + badLine.getMessage());<a name="line.192"></a>
+<span class="sourceLineNo">193</span>      if (skipBadLines) {<a name="line.193"></a>
+<span class="sourceLineNo">194</span>        incrementBadLineCount(1);<a name="line.194"></a>
+<span class="sourceLineNo">195</span>        return;<a name="line.195"></a>
+<span class="sourceLineNo">196</span>      }<a name="line.196"></a>
+<span class="sourceLineNo">197</span>      throw new IOException(badLine);<a name="line.197"></a>
+<span class="sourceLineNo">198</span>    } catch (InterruptedException e) {<a name="line.198"></a>
+<span class="sourceLineNo">199</span>      e.printStackTrace();<a name="line.199"></a>
+<span class="sourceLineNo">200</span>    }<a name="line.200"></a>
+<span class="sourceLineNo">201</span>  }<a name="line.201"></a>
+<span class="sourceLineNo">202</span><a name="line.202"></a>
+<span class="sourceLineNo">203</span>  protected void populatePut(byte[] lineBytes, ImportTsv.TsvParser.ParsedLine parsed, Put put,<a name="line.203"></a>
+<span class="sourceLineNo">204</span>      int i) throws BadTsvLineException, IOException {<a name="line.204"></a>
+<span class="sourceLineNo">205</span>    Cell cell = null;<a name="line.205"></a>
+<span class="sourceLineNo">206</span>    if (hfileOutPath == null) {<a name="line.206"></a>
+<span class="sourceLineNo">207</span>      cell = new KeyValue(lineBytes, parsed.getRowKeyOffset(), parsed.getRowKeyLength(),<a name="line.207"></a>
+<span class="sourceLineNo">208</span>          parser.getFamily(i), 0, parser.getFamily(i).length, parser.getQualifier(i), 0,<a name="line.208"></a>
+<span class="sourceLineNo">209</span>          parser.getQualifier(i).length, ts, KeyValue.Type.Put, lineBytes,<a name="line.209"></a>
+<span class="sourceLineNo">210</span>          parsed.getColumnOffset(i), parsed.getColumnLength(i));<a name="line.210"></a>
+<span class="sourceLineNo">211</span>      if (cellVisibilityExpr != null) {<a name="line.211"></a>
+<span class="sourceLineNo">212</span>        // We won't be validating the expression here. The Visibility CP will do<a name="line.212"></a>
+<span class="sourceLineNo">213</span>        // the validation<a name="line.213"></a>
+<span class="sourceLineNo">214</span>        put.setCellVisibility(new CellVisibility(cellVisibilityExpr));<a name="line.214"></a>
+<span class="sourceLineNo">215</span>      }<a name="line.215"></a>
+<span class="sourceLineNo">216</span>      if (ttl &gt; 0) {<a name="line.216"></a>
+<span class="sourceLineNo">217</span>        put.setTTL(ttl);<a name="line.217"></a>
+<span class="sourceLineNo">218</span>      }<a name="line.218"></a>
+<span class="sourceLineNo">219</span>    } else {<a name="line.219"></a>
+<span class="sourceLineNo">220</span>      // Creating the KV which needs to be directly written to HFiles. Using the Facade<a name="line.220"></a>
+<span class="sourceLineNo">221</span>      // KVCreator for creation of kvs.<a name="line.221"></a>
+<span class="sourceLineNo">222</span>      cell = this.kvCreator.create(lineBytes, parsed.getRowKeyOffset(), parsed.getRowKeyLength(),<a name="line.222"></a>
+<span class="sourceLineNo">223</span>          parser.getFamily(i), 0, parser.getFamily(i).length, parser.getQualifier(i), 0,<a name="line.223"></a>
+<span class="sourceLineNo">224</span>          parser.getQualifier(i).length, ts, lineBytes, parsed.getColumnOffset(i),<a name="line.224"></a>
+<span class="sourceLineNo">225</span>          parsed.getColumnLength(i), tags);<a name="line.225"></a>
+<span class="sourceLineNo">226</span>    }<a name="line.226"></a>
+<span class="sourceLineNo">227</span>    put.add(cell);<a name="line.227"></a>
+<span class="sourceLineNo">228</span>  }<a name="line.228"></a>
+<span class="sourceLineNo">229</span>}<a name="line.229"></a>