HBASE-9345 Add support for specifying filters in scan (Virag Kothari)

hbase-server/src/main/java/org/apache/hadoop/hbase/rest/Constants.java

@@ -68,6 +68,8 @@ public interface Constants {
   String SCAN_BATCH_SIZE = "batchsize";
   String SCAN_LIMIT = "limit";
   String SCAN_FETCH_SIZE = "hbase.rest.scan.fetchsize";
+  String SCAN_FILTER = "filter"; 
+  String CUSTOM_FILTERS = "hbase.rest.custom.filters"; 
 
   String ROW_KEYS_PARAM_NAME = "row";
   /** If this query parameter is present when processing row or scanner resources,

hbase-server/src/main/java/org/apache/hadoop/hbase/rest/RESTServlet.java

@@ -31,6 +31,7 @@ import org.apache.hadoop.hbase.client.HBaseAdmin;
 import org.apache.hadoop.hbase.client.HConnection;
 import org.apache.hadoop.hbase.client.HConnectionManager;
 import org.apache.hadoop.hbase.client.HTableInterface;
+import org.apache.hadoop.hbase.filter.ParseFilter;
 import org.apache.hadoop.hbase.security.User;
 import org.apache.hadoop.hbase.security.UserProvider;
 import org.apache.hadoop.hbase.util.EnvironmentEdgeManager;
@@ -186,6 +187,7 @@ public class RESTServlet implements Constants {
 
     this.realUser = realUser;
     this.conf = conf;
+    registerCustomFilter(conf);
   }
 
   /**
@@ -255,4 +257,19 @@ public class RESTServlet implements Constants {
     }
     return connInfo;
   }
+  
+  private void registerCustomFilter(Configuration conf) {
+    String[] filterList = conf.getStrings(Constants.CUSTOM_FILTERS);
+    if (filterList != null) {
+      for (String filterClass : filterList) {
+        String[] filterPart = filterClass.split(":");
+        if (filterPart.length != 2) {
+          LOG.warn(
+            "Invalid filter specification " + filterClass + " - skipping");
+        } else {
+          ParseFilter.registerFilter(filterPart[0], filterPart[1]);
+        }
+      }
+    }
+  }
 }

hbase-server/src/main/java/org/apache/hadoop/hbase/rest/TableResource.java

@@ -38,6 +38,8 @@ import org.apache.hadoop.classification.InterfaceAudience;
 import org.apache.hadoop.hbase.client.HTableInterface;
 import org.apache.hadoop.hbase.client.Scan;
 import org.apache.hadoop.hbase.filter.Filter;
+import org.apache.hadoop.hbase.filter.FilterList;
+import org.apache.hadoop.hbase.filter.ParseFilter;
 import org.apache.hadoop.hbase.filter.PrefixFilter;
 import org.apache.hadoop.hbase.util.Bytes;
 
@@ -130,7 +132,8 @@ public class TableResource extends ResourceBase {
       @DefaultValue("-1") @QueryParam(Constants.SCAN_BATCH_SIZE) int batchSize,
       @DefaultValue("0") @QueryParam(Constants.SCAN_START_TIME) long startTime,
       @DefaultValue(Long.MAX_VALUE + "") @QueryParam(Constants.SCAN_END_TIME) long endTime,
-      @DefaultValue("true") @QueryParam(Constants.SCAN_BATCH_SIZE) boolean cacheBlocks) {
+      @DefaultValue("true") @QueryParam(Constants.SCAN_BATCH_SIZE) boolean cacheBlocks, 
+      @DefaultValue("") @QueryParam(Constants.SCAN_FILTER) String filters) {
     try {
       Filter filter = null;
       if (scanSpec.indexOf('*') > 0) {
@@ -164,7 +167,20 @@ public class TableResource extends ResourceBase {
           tableScan.addFamily(Bytes.toBytes(familysplit[0]));
         }
       }
-      if (filter != null) {
+      FilterList filterList = null;
+      if (StringUtils.isNotEmpty(filters)) {
+          ParseFilter pf = new ParseFilter();
+          Filter filterParam = pf.parseFilterString(filters);
+          if (filter != null) {
+            filterList = new FilterList(filter, filterParam);
+          }
+          else {
+            filter = filterParam;
+          }
+      }
+      if (filterList != null) {
+        tableScan.setFilter(filterList);
+      } else if (filter != null) {
         tableScan.setFilter(filter);
       }
       int fetchSize = this.servlet.getConfiguration().getInt(Constants.SCAN_FETCH_SIZE, 10);

hbase-server/src/test/java/org/apache/hadoop/hbase/rest/TestTableScan.java

@@ -27,6 +27,7 @@ import java.io.EOFException;
 import java.io.IOException;
 import java.io.InputStream;
 import java.io.Serializable;
+import java.net.URLEncoder;
 import java.util.ArrayList;
 import java.util.List;
 
@@ -48,6 +49,9 @@ import org.apache.hadoop.hbase.HTableDescriptor;
 import org.apache.hadoop.hbase.MediumTests;
 import org.apache.hadoop.hbase.TableName;
 import org.apache.hadoop.hbase.client.HBaseAdmin;
+import org.apache.hadoop.hbase.filter.Filter;
+import org.apache.hadoop.hbase.filter.ParseFilter;
+import org.apache.hadoop.hbase.filter.PrefixFilter;
 import org.apache.hadoop.hbase.rest.client.Client;
 import org.apache.hadoop.hbase.rest.client.Cluster;
 import org.apache.hadoop.hbase.rest.client.Response;
@@ -87,6 +91,7 @@ public class TestTableScan {
   @BeforeClass
   public static void setUpBeforeClass() throws Exception {
     conf = TEST_UTIL.getConfiguration();
+    conf.set(Constants.CUSTOM_FILTERS, "CustomFilter:" + CustomFilter.class.getName()); 
     TEST_UTIL.startMiniCluster();
     REST_TEST_UTIL.startServletContainer(conf);
     client = new Client(new Cluster().add("localhost",
@@ -184,7 +189,6 @@ public class TestTableScan {
     count = TestScannerResource.countCellSet(model);
     assertEquals(15, count);
     checkRowsNotNull(model);
-
   }
 
   @Test
@@ -459,6 +463,108 @@ public class TestTableScan {
     int count = TestScannerResource.countCellSet(model);
     assertEquals(0, count);
   }
+  
+  @Test
+  public void testSimpleFilter() throws IOException, JAXBException {
+    StringBuilder builder = new StringBuilder();
+    builder = new StringBuilder();
+    builder.append("/*");
+    builder.append("?");
+    builder.append(Constants.SCAN_COLUMN + "=" + COLUMN_1);
+    builder.append("&");
+    builder.append(Constants.SCAN_START_ROW + "=aaa");
+    builder.append("&");
+    builder.append(Constants.SCAN_END_ROW + "=aay");
+    builder.append("&");
+    builder.append(Constants.SCAN_FILTER + "=" + URLEncoder.encode("PrefixFilter('aab')", "UTF-8"));
+    Response response =
+        client.get("/" + TABLE + builder.toString(), Constants.MIMETYPE_XML);
+    assertEquals(200, response.getCode());
+    JAXBContext ctx = JAXBContext.newInstance(CellSetModel.class);
+    Unmarshaller ush = ctx.createUnmarshaller();
+    CellSetModel model = (CellSetModel) ush.unmarshal(response.getStream());
+    int count = TestScannerResource.countCellSet(model);
+    assertEquals(1, count);
+    assertEquals("aab", new String(model.getRows().get(0).getCells().get(0).getValue()));
+  }
+
+  @Test
+  public void testCompoundFilter() throws IOException, JAXBException {
+    StringBuilder builder = new StringBuilder();
+    builder = new StringBuilder();
+    builder.append("/*");
+    builder.append("?");
+    builder.append(Constants.SCAN_FILTER + "="
+        + URLEncoder.encode("PrefixFilter('abc') AND QualifierFilter(=,'binary:1')", "UTF-8"));
+    Response response =
+        client.get("/" + TABLE + builder.toString(), Constants.MIMETYPE_XML);
+    assertEquals(200, response.getCode());
+    JAXBContext ctx = JAXBContext.newInstance(CellSetModel.class);
+    Unmarshaller ush = ctx.createUnmarshaller();
+    CellSetModel model = (CellSetModel) ush.unmarshal(response.getStream());
+    int count = TestScannerResource.countCellSet(model);
+    assertEquals(1, count);
+    assertEquals("abc", new String(model.getRows().get(0).getCells().get(0).getValue()));
+  }
+
+  @Test
+  public void testCustomFilter() throws IOException, JAXBException {
+    StringBuilder builder = new StringBuilder();
+    builder = new StringBuilder();
+    builder.append("/a*");
+    builder.append("?");
+    builder.append(Constants.SCAN_COLUMN + "=" + COLUMN_1);
+    builder.append("&");
+    builder.append(Constants.SCAN_FILTER + "=" + URLEncoder.encode("CustomFilter('abc')", "UTF-8"));
+    Response response =
+        client.get("/" + TABLE + builder.toString(), Constants.MIMETYPE_XML);
+    assertEquals(200, response.getCode());
+    JAXBContext ctx = JAXBContext.newInstance(CellSetModel.class);
+    Unmarshaller ush = ctx.createUnmarshaller();
+    CellSetModel model = (CellSetModel) ush.unmarshal(response.getStream());
+    int count = TestScannerResource.countCellSet(model);
+    assertEquals(1, count);
+    assertEquals("abc", new String(model.getRows().get(0).getCells().get(0).getValue()));
+  }
+  
+  @Test
+  public void testNegativeCustomFilter() throws IOException, JAXBException {
+    StringBuilder builder = new StringBuilder();
+    builder = new StringBuilder();
+    builder.append("/b*");
+    builder.append("?");
+    builder.append(Constants.SCAN_COLUMN + "=" + COLUMN_1);
+    builder.append("&");
+    builder.append(Constants.SCAN_FILTER + "=" + URLEncoder.encode("CustomFilter('abc')", "UTF-8"));
+    Response response =
+        client.get("/" + TABLE + builder.toString(), Constants.MIMETYPE_XML);
+    assertEquals(200, response.getCode());
+    JAXBContext ctx = JAXBContext.newInstance(CellSetModel.class);
+    Unmarshaller ush = ctx.createUnmarshaller();
+    CellSetModel model = (CellSetModel) ush.unmarshal(response.getStream());
+    int count = TestScannerResource.countCellSet(model);
+    // Should return no rows as the filters conflict
+    assertEquals(0, count);
+  }
+
+  public static class CustomFilter extends PrefixFilter {
+    private byte[] key = null;
+
+    public CustomFilter(byte[] key) {
+      super(key);
+    }
+    
+    @Override
+    public boolean filterRowKey(byte[] buffer, int offset, int length) {
+      int cmp = Bytes.compareTo(buffer, offset, length, this.key, 0, this.key.length);
+      return cmp != 0;
+    }
+
+    public static Filter createFilterFromArguments(ArrayList<byte[]> filterArguments) {
+      byte[] prefix = ParseFilter.removeQuotesFromByteArray(filterArguments.get(0));
+      return new CustomFilter(prefix);
+    }
+  }
 
   /**
    * The Class ClientSideCellSetModel which mimics cell set model, and contains listener to perform