com.google.cloud.bigtable.hbase.adapters.filters

Class SingleColumnValueFilterAdapter

java.lang.Object

com.google.cloud.bigtable.hbase.adapters.filters.TypedFilterAdapterBase<org.apache.hadoop.hbase.filter.SingleColumnValueFilter>

com.google.cloud.bigtable.hbase.adapters.filters.SingleColumnValueFilterAdapter

All Implemented Interfaces:

TypedFilterAdapter<org.apache.hadoop.hbase.filter.SingleColumnValueFilter>

@InternalApi(value="For internal usage only")
public class SingleColumnValueFilterAdapter
extends TypedFilterAdapterBase<org.apache.hadoop.hbase.filter.SingleColumnValueFilter>

Adapt SingleColumnValueFilter instances into bigtable RowFilters.

For internal use only - public for technical reasons.

Constructor Summary

Constructors

Constructor and Description
SingleColumnValueFilterAdapter(ValueFilterAdapter delegateAdapter) Constructor for SingleColumnValueFilterAdapter.

Method Summary

Modifier and Type	Method and Description
com.google.cloud.bigtable.data.v2.models.Filters.Filter	adapt(FilterAdapterContext context, org.apache.hadoop.hbase.filter.SingleColumnValueFilter filter) SingleColumnValueFilter is a filter that will return a row if a family/qualifier value matches some condition.
FilterSupportStatus	isFilterSupported(FilterAdapterContext context, org.apache.hadoop.hbase.filter.SingleColumnValueFilter filter) Determine if the given filter can be adapted to a Bigtable RowFilter.

Methods inherited from class com.google.cloud.bigtable.hbase.adapters.filters.TypedFilterAdapterBase

getIndexScanHint

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

SingleColumnValueFilterAdapter

public SingleColumnValueFilterAdapter(ValueFilterAdapter delegateAdapter)

Constructor for SingleColumnValueFilterAdapter.

Parameters:

delegateAdapter - a ValueFilterAdapter object.

Method Detail

adapt

public com.google.cloud.bigtable.data.v2.models.Filters.Filter adapt(FilterAdapterContext context,
                                                                     org.apache.hadoop.hbase.filter.SingleColumnValueFilter filter)
                                                              throws IOException

SingleColumnValueFilter is a filter that will return a row if a family/qualifier value matches some condition. Optionally, if SingleColumnValueFilter.getFilterIfMissing() is set to false, then also return the row if the family/column is not present on the row. There's a

Here's a rough translation of SingleColumnValueFilter.getFilterIfMissing() == true.

 IF a single family/column exists AND
    the value of the family/column meets some condition THEN
       return the ROW
 END
 

Here's a rough translation of SingleColumnValueFilter.getFilterIfMissing() == false.

 IF a single family/column exists THEN
   IF the value of the family/column meets some condition THEN
     return the ROW
   END
 ELSE IF filter.filter_if_missing == false THEN
   return the ROW
 END
 

The Cloud Bigtable filter translation for the SingleColumnValueFilter.getFilterIfMissing() true case here's the resulting filter is as follows:

   condition: {
      predicate: {
        chain: {
           family: [filter.family]
           qualifier: [filter.qualifier],
           // if filter.latestOnly, then add
           // cells_per_column: 1
           value: // something interesting
        }
      }
      true_filter: {
         pass_all: true
      }
   }
 

In addition to the default filter, there's a bit more if SingleColumnValueFilter.getFilterIfMissing() is false. Here's what the filter would look like:

   interleave: [ // either
     {
       // If the family/qualifier exists and matches a value
       // Then return the row
       // Else return nothing
       condition: {
         predicate: {
           chain: {
             family: [filter.family]
             qualifier: [filter.qualifier],
             // if filter.latestOnly, then add
             // cells_per_column: 1
             value: // something interesting
           }
         },
         true_filter: { pass_all: true }
       }
     }, {
       // If the family/qualifer exists
       // Then return nothing
       // Else return row
       condition: {
         predicate: {
           chain: {
             family: [filter.family]
             qualifier: [filter.qualifier],
           }
         },
         false_filter: { pass_all: true }
       }
     }
   ]
 

NOTE: This logic can also be expressed as nested predicates, but that approach creates really poor performance on the server side.

Parameters:

context - a FilterAdapterContext object.

filter - a S object.

Returns:

a Filters.Filter object.

Throws:

IOException - if any.

isFilterSupported

public FilterSupportStatus isFilterSupported(FilterAdapterContext context,
                                             org.apache.hadoop.hbase.filter.SingleColumnValueFilter filter)

Determine if the given filter can be adapted to a Bigtable RowFilter.

Parameters:

context - a FilterAdapterContext object.

filter - a S object.

Returns:

a FilterSupportStatus object.