com.intel.hadoop.graphbuilder.job
Class AbstractIdNormalizationJob
java.lang.Object
com.intel.hadoop.graphbuilder.job.AbstractIdNormalizationJob
- Type Parameters:
VidType
- VertexData
- EdgeData
-
- Direct Known Subclasses:
- IdNormalizationJobTest.Job
public abstract class AbstractIdNormalizationJob
- extends java.lang.Object
An abstract wrapper class for running the Graph Id Normalization Job. User
needs to override 3 methods:
vidClass(), vdataClass(), and edataClass()
to generate the right
parsers for parsing the input data. BasicGraphParser
is used, and can
be replaced by overriding the
method.
See an example in IdNormalizationJobTest
.
Input directory:
- $inputdir/edata input edge data
- $inputdir/vdata input vertex data
Output directories:
- $outputdir/edata edges for partition{$i}
- $outputdir/vdata vertex records for partition{$i}
- $outputdir/vidmap dictioanry for raw vid to normalized vid
- See Also:
BasicGraphParser
,
GraphParser
,
IdNormalizationJobTest
Methods inherited from class java.lang.Object |
equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait |
AbstractIdNormalizationJob
public AbstractIdNormalizationJob()
vidClass
public abstract java.lang.Class vidClass()
- Returns:
- the class of vertex id type
vdataClass
public abstract java.lang.Class vdataClass()
- Returns:
- the class of vertex data type
edataClass
public abstract java.lang.Class edataClass()
- Returns:
- the class of edge data type
graphParserClass
public java.lang.Class graphParserClass()
- Returns:
- the class of graph parser type
setDictionaryParts
public void setDictionaryParts(int n)
- Parameters:
n
- number of partitions of the dictionary
run
public boolean run(java.lang.String input,
java.lang.String output)
throws javassist.NotFoundException,
javassist.CannotCompileException
- Running the normalization job. Reads input from
input
, and outputs
to output
directory.
- Parameters:
nparts
- inputs
- output
-
- Returns:
-
- Throws:
javassist.NotFoundException
javassist.CannotCompileException