HadoopPipeline (Titan-Hadoop: Version-independent Core 0.5.4-jboss-1 API)

java.lang.Object
- com.thinkaurelius.titan.hadoop.HadoopPipeline

```
public class HadoopPipeline
extends Object
```
A HadoopPipeline defines a breadth-first traversal through a property graph representation. Gremlin/Hadoop compiles down to a HadoopPipeline which is ultimately a series of MapReduce jobs.

Author:

Marko A. Rodriguez (http://markorodriguez.com)

Nested Class Summary

Nested Classes
Modifier and Type Class and Description

protected class HadoopPipeline.State

Nested Classes
Modifier and Type	Class and Description
`protected class`	`HadoopPipeline.State`

Field Summary

Fields
Modifier and Type	Field and Description
`protected HadoopCompiler`	`compiler`
`protected static ScriptEngine`	`engine`
`protected HadoopGraph`	`graph`
`static String`	`PIPELINE_IS_LOCKED`
`protected HadoopPipeline.State`	`state`
`protected List<String>`	`stringRepresentation`

Constructor Summary

Constructors
Constructor and Description

HadoopPipeline(HadoopGraph graph)
Construct a HadoopPipeline

Constructors
Constructor and Description
`HadoopPipeline(HadoopGraph graph)` Construct a HadoopPipeline

Method Summary

All Methods Instance Methods Concrete Methods
Modifier and Type	Method and Description
`HadoopPipeline`	`_()` The identity step does not alter the graph in anyway.
`HadoopPipeline`	`as(String name)` Name a step in order to reference it later in the expression.
`HadoopPipeline`	`back(String step)` Go back to an element a named step ago.
`HadoopPipeline`	`both(String... labels)` Take both incoming and outgoing labeled edges to adjacent vertices.
`HadoopPipeline`	`bothE(String... labels)` Take both incoming and outgoing labeled edges to incident edges.
`HadoopPipeline`	`bothV()` Go to both the incoming/head and outgoing/tail vertices of the edge.
`HadoopPipeline`	`count()` Count the number of traversers currently in the graph
`HadoopPipeline`	`dedup()` Remove any duplicate traversers at a single element.
`HadoopPipeline`	`drop()` Drop all the elements of respective type at the current step.
`HadoopPipeline`	`E()` Start a traversal at all edges in the graph.
`HadoopPipeline`	`filter(String closure)` Emit or deny the current element based upon the provided boolean-based closure.
`HadoopCompiler`	`getCompiler()`
`HadoopGraph`	`getGraph()` Get a reference to the graph currently being used in this HadoopPipeline.
`HadoopPipeline`	`groupCount()` Count the number of times the previous element (or property) has been traversed to.
`HadoopPipeline`	`groupCount(String keyClosure)` Apply the provided closure to the incoming element to determine the grouping key.
`HadoopPipeline`	`groupCount(String keyClosure, String valueClosure)` Apply the provided closure to the incoming element to determine the grouping key.
`HadoopPipeline`	`has(String key, com.tinkerpop.blueprints.Compare compare, Object... values)` Emit the current element if it has a property value comparable to the provided values.
`HadoopPipeline`	`has(String key, Object... values)` Emit the current element it has a property value equal to the provided values.
`HadoopPipeline`	`has(String key, com.tinkerpop.gremlin.Tokens.T compare, Object... values)` Emit the current element if it has a property value comparable to the provided values.
`HadoopPipeline`	`hasNot(String key, com.tinkerpop.blueprints.Compare compare, Object... values)` Emit the current element if it does not have a property value comparable to the provided values.
`HadoopPipeline`	`hasNot(String key, Object... values)` Emit the current element it does not have a property value equal to the provided values.
`HadoopPipeline`	`hasNot(String key, com.tinkerpop.gremlin.Tokens.T compare, Object... values)` Emit the current element if it does not have a property value comparable to the provided values.
`HadoopPipeline`	`in(String... labels)` Take incoming labeled edges to adjacent vertices.
`HadoopPipeline`	`inE(String... labels)` Take incoming labeled edges to incident edges.
`HadoopPipeline`	`interval(String key, Object startValue, Object endValue)` Emit the current element it has a property value equal within the provided range.
`HadoopPipeline`	`inV()` Go to the incoming/head vertex of the edge.
`HadoopPipeline`	`keep()` Keep all the elements of the respetive type at the current step.
`HadoopPipeline`	`label()` Emit the label of the current edge.
`HadoopPipeline`	`linkIn(String label, String step)` Have the elements for the named step previous project an edge to the current vertex with provided label.
`HadoopPipeline`	`linkIn(String label, String step, String mergeWeightKey)` Have the elements for the named step previous project an edge to the current vertex with provided label.
`HadoopPipeline`	`linkOut(String label, String step)` Have the elements for the named step previous project an edge from the current vertex with provided label.
`HadoopPipeline`	`linkOut(String label, String step, String mergeWeightKey)` Have the elements for the named step previous project an edge from the current vertex with provided label.
`HadoopPipeline`	`map()` Emit a string representation of the property map.
`HadoopPipeline`	`order(com.tinkerpop.gremlin.Tokens.T order)` Order the previous property value results.
`HadoopPipeline`	`order(com.tinkerpop.gremlin.Tokens.T order, String elementKey)` Order the previous property value results and emit them with another element property value.
`HadoopPipeline`	`order(com.tinkerpop.pipes.transform.TransformPipe.Order order)` Order the previous property value results.
`HadoopPipeline`	`order(com.tinkerpop.pipes.transform.TransformPipe.Order order, String elementKey)` Order the previous property value results and emit them with another element property value.
`HadoopPipeline`	`out(String... labels)` Take outgoing labeled edges to adjacent vertices.
`HadoopPipeline`	`outE(String... labels)` Take outgoing labeled edges to incident edges.
`HadoopPipeline`	`outV()` Go to the outgoing/tail vertex of the edge.
`HadoopPipeline`	`path()` Emit the path taken from start to current element.
`HadoopPipeline`	`property(String key)` Emit the property value of an element.
`HadoopPipeline`	`property(String key, Class type)` Emit the property value of an element.
`HadoopPipeline`	`script(String scriptUri, String... args)`
`HadoopPipeline`	`sideEffect(String closure)` Emit the element, but compute some sideeffect in the process.
`HadoopPipeline`	`simplePath()` Emit the element only if it was arrived at via a path that does not have cycles in it.
`int`	`submit()` Submit the HadoopPipeline to the Hadoop cluster.
`int`	`submit(String script, Boolean showHeader)` Submit the HadoopPipeline to the Hadoop cluster and ensure that a header is emitted in the logs.
`String`	`toString()`
`HadoopPipeline`	`transform(String closure)` Apply the provided closure to the current element and emit the result.
`HadoopPipeline`	`V()` Start a traversal at all vertices in the graph.
`HadoopPipeline`	`v(long... ids)` Start a traversal at the vertices identified by the provided ids.

Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, wait, wait, wait

- Field Detail
  - engine
```
protected static final ScriptEngine engine
```
  - PIPELINE_IS_LOCKED
```
public static final String PIPELINE_IS_LOCKED
```
    See Also:
    
    Constant Field Values
  - compiler
```
protected final HadoopCompiler compiler
```
  - graph
```
protected final HadoopGraph graph
```
  - state
```
protected final HadoopPipeline.State state
```
  - stringRepresentation
```
protected final List<String> stringRepresentation
```
- Constructor Detail
  - HadoopPipeline
```
public HadoopPipeline(HadoopGraph graph)
```
    Construct a HadoopPipeline
    
    Parameters:
    
    graph - the HadoopGraph that is the source of the traversal
- Method Detail
  - _
```
public HadoopPipeline _()
```
    The identity step does not alter the graph in anyway. It has the benefit of emitting various useful graph statistic counters.
    
    Returns:
    
    the extended HadoopPipeline
  - transform
```
public HadoopPipeline transform(String closure)
```
    Apply the provided closure to the current element and emit the result.
    
    Parameters:
    
    closure - the closure to apply to the element
    
    Returns:
    
    the extended HadoopPipeline
  - V
```
public HadoopPipeline V()
```
    Start a traversal at all vertices in the graph.
    
    Returns:
    
    the extended HadoopPipeline
  - E
```
public HadoopPipeline E()
```
    Start a traversal at all edges in the graph.
    
    Returns:
    
    the extended HadoopPipeline
  - v
```
public HadoopPipeline v(long... ids)
```
    Start a traversal at the vertices identified by the provided ids.
    
    Parameters:
    
    ids - the long ids of the vertices to start the traversal from
    
    Returns:
    
    the extended HadoopPipeline
  - out
```
public HadoopPipeline out(String... labels)
```
    Take outgoing labeled edges to adjacent vertices.
    
    Parameters:
    
    labels - the labels of the edges to traverse over
    
    Returns:
    
    the extended HadoopPipeline
  - in
```
public HadoopPipeline in(String... labels)
```
    Take incoming labeled edges to adjacent vertices.
    
    Parameters:
    
    labels - the labels of the edges to traverse over
    
    Returns:
    
    the extended HadoopPipeline
  - both
```
public HadoopPipeline both(String... labels)
```
    Take both incoming and outgoing labeled edges to adjacent vertices.
    
    Parameters:
    
    labels - the labels of the edges to traverse over
    
    Returns:
    
    the extended HadoopPipeline
  - outE
```
public HadoopPipeline outE(String... labels)
```
    Take outgoing labeled edges to incident edges.
    
    Parameters:
    
    labels - the labels of the edges to traverse over
    
    Returns:
    
    the extended HadoopPipeline
  - inE
```
public HadoopPipeline inE(String... labels)
```
    Take incoming labeled edges to incident edges.
    
    Parameters:
    
    labels - the labels of the edges to traverse over
    
    Returns:
    
    the extended HadoopPipeline
  - bothE
```
public HadoopPipeline bothE(String... labels)
```
    Take both incoming and outgoing labeled edges to incident edges.
    
    Parameters:
    
    labels - the labels of the edges to traverse over
    
    Returns:
    
    the extended HadoopPipeline
  - outV
```
public HadoopPipeline outV()
```
    Go to the outgoing/tail vertex of the edge.
    
    Returns:
    
    the extended HadoopPipeline
  - inV
```
public HadoopPipeline inV()
```
    Go to the incoming/head vertex of the edge.
    
    Returns:
    
    the extended HadoopPipeline
  - bothV
```
public HadoopPipeline bothV()
```
    Go to both the incoming/head and outgoing/tail vertices of the edge.
    
    Returns:
    
    the extended HadoopPipeline
  - property
```
public HadoopPipeline property(String key,
                               Class type)
```
    Emit the property value of an element.
    
    Parameters:
    
    key - the key identifying the property
    
    type - the class of the property value (so Hadoop can intelligently handle the result)
    
    Returns:
    
    the extended HadoopPipeline
  - property
```
public HadoopPipeline property(String key)
```
    Emit the property value of an element.
    
    Parameters:
    
    key - the key identifying the property
    
    Returns:
    
    the extended HadoopPipeline
  - map
```
public HadoopPipeline map()
```
    Emit a string representation of the property map.
    
    Returns:
    
    the extended HadoopPipeline
  - label
```
public HadoopPipeline label()
```
    Emit the label of the current edge.
    
    Returns:
    
    the extended HadoopPipeline
  - path
```
public HadoopPipeline path()
```
    Emit the path taken from start to current element.
    
    Returns:
    
    the extended HadoopPipeline
  - order
```
public HadoopPipeline order(com.tinkerpop.pipes.transform.TransformPipe.Order order,
                            String elementKey)
```
    Order the previous property value results and emit them with another element property value. It is important to emit the previous property with a provided type else it is ordered lexigraphically.
    
    Parameters:
    
    order - increasing and descending order
    
    elementKey - the key of the element to associate it with
    
    Returns:
    
    the extended HadoopPipeline
  - order
```
public HadoopPipeline order(com.tinkerpop.pipes.transform.TransformPipe.Order order)
```
    Order the previous property value results.
    
    Parameters:
    
    order - increasing and descending order
    
    Returns:
    
    the extended HadoopPipeline
  - order
```
public HadoopPipeline order(com.tinkerpop.gremlin.Tokens.T order,
                            String elementKey)
```
    Order the previous property value results and emit them with another element property value. It is important to emit the previous property with a provided type else it is ordered lexigraphically.
    
    Parameters:
    
    order - increasing and descending order
    
    elementKey - the key of the element to associate it with
    
    Returns:
    
    the extended HadoopPipeline
  - order
```
public HadoopPipeline order(com.tinkerpop.gremlin.Tokens.T order)
```
    Order the previous property value results.
    
    Parameters:
    
    order - increasing and descending order
    
    Returns:
    
    the extended HadoopPipeline
  - filter
```
public HadoopPipeline filter(String closure)
```
    Emit or deny the current element based upon the provided boolean-based closure.
    
    Parameters:
    
    closure - return true to emit and false to remove.
    
    Returns:
    
    the extended HadoopPipeline
  - has
```
public HadoopPipeline has(String key,
                          com.tinkerpop.gremlin.Tokens.T compare,
                          Object... values)
```
    Emit the current element if it has a property value comparable to the provided values.
    
    Parameters:
    
    key - the property key of the element
    
    compare - the comparator
    
    values - the values to compare against where only one needs to succeed (or'd)
    
    Returns:
    
    the extended HadoopPipeline
  - hasNot
```
public HadoopPipeline hasNot(String key,
                             com.tinkerpop.gremlin.Tokens.T compare,
                             Object... values)
```
    Emit the current element if it does not have a property value comparable to the provided values.
    
    Parameters:
    
    key - the property key of the element
    
    compare - the comparator (will be not'd)
    
    values - the values to compare against where only one needs to succeed (or'd)
    
    Returns:
    
    the extended HadoopPipeline
  - has
```
public HadoopPipeline has(String key,
                          com.tinkerpop.blueprints.Compare compare,
                          Object... values)
```
    Emit the current element if it has a property value comparable to the provided values.
    
    Parameters:
    
    key - the property key of the element
    
    compare - the comparator
    
    values - the values to compare against where only one needs to succeed (or'd)
    
    Returns:
    
    the extended HadoopPipeline
  - hasNot
```
public HadoopPipeline hasNot(String key,
                             com.tinkerpop.blueprints.Compare compare,
                             Object... values)
```
    Emit the current element if it does not have a property value comparable to the provided values.
    
    Parameters:
    
    key - the property key of the element
    
    compare - the comparator (will be not'd)
    
    values - the values to compare against where only one needs to succeed (or'd)
    
    Returns:
    
    the extended HadoopPipeline
  - has
```
public HadoopPipeline has(String key,
                          Object... values)
```
    Emit the current element it has a property value equal to the provided values.
    
    Parameters:
    
    key - the property key of the element
    
    values - the values to compare against where only one needs to succeed (or'd)
    
    Returns:
    
    the extended HadoopPipeline
  - hasNot
```
public HadoopPipeline hasNot(String key,
                             Object... values)
```
    Emit the current element it does not have a property value equal to the provided values.
    
    Parameters:
    
    key - the property key of the element
    
    values - the values to compare against where only one needs to succeed (or'd)
    
    Returns:
    
    the extended HadoopPipeline
  - interval
```
public HadoopPipeline interval(String key,
                               Object startValue,
                               Object endValue)
```
    Emit the current element it has a property value equal within the provided range.
    
    Parameters:
    
    key - the property key of the element
    
    startValue - the start of the range (inclusive)
    
    endValue - the end of the range (exclusive)
    
    Returns:
    
    the extended HadoopPipeline
  - dedup
```
public HadoopPipeline dedup()
```
    Remove any duplicate traversers at a single element.
    
    Returns:
    
    the extended HadoopPipeline
  - back
```
public HadoopPipeline back(String step)
```
    Go back to an element a named step ago. Currently only backing up to vertices is supported.
    
    Parameters:
    
    step - the name of the step to back up to
    
    Returns:
    
    the extended HadoopPipeline
  - simplePath
```
public HadoopPipeline simplePath()
```
    Emit the element only if it was arrived at via a path that does not have cycles in it.
    
    Returns:
    
    the extended HadoopPipeline
  - sideEffect
```
public HadoopPipeline sideEffect(String closure)
```
    Emit the element, but compute some sideeffect in the process. For example, mutate the properties of the element.
    
    Parameters:
    
    closure - the sideeffect closure whose results are ignored.
    
    Returns:
    
    the extended HadoopPipeline
  - as
```
public HadoopPipeline as(String name)
```
    Name a step in order to reference it later in the expression.
    
    Parameters:
    
    name - the string representation of the name
    
    Returns:
    
    the extended HadoopPipeline
  - linkIn
```
public HadoopPipeline linkIn(String label,
                             String step,
                             String mergeWeightKey)
```
    Have the elements for the named step previous project an edge to the current vertex with provided label. If a merge weight key is provided, then count the number of duplicate edges between the same two vertices and add a weight. No weight key is specified by "_" and then all duplicates are merged, but no weight is added to the resultant edge.
    
    Parameters:
    
    step - the name of the step where the source vertices were
    
    label - the label of the edge to project
    
    mergeWeightKey - the property key to use for weight
    
    Returns:
    
    the extended HadoopPipeline
  - linkIn
```
public HadoopPipeline linkIn(String label,
                             String step)
```
    Have the elements for the named step previous project an edge to the current vertex with provided label.
    
    Parameters:
    
    step - the name of the step where the source vertices were
    
    label - the label of the edge to project
    
    Returns:
    
    the extended HadoopPipeline
  - linkOut
```
public HadoopPipeline linkOut(String label,
                              String step,
                              String mergeWeightKey)
```
    Have the elements for the named step previous project an edge from the current vertex with provided label. If a merge weight key is provided, then count the number of duplicate edges between the same two vertices and add a weight. No weight key is specified by "_" and then all duplicates are merged, but no weight is added to the resultant edge.
    
    Parameters:
    
    step - the name of the step where the source vertices were
    
    label - the label of the edge to project
    
    mergeWeightKey - the property key to use for weight
    
    Returns:
    
    the extended HadoopPipeline
  - linkOut
```
public HadoopPipeline linkOut(String label,
                              String step)
```
    Have the elements for the named step previous project an edge from the current vertex with provided label.
    
    Parameters:
    
    step - the name of the step where the source vertices were
    
    label - the label of the edge to project
    
    Returns:
    
    the extended HadoopPipeline
  - groupCount
```
public HadoopPipeline groupCount()
```
    Count the number of times the previous element (or property) has been traversed to. The results are stored in the jobs sideeffect file in HDFS.
    
    Returns:
    
    the extended HadoopPipeline.
  - groupCount
```
public HadoopPipeline groupCount(String keyClosure)
```
    Apply the provided closure to the incoming element to determine the grouping key. The value of the count is incremented by 1 The results are stored in the jobs sideeffect file in HDFS.
    
    Returns:
    
    the extended HadoopPipeline.
  - groupCount
```
public HadoopPipeline groupCount(String keyClosure,
                                 String valueClosure)
```
    Apply the provided closure to the incoming element to determine the grouping key. Then apply the value closure to the current element to determine the count increment. The results are stored in the jobs sideeffect file in HDFS.
    
    Returns:
    
    the extended HadoopPipeline.
  - drop
```
public HadoopPipeline drop()
```
    Drop all the elements of respective type at the current step. Keep all others.
    
    Returns:
    
    the extended HadoopPipeline
  - keep
```
public HadoopPipeline keep()
```
    Keep all the elements of the respetive type at the current step. Drop all others.
    
    Returns:
    
    the extended HadoopPipeline
  - script
```
public HadoopPipeline script(String scriptUri,
                             String... args)
```
  - count
```
public HadoopPipeline count()
```
    Count the number of traversers currently in the graph
    
    Returns:
    
    the count
  - toString
```
public String toString()
```
    Overrides:
    
    toString in class Object
  - submit
```
public int submit()
           throws Exception
```
    Submit the HadoopPipeline to the Hadoop cluster.
    
    Throws:
    
    Exception
  - submit
```
public int submit(String script,
                  Boolean showHeader)
           throws Exception
```
    Submit the HadoopPipeline to the Hadoop cluster and ensure that a header is emitted in the logs.
    
    Parameters:
    
    script - the Gremlin script
    
    showHeader - the Titan/Hadoop header
    
    Throws:
    
    Exception
  - getGraph
```
public HadoopGraph getGraph()
```
    Get a reference to the graph currently being used in this HadoopPipeline.
    
    Returns:
    
    the HadoopGraph
  - getCompiler
```
public HadoopCompiler getCompiler()
```

Class HadoopPipeline

Nested Class Summary

Field Summary

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Field Detail

engine

PIPELINE_IS_LOCKED

compiler

graph

state

stringRepresentation

Constructor Detail

HadoopPipeline

Method Detail

_

transform

V

E

v

out

in

both

outE

inE

bothE

outV

inV

bothV

property

property

map

label

path

order

order

order

order

filter

has

hasNot

has

hasNot

has

hasNot

interval

dedup

back

simplePath

sideEffect

as

linkIn

linkIn

linkOut

linkOut

groupCount

groupCount

groupCount

drop

keep

script

count

toString

submit

submit

getGraph

getCompiler