scalding avro

SiteCatNsTuple.java

/**
 * Autogenerated by Avro
 * 
 * DO NOT EDIT DIRECTLY
 */
package com.bluekai.analytics.jobs;  
@SuppressWarnings("all")
@org.apache.avro.specific.AvroGenerated
public class SiteCatNsTuple extends org.apache.avro.specific.SpecificRecordBase implements org.apache.avro.specific.SpecificRecord {
  public static final org.apache.avro.Schema SCHEMA$ = new org.apache.avro.Schema.Parser().parse("{\"type\":\"record\",\"name\":\"SiteCatNsTuple\",\"namespace\":\"com.bluekai.analytics.jobs\",\"fields\":[{\"name\":\"site\",\"type\":\"int\"},{\"name\":\"ns\",\"type\":\"int\"},{\"name\":\"cat\",\"type\":\"int\"}]}");
  public static org.apache.avro.Schema getClassSchema() { return SCHEMA$; }
  @Deprecated public int site;
  @Deprecated public int ns;
  @Deprecated public int cat;

  /**
   * Default constructor.  Note that this does not initialize fields
   * to their default values from the schema.  If that is desired then
   * one should use <code>newBuilder()</code>. 
   */
  public SiteCatNsTuple() {}

  /**
   * All-args constructor.
   */
  public SiteCatNsTuple(java.lang.Integer site, java.lang.Integer ns, java.lang.Integer cat) {
    this.site = site;
    this.ns = ns;
    this.cat = cat;
  }

  public org.apache.avro.Schema getSchema() { return SCHEMA$; }
  // Used by DatumWriter.  Applications should not call. 
  public java.lang.Object get(int field$) {
    switch (field$) {
    case 0: return site;
    case 1: return ns;
    case 2: return cat;
    default: throw new org.apache.avro.AvroRuntimeException("Bad index");
    }
  }
  // Used by DatumReader.  Applications should not call. 
  @SuppressWarnings(value="unchecked")
  public void put(int field$, java.lang.Object value$) {
    switch (field$) {
    case 0: site = (java.lang.Integer)value$; break;
    case 1: ns = (java.lang.Integer)value$; break;
    case 2: cat = (java.lang.Integer)value$; break;
    default: throw new org.apache.avro.AvroRuntimeException("Bad index");
    }
  }

  /**
   * Gets the value of the 'site' field.
   */
  public java.lang.Integer getSite() {
    return site;
  }

  /**
   * Sets the value of the 'site' field.
   * @param value the value to set.
   */
  public void setSite(java.lang.Integer value) {
    this.site = value;
  }

  /**
   * Gets the value of the 'ns' field.
   */
  public java.lang.Integer getNs() {
    return ns;
  }

  /**
   * Sets the value of the 'ns' field.
   * @param value the value to set.
   */
  public void setNs(java.lang.Integer value) {
    this.ns = value;
  }

  /**
   * Gets the value of the 'cat' field.
   */
  public java.lang.Integer getCat() {
    return cat;
  }

  /**
   * Sets the value of the 'cat' field.
   * @param value the value to set.
   */
  public void setCat(java.lang.Integer value) {
    this.cat = value;
  }

  /** Creates a new SiteCatNsTuple RecordBuilder */
  public static com.bluekai.analytics.jobs.SiteCatNsTuple.Builder newBuilder() {
    return new com.bluekai.analytics.jobs.SiteCatNsTuple.Builder();
  }
  
  /** Creates a new SiteCatNsTuple RecordBuilder by copying an existing Builder */
  public static com.bluekai.analytics.jobs.SiteCatNsTuple.Builder newBuilder(com.bluekai.analytics.jobs.SiteCatNsTuple.Builder other) {
    return new com.bluekai.analytics.jobs.SiteCatNsTuple.Builder(other);
  }
  
  /** Creates a new SiteCatNsTuple RecordBuilder by copying an existing SiteCatNsTuple instance */
  public static com.bluekai.analytics.jobs.SiteCatNsTuple.Builder newBuilder(com.bluekai.analytics.jobs.SiteCatNsTuple other) {
    return new com.bluekai.analytics.jobs.SiteCatNsTuple.Builder(other);
  }
  
  /**
   * RecordBuilder for SiteCatNsTuple instances.
   */
  public static class Builder extends org.apache.avro.specific.SpecificRecordBuilderBase<SiteCatNsTuple>
    implements org.apache.avro.data.RecordBuilder<SiteCatNsTuple> {

    private int site;
    private int ns;
    private int cat;

    /** Creates a new Builder */
    private Builder() {
      super(SCHEMA$);
    }
    
    /** Creates a Builder by copying an existing Builder */
    private Builder(com.bluekai.analytics.jobs.SiteCatNsTuple.Builder other) {
      super(other);
      if (isValidValue(fields()[0], other.site)) {
        this.site = data().deepCopy(fields()[0].schema(), other.site);
        fieldSetFlags()[0] = true;
      }
      if (isValidValue(fields()[1], other.ns)) {
        this.ns = data().deepCopy(fields()[1].schema(), other.ns);
        fieldSetFlags()[1] = true;
      }
      if (isValidValue(fields()[2], other.cat)) {
        this.cat = data().deepCopy(fields()[2].schema(), other.cat);
        fieldSetFlags()[2] = true;
      }
    }
    
    /** Creates a Builder by copying an existing SiteCatNsTuple instance */
    private Builder(com.bluekai.analytics.jobs.SiteCatNsTuple other) {
            super(SCHEMA$);
      if (isValidValue(fields()[0], other.site)) {
        this.site = data().deepCopy(fields()[0].schema(), other.site);
        fieldSetFlags()[0] = true;
      }
      if (isValidValue(fields()[1], other.ns)) {
        this.ns = data().deepCopy(fields()[1].schema(), other.ns);
        fieldSetFlags()[1] = true;
      }
      if (isValidValue(fields()[2], other.cat)) {
        this.cat = data().deepCopy(fields()[2].schema(), other.cat);
        fieldSetFlags()[2] = true;
      }
    }

    /** Gets the value of the 'site' field */
    public java.lang.Integer getSite() {
      return site;
    }
    
    /** Sets the value of the 'site' field */
    public com.bluekai.analytics.jobs.SiteCatNsTuple.Builder setSite(int value) {
      validate(fields()[0], value);
      this.site = value;
      fieldSetFlags()[0] = true;
      return this; 
    }
    
    /** Checks whether the 'site' field has been set */
    public boolean hasSite() {
      return fieldSetFlags()[0];
    }
    
    /** Clears the value of the 'site' field */
    public com.bluekai.analytics.jobs.SiteCatNsTuple.Builder clearSite() {
      fieldSetFlags()[0] = false;
      return this;
    }

    /** Gets the value of the 'ns' field */
    public java.lang.Integer getNs() {
      return ns;
    }
    
    /** Sets the value of the 'ns' field */
    public com.bluekai.analytics.jobs.SiteCatNsTuple.Builder setNs(int value) {
      validate(fields()[1], value);
      this.ns = value;
      fieldSetFlags()[1] = true;
      return this; 
    }
    
    /** Checks whether the 'ns' field has been set */
    public boolean hasNs() {
      return fieldSetFlags()[1];
    }
    
    /** Clears the value of the 'ns' field */
    public com.bluekai.analytics.jobs.SiteCatNsTuple.Builder clearNs() {
      fieldSetFlags()[1] = false;
      return this;
    }

    /** Gets the value of the 'cat' field */
    public java.lang.Integer getCat() {
      return cat;
    }
    
    /** Sets the value of the 'cat' field */
    public com.bluekai.analytics.jobs.SiteCatNsTuple.Builder setCat(int value) {
      validate(fields()[2], value);
      this.cat = value;
      fieldSetFlags()[2] = true;
      return this; 
    }
    
    /** Checks whether the 'cat' field has been set */
    public boolean hasCat() {
      return fieldSetFlags()[2];
    }
    
    /** Clears the value of the 'cat' field */
    public com.bluekai.analytics.jobs.SiteCatNsTuple.Builder clearCat() {
      fieldSetFlags()[2] = false;
      return this;
    }

    @Override
    public SiteCatNsTuple build() {
      try {
        SiteCatNsTuple record = new SiteCatNsTuple();
        record.site = fieldSetFlags()[0] ? this.site : (java.lang.Integer) defaultValue(fields()[0]);
        record.ns = fieldSetFlags()[1] ? this.ns : (java.lang.Integer) defaultValue(fields()[1]);
        record.cat = fieldSetFlags()[2] ? this.cat : (java.lang.Integer) defaultValue(fields()[2]);
        return record;
      } catch (Exception e) {
        throw new org.apache.avro.AvroRuntimeException(e);
      }
    }
  }
}

Test.scala

package com.bluekai.ds


import com.twitter.scalding._
import scala.collection.JavaConverters._
import com.twitter.scalding.avro._
import com.bluekai.analytics.jobs.{UserProfileFlatAggregate}

class Test(args: Args) extends Job(args) {
  override def config: Map[AnyRef, AnyRef] = {
    super.config ++ Map("cascading.app.appjar.class" -> classOf[Test])
  }

  val input = args("input")
  val avro = PackedAvroSource[UserProfileFlatAggregate](args("input")).read
    .flatMapTo('UserProfileFlatAggregate -> ('uuid, 'cat)) {
    userProfile: UserProfileFlatAggregate =>
      val uuid = userProfile.getUuid.toString
      val tuples = userProfile.getProfile.asScala
      for (t <- tuples) yield (uuid, t.getCat)
  }.debug
  .write(Tsv("/temp/output.tsv"))
}

UserProfileFlatAggregate.java

/**
 * Autogenerated by Avro
 * 
 * DO NOT EDIT DIRECTLY
 */
package com.bluekai.analytics.jobs;  
@SuppressWarnings("all")
@org.apache.avro.specific.AvroGenerated
public class UserProfileFlatAggregate extends org.apache.avro.specific.SpecificRecordBase implements org.apache.avro.specific.SpecificRecord {
  public static final org.apache.avro.Schema SCHEMA$ = new org.apache.avro.Schema.Parser().parse("{\"type\":\"record\",\"name\":\"UserProfileFlatAggregate\",\"namespace\":\"com.bluekai.analytics.jobs\",\"fields\":[{\"name\":\"uuid\",\"type\":\"string\"},{\"name\":\"profile\",\"type\":{\"type\":\"array\",\"items\":{\"type\":\"record\",\"name\":\"SiteCatNsTuple\",\"fields\":[{\"name\":\"site\",\"type\":\"int\"},{\"name\":\"ns\",\"type\":\"int\"},{\"name\":\"cat\",\"type\":\"int\"}]}},\"aliases\":[\"bag\"]}]}");
  public static org.apache.avro.Schema getClassSchema() { return SCHEMA$; }
  @Deprecated public java.lang.CharSequence uuid;
  @Deprecated public java.util.List<com.bluekai.analytics.jobs.SiteCatNsTuple> profile;

  /**
   * Default constructor.  Note that this does not initialize fields
   * to their default values from the schema.  If that is desired then
   * one should use <code>newBuilder()</code>. 
   */
  public UserProfileFlatAggregate() {}

  /**
   * All-args constructor.
   */
  public UserProfileFlatAggregate(java.lang.CharSequence uuid, java.util.List<com.bluekai.analytics.jobs.SiteCatNsTuple> profile) {
    this.uuid = uuid;
    this.profile = profile;
  }

  public org.apache.avro.Schema getSchema() { return SCHEMA$; }
  // Used by DatumWriter.  Applications should not call. 
  public java.lang.Object get(int field$) {
    switch (field$) {
    case 0: return uuid;
    case 1: return profile;
    default: throw new org.apache.avro.AvroRuntimeException("Bad index");
    }
  }
  // Used by DatumReader.  Applications should not call. 
  @SuppressWarnings(value="unchecked")
  public void put(int field$, java.lang.Object value$) {
    switch (field$) {
    case 0: uuid = (java.lang.CharSequence)value$; break;
    case 1: profile = (java.util.List<com.bluekai.analytics.jobs.SiteCatNsTuple>)value$; break;
    default: throw new org.apache.avro.AvroRuntimeException("Bad index");
    }
  }

  /**
   * Gets the value of the 'uuid' field.
   */
  public java.lang.CharSequence getUuid() {
    return uuid;
  }

  /**
   * Sets the value of the 'uuid' field.
   * @param value the value to set.
   */
  public void setUuid(java.lang.CharSequence value) {
    this.uuid = value;
  }

  /**
   * Gets the value of the 'profile' field.
   */
  public java.util.List<com.bluekai.analytics.jobs.SiteCatNsTuple> getProfile() {
    return profile;
  }

  /**
   * Sets the value of the 'profile' field.
   * @param value the value to set.
   */
  public void setProfile(java.util.List<com.bluekai.analytics.jobs.SiteCatNsTuple> value) {
    this.profile = value;
  }

  /** Creates a new UserProfileFlatAggregate RecordBuilder */
  public static com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder newBuilder() {
    return new com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder();
  }
  
  /** Creates a new UserProfileFlatAggregate RecordBuilder by copying an existing Builder */
  public static com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder newBuilder(com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder other) {
    return new com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder(other);
  }
  
  /** Creates a new UserProfileFlatAggregate RecordBuilder by copying an existing UserProfileFlatAggregate instance */
  public static com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder newBuilder(com.bluekai.analytics.jobs.UserProfileFlatAggregate other) {
    return new com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder(other);
  }
  
  /**
   * RecordBuilder for UserProfileFlatAggregate instances.
   */
  public static class Builder extends org.apache.avro.specific.SpecificRecordBuilderBase<UserProfileFlatAggregate>
    implements org.apache.avro.data.RecordBuilder<UserProfileFlatAggregate> {

    private java.lang.CharSequence uuid;
    private java.util.List<com.bluekai.analytics.jobs.SiteCatNsTuple> profile;

    /** Creates a new Builder */
    private Builder() {
      super(SCHEMA$);
    }
    
    /** Creates a Builder by copying an existing Builder */
    private Builder(com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder other) {
      super(other);
      if (isValidValue(fields()[0], other.uuid)) {
        this.uuid = data().deepCopy(fields()[0].schema(), other.uuid);
        fieldSetFlags()[0] = true;
      }
      if (isValidValue(fields()[1], other.profile)) {
        this.profile = data().deepCopy(fields()[1].schema(), other.profile);
        fieldSetFlags()[1] = true;
      }
    }
    
    /** Creates a Builder by copying an existing UserProfileFlatAggregate instance */
    private Builder(com.bluekai.analytics.jobs.UserProfileFlatAggregate other) {
            super(SCHEMA$);
      if (isValidValue(fields()[0], other.uuid)) {
        this.uuid = data().deepCopy(fields()[0].schema(), other.uuid);
        fieldSetFlags()[0] = true;
      }
      if (isValidValue(fields()[1], other.profile)) {
        this.profile = data().deepCopy(fields()[1].schema(), other.profile);
        fieldSetFlags()[1] = true;
      }
    }

    /** Gets the value of the 'uuid' field */
    public java.lang.CharSequence getUuid() {
      return uuid;
    }
    
    /** Sets the value of the 'uuid' field */
    public com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder setUuid(java.lang.CharSequence value) {
      validate(fields()[0], value);
      this.uuid = value;
      fieldSetFlags()[0] = true;
      return this; 
    }
    
    /** Checks whether the 'uuid' field has been set */
    public boolean hasUuid() {
      return fieldSetFlags()[0];
    }
    
    /** Clears the value of the 'uuid' field */
    public com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder clearUuid() {
      uuid = null;
      fieldSetFlags()[0] = false;
      return this;
    }

    /** Gets the value of the 'profile' field */
    public java.util.List<com.bluekai.analytics.jobs.SiteCatNsTuple> getProfile() {
      return profile;
    }
    
    /** Sets the value of the 'profile' field */
    public com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder setProfile(java.util.List<com.bluekai.analytics.jobs.SiteCatNsTuple> value) {
      validate(fields()[1], value);
      this.profile = value;
      fieldSetFlags()[1] = true;
      return this; 
    }
    
    /** Checks whether the 'profile' field has been set */
    public boolean hasProfile() {
      return fieldSetFlags()[1];
    }
    
    /** Clears the value of the 'profile' field */
    public com.bluekai.analytics.jobs.UserProfileFlatAggregate.Builder clearProfile() {
      profile = null;
      fieldSetFlags()[1] = false;
      return this;
    }

    @Override
    public UserProfileFlatAggregate build() {
      try {
        UserProfileFlatAggregate record = new UserProfileFlatAggregate();
        record.uuid = fieldSetFlags()[0] ? this.uuid : (java.lang.CharSequence) defaultValue(fields()[0]);
        record.profile = fieldSetFlags()[1] ? this.profile : (java.util.List<com.bluekai.analytics.jobs.SiteCatNsTuple>) defaultValue(fields()[1]);
        return record;
      } catch (Exception e) {
        throw new org.apache.avro.AvroRuntimeException(e);
      }
    }
  }
}