feat: add collation for Case sensitive string column (#2490)

Neenu1995 · gcf-owl-bot[bot] · web-flow · commit 325773757e81 · 2023-01-26T13:31:58.000-05:00
* feat: add collation for Case sensitive string column * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * 🦉 Updates from OwlBot post-processor See https://github.com/googleapis/repo-automation-bots/blob/main/packages/owl-bot/README.md * chore: fix typo Co-authored-by: Owl Bot <gcf-owl-bot[bot]@users.noreply.github.com>
diff --git a/README.md b/README.md
@@ -59,13 +59,13 @@ implementation 'com.google.cloud:google-cloud-bigquery'
 If you are using Gradle without BOM, add this to your dependencies:
 
 ```Groovy
-implementation 'com.google.cloud:google-cloud-bigquery:2.20.2'
+implementation 'com.google.cloud:google-cloud-bigquery:2.21.0'
 ```
 
 If you are using SBT, add this to your dependencies:
 
 ```Scala
-libraryDependencies += "com.google.cloud" % "google-cloud-bigquery" % "2.20.2"
+libraryDependencies += "com.google.cloud" % "google-cloud-bigquery" % "2.21.0"
 ```
 
 ## Authentication
diff --git a/google-cloud-bigquery/clirr-ignored-differences.xml b/google-cloud-bigquery/clirr-ignored-differences.xml
@@ -34,4 +34,14 @@
     <className>com/google/cloud/bigquery/Connection</className>
     <method>com.google.common.util.concurrent.ListenableFuture executeSelectAsync(java.lang.String, java.util.List, java.util.Map[])</method>
   </difference>
+  <difference>
+    <differenceType>7013</differenceType>
+    <className>com/google/cloud/bigquery/DatasetInfo*</className>
+    <method>*DefaultCollation(*)</method>
+  </difference>
+  <difference>
+    <differenceType>7013</differenceType>
+    <className>com/google/cloud/bigquery/TableInfo*</className>
+    <method>*DefaultCollation(*)</method>
+  </difference>
   </differences>
diff --git a/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/Dataset.java b/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/Dataset.java
@@ -146,6 +146,12 @@ public Builder setDefaultPartitionExpirationMs(Long defaultPartitionExpirationMs
       return this;
     }
 
+    @Override
+    public Builder setDefaultCollation(String defaultCollation) {
+      infoBuilder.setDefaultCollation(defaultCollation);
+      return this;
+    }
+
     @Override
     public Dataset build() {
       return new Dataset(bigquery, infoBuilder);
diff --git a/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/DatasetInfo.java b/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/DatasetInfo.java
@@ -72,6 +72,7 @@ public Dataset apply(DatasetInfo datasetInfo) {
   private final Labels labels;
   private final EncryptionConfiguration defaultEncryptionConfiguration;
   private final Long defaultPartitionExpirationMs;
+  private final String defaultCollation;
 
   /** A builder for {@code DatasetInfo} objects. */
   public abstract static class Builder {
@@ -148,6 +149,19 @@ public abstract Builder setDefaultEncryptionConfiguration(
      */
     public abstract Builder setDefaultPartitionExpirationMs(Long defaultPartitionExpirationMs);
 
+    /**
+     * Optional. Defines the default collation specification of future tables created in the
+     * dataset. If a table is created in this dataset without table-level default collation, then
+     * the table inherits the dataset default collation, which is applied to the string fields that
+     * do not have explicit collation specified. A change to this field affects only tables created
+     * afterwards, and does not alter the existing tables. The following values are supported:
+     *
+     * <p>* 'und:ci': undetermined locale, case insensitive. * '': empty string. Default to
+     * case-sensitive behavior. (-- A wrapper is used here because it is possible to set the value
+     * to the empty string. --) (-- api-linter: standard-fields=disabled --)
+     */
+    public abstract Builder setDefaultCollation(String defaultCollation);
+
     /** Creates a {@code DatasetInfo} object. */
     public abstract DatasetInfo build();
   }
@@ -168,6 +182,7 @@ static final class BuilderImpl extends Builder {
     private Labels labels = Labels.ZERO;
     private EncryptionConfiguration defaultEncryptionConfiguration;
     private Long defaultPartitionExpirationMs;
+    private String defaultCollation;
 
     BuilderImpl() {}
 
@@ -186,6 +201,7 @@ static final class BuilderImpl extends Builder {
       this.labels = datasetInfo.labels;
       this.defaultEncryptionConfiguration = datasetInfo.defaultEncryptionConfiguration;
       this.defaultPartitionExpirationMs = datasetInfo.defaultPartitionExpirationMs;
+      this.defaultCollation = datasetInfo.defaultCollation;
     }
 
     BuilderImpl(com.google.api.services.bigquery.model.Dataset datasetPb) {
@@ -219,6 +235,7 @@ public Acl apply(Dataset.Access accessPb) {
                 .build();
       }
       this.defaultPartitionExpirationMs = datasetPb.getDefaultPartitionExpirationMs();
+      this.defaultCollation = datasetPb.getDefaultCollation();
     }
 
     @Override
@@ -313,6 +330,12 @@ public Builder setDefaultPartitionExpirationMs(Long defaultPartitionExpirationMs
       return this;
     }
 
+    @Override
+    public Builder setDefaultCollation(String defaultCollation) {
+      this.defaultCollation = defaultCollation;
+      return this;
+    }
+
     @Override
     public DatasetInfo build() {
       return new DatasetInfo(this);
@@ -334,6 +357,7 @@ public DatasetInfo build() {
     labels = builder.labels;
     defaultEncryptionConfiguration = builder.defaultEncryptionConfiguration;
     defaultPartitionExpirationMs = builder.defaultPartitionExpirationMs;
+    defaultCollation = builder.defaultCollation;
   }
 
   /** Returns the dataset identity. */
@@ -459,6 +483,10 @@ public Long getDefaultPartitionExpirationMs() {
     return defaultPartitionExpirationMs;
   }
 
+  public String getDefaultCollation() {
+    return defaultCollation;
+  }
+
   /** Returns a builder for the dataset object. */
   public Builder toBuilder() {
     return new BuilderImpl(this);
@@ -481,6 +509,7 @@ public String toString() {
         .add("labels", labels)
         .add("defaultEncryptionConfiguration", defaultEncryptionConfiguration)
         .add("defaultPartitionExpirationMs", defaultPartitionExpirationMs)
+        .add("defaultCollation", defaultCollation)
         .toString();
   }
 
@@ -556,6 +585,9 @@ public Dataset.Access apply(Acl acl) {
     if (defaultPartitionExpirationMs != null) {
       datasetPb.setDefaultPartitionExpirationMs(defaultPartitionExpirationMs);
     }
+    if (defaultCollation != null) {
+      datasetPb.setDefaultCollation(defaultCollation);
+    }
     return datasetPb;
   }
 
diff --git a/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/Field.java b/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/Field.java
@@ -63,6 +63,7 @@ public TableFieldSchema apply(Field field) {
   private final Long scale;
   private final Long precision;
   private final String defaultValueExpression;
+  private final String collation;
 
   /**
    * Mode for a BigQuery Table field. {@link Mode#NULLABLE} fields can be set to {@code null},
@@ -87,6 +88,7 @@ public static final class Builder {
     private Long scale;
     private Long precision;
     private String defaultValueExpression;
+    private String collation;
 
     private Builder() {}
 
@@ -101,6 +103,7 @@ private Builder(Field field) {
       this.scale = field.scale;
       this.precision = field.precision;
       this.defaultValueExpression = field.defaultValueExpression;
+      this.collation = field.collation;
     }
 
     /**
@@ -285,6 +288,19 @@ public Builder setDefaultValueExpression(String defaultValueExpression) {
       return this;
     }
 
+    /**
+     * Optional. Field collation can be set only when the type of field is STRING. The following
+     * values are supported:
+     *
+     * <p>* 'und:ci': undetermined locale, case insensitive. * '': empty string. Default to
+     * case-sensitive behavior. (-- A wrapper is used here because it is possible to set the value
+     * to the empty string. --)
+     */
+    public Builder setCollation(String collation) {
+      this.collation = collation;
+      return this;
+    }
+
     /** Creates a {@code Field} object. */
     public Field build() {
       return new Field(this);
@@ -302,6 +318,7 @@ private Field(Builder builder) {
     this.scale = builder.scale;
     this.precision = builder.precision;
     this.defaultValueExpression = builder.defaultValueExpression;
+    this.collation = builder.collation;
   }
 
   /** Returns the field name. */
@@ -357,6 +374,10 @@ public String getDefaultValueExpression() {
     return defaultValueExpression;
   }
 
+  public String getCollation() {
+    return collation;
+  }
+
   /**
    * Returns the list of sub-fields if {@link #getType()} is a {@link LegacySQLTypeName#RECORD}.
    * Returns {@code null} otherwise.
@@ -382,6 +403,7 @@ public String toString() {
         .add("scale", scale)
         .add("precision", precision)
         .add("defaultValueExpression", defaultValueExpression)
+        .add("collation", collation)
         .toString();
   }
 
@@ -468,6 +490,9 @@ TableFieldSchema toPb() {
       List<TableFieldSchema> fieldsPb = Lists.transform(getSubFields(), TO_PB_FUNCTION);
       fieldSchemaPb.setFields(fieldsPb);
     }
+    if (collation != null) {
+      fieldSchemaPb.setCollation(collation);
+    }
     return fieldSchemaPb;
   }
 
@@ -500,6 +525,9 @@ static Field fromPb(TableFieldSchema fieldSchemaPb) {
             ? FieldList.of(Lists.transform(fieldSchemaPb.getFields(), FROM_PB_FUNCTION))
             : null;
     fieldBuilder.setType(LegacySQLTypeName.valueOf(fieldSchemaPb.getType()), subFields);
+    if (fieldSchemaPb.getCollation() != null) {
+      fieldBuilder.setCollation(fieldSchemaPb.getCollation());
+    }
     return fieldBuilder.build();
   }
 }
diff --git a/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/Table.java b/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/Table.java
@@ -156,6 +156,12 @@ public Builder setRequirePartitionFilter(Boolean requirePartitionFilter) {
       return this;
     }
 
+    @Override
+    public Builder setDefaultCollation(String defaultCollation) {
+      infoBuilder.setDefaultCollation(defaultCollation);
+      return this;
+    }
+
     @Override
     public Table build() {
       return new Table(bigquery, infoBuilder);
diff --git a/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/TableInfo.java b/google-cloud-bigquery/src/main/java/com/google/cloud/bigquery/TableInfo.java
@@ -72,6 +72,7 @@ public Table apply(TableInfo tableInfo) {
   private final EncryptionConfiguration encryptionConfiguration;
   private final Labels labels;
   private final Boolean requirePartitionFilter;
+  private final String defaultCollation;
 
   /** A builder for {@code TableInfo} objects. */
   public abstract static class Builder {
@@ -135,6 +136,8 @@ public abstract static class Builder {
     public Builder setRequirePartitionFilter(Boolean requirePartitionFilter) {
       return this;
     }
+
+    public abstract Builder setDefaultCollation(String defaultCollation);
   }
 
   static class BuilderImpl extends Builder {
@@ -155,6 +158,7 @@ static class BuilderImpl extends Builder {
     private EncryptionConfiguration encryptionConfiguration;
     private Labels labels = Labels.ZERO;
     private Boolean requirePartitionFilter;
+    private String defaultCollation;
 
     BuilderImpl() {}
 
@@ -175,6 +179,7 @@ static class BuilderImpl extends Builder {
       this.encryptionConfiguration = tableInfo.encryptionConfiguration;
       this.labels = tableInfo.labels;
       this.requirePartitionFilter = tableInfo.requirePartitionFilter;
+      this.defaultCollation = tableInfo.defaultCollation;
     }
 
     BuilderImpl(Table tablePb) {
@@ -199,6 +204,7 @@ static class BuilderImpl extends Builder {
       }
       this.labels = Labels.fromPb(tablePb.getLabels());
       this.requirePartitionFilter = tablePb.getRequirePartitionFilter();
+      this.defaultCollation = tablePb.getDefaultCollation();
     }
 
     @Override
@@ -297,6 +303,12 @@ public Builder setRequirePartitionFilter(Boolean requirePartitionFilter) {
       return this;
     }
 
+    @Override
+    public Builder setDefaultCollation(String defaultCollation) {
+      this.defaultCollation = defaultCollation;
+      return this;
+    }
+
     @Override
     public TableInfo build() {
       return new TableInfo(this);
@@ -318,8 +330,9 @@ public TableInfo build() {
     this.numRows = builder.numRows;
     this.definition = builder.definition;
     this.encryptionConfiguration = builder.encryptionConfiguration;
-    labels = builder.labels;
+    this.labels = builder.labels;
     this.requirePartitionFilter = builder.requirePartitionFilter;
+    this.defaultCollation = builder.defaultCollation;
   }
 
   /** Returns the hash of the table resource. */
@@ -422,6 +435,10 @@ public Boolean getRequirePartitionFilter() {
     return requirePartitionFilter;
   }
 
+  public String getDefaultCollation() {
+    return defaultCollation;
+  }
+
   /** Returns a builder for the table object. */
   public Builder toBuilder() {
     return new BuilderImpl(this);
@@ -446,6 +463,7 @@ public String toString() {
         .add("encryptionConfiguration", encryptionConfiguration)
         .add("labels", labels)
         .add("requirePartitionFilter", requirePartitionFilter)
+        .add("defaultCollation", defaultCollation)
         .toString();
   }
 
@@ -507,6 +525,9 @@ Table toPb() {
     }
     tablePb.setLabels(labels.toPb());
     tablePb.setRequirePartitionFilter(requirePartitionFilter);
+    if (defaultCollation != null) {
+      tablePb.setDefaultCollation(defaultCollation);
+    }
     return tablePb;
   }
 
diff --git a/google-cloud-bigquery/src/test/java/com/google/cloud/bigquery/it/ITBigQueryTest.java b/google-cloud-bigquery/src/test/java/com/google/cloud/bigquery/it/ITBigQueryTest.java
