Create index command implementation (#1526)

src/main/java/io/stargate/sgv2/jsonapi/api/model/command/CollectionCommand.java

@@ -20,7 +20,8 @@
   @JsonSubTypes.Type(value = InsertManyCommand.class),
   @JsonSubTypes.Type(value = UpdateManyCommand.class),
   @JsonSubTypes.Type(value = UpdateOneCommand.class),
-  @JsonSubTypes.Type(value = AddIndexCommand.class),
+  // We have only collection resource that is used for api tables
+  @JsonSubTypes.Type(value = CreateIndexCommand.class),
   @JsonSubTypes.Type(value = DropIndexCommand.class),
 })
 public interface CollectionCommand extends Command {}

src/main/java/io/stargate/sgv2/jsonapi/api/model/command/Command.java

@@ -39,11 +39,11 @@ public interface Command {
 
   /** Enum class for API command name. This is what user uses for command json body. */
   enum CommandName {
-    ADD_INDEX("addIndex"),
     COUNT_DOCUMENTS("countDocuments"),
     CREATE_COLLECTION("createCollection"),
-    CREATE_NAMESPACE("createNamespace"),
+    CREATE_INDEX("createIndex"),
     CREATE_KEYSPACE("createKeyspace"),
+    CREATE_NAMESPACE("createNamespace"),
     CREATE_TABLE("createTable"),
     DELETE_COLLECTION("deleteCollection"),
     DELETE_MANY("deleteMany"),

src/main/java/io/stargate/sgv2/jsonapi/api/model/command/impl/CreateIndexCommand.java

@@ -0,0 +1,94 @@
+package io.stargate.sgv2.jsonapi.api.model.command.impl;
+
+import com.fasterxml.jackson.annotation.JsonTypeName;
+import io.stargate.sgv2.jsonapi.api.model.command.CollectionCommand;
+import io.stargate.sgv2.jsonapi.service.schema.SimilarityFunction;
+import jakarta.annotation.Nullable;
+import jakarta.validation.constraints.NotNull;
+import jakarta.validation.constraints.Pattern;
+import jakarta.validation.constraints.Size;
+import org.eclipse.microprofile.openapi.annotations.enums.SchemaType;
+import org.eclipse.microprofile.openapi.annotations.media.Schema;
+
+// TODO, hide table feature detail before it goes public
+@Schema(description = "Command that creates an index for a column in a table.")
+@JsonTypeName("createIndex")
+public record CreateIndexCommand(
+    @NotNull
+        @Size(min = 1, max = 48)
+        @Pattern(regexp = "[a-zA-Z][a-zA-Z0-9_]*")
+        @Schema(description = "Name of the column to create the index on")
+        String name,
+    @NotNull
+        @Schema(
+            description = "Definition for created index for a column.",
+            type = SchemaType.OBJECT)
+        Definition definition,
+    @Nullable @Schema(description = "Creating index command option.", type = SchemaType.OBJECT)
+        Options options)
+    implements CollectionCommand {
+  public record Definition(
+      @NotNull
+          @Size(min = 1, max = 48)
+          @Pattern(regexp = "[a-zA-Z][a-zA-Z0-9_]*")
+          @Schema(description = "Name of the column for which index to be created.")
+          String column,
+      @Nullable @Schema(description = "Different indexing options.", type = SchemaType.OBJECT)
+          Options options) {
+    // This is index definition options for text and vector column types.
+    public record Options(
+        @Nullable
+            @Schema(
+                description = "Ignore case in matching string values.",
+                defaultValue = "true",
+                type = SchemaType.BOOLEAN,
+                implementation = Boolean.class)
+            Boolean caseSensitive,
+        @Nullable
+            @Schema(
+                description = "When set to true, perform Unicode normalization on indexed strings.",
+                defaultValue = "false",
+                type = SchemaType.BOOLEAN,
+                implementation = Boolean.class)
+            Boolean normalize,
+        @Nullable
+            @Schema(
+                description =
+                    "When set to true, index will converts alphabetic, numeric, and symbolic characters to the ascii equivalent, if one exists.",
+                defaultValue = "false",
+                type = SchemaType.BOOLEAN,
+                implementation = Boolean.class)
+            Boolean ascii,
+        @Nullable
+            @Pattern(
+                regexp = "(dot_product|cosine|euclidean)",
+                message = "function name can only be 'dot_product', 'cosine' or 'euclidean'")
+            @Schema(
+                description =
+                    "Similarity function algorithm that needs to be used for vector search",
+                defaultValue = "cosine",
+                type = SchemaType.STRING,
+                implementation = String.class)
+            SimilarityFunction metric,
+        @Nullable
+            @Size(min = 1, max = 48)
+            @Pattern(regexp = "[a-zA-Z][a-zA-Z0-9_]*")
+            @Schema(description = "Model name used to generate the embeddings.")
+            String sourceModel) {}
+  }
+
+  // This is index command option irrespective of column definition.
+  public record Options(
+      @Schema(
+              description = "Flag to ignore if index already exists",
+              defaultValue = "false",
+              type = SchemaType.BOOLEAN,
+              implementation = Boolean.class)
+          Boolean ifNotExists) {}
+
+  /** {@inheritDoc} */
+  @Override
+  public CommandName commandName() {
+    return CommandName.CREATE_INDEX;
+  }
+}

src/main/java/io/stargate/sgv2/jsonapi/api/v1/CollectionResource.java

@@ -109,10 +109,10 @@ public CollectionResource(MeteredCommandProcessor meteredCommandProcessor) {
                         InsertOneCommand.class,
                         InsertManyCommand.class,
                         UpdateManyCommand.class,
-                        UpdateOneCommand.class
+                        UpdateOneCommand.class,
                         // TODO, hide table feature detail before it goes public,
                         // https://github.com/stargate/data-api/pull/1360
-                        //                        AddIndexCommand.class,
+                        //                        CreateIndexCommand.class,
                         //                        DropIndexCommand.class
                       }),
               examples = {

src/main/java/io/stargate/sgv2/jsonapi/config/constants/VectorConstant.java

@@ -0,0 +1,14 @@
+package io.stargate.sgv2.jsonapi.config.constants;
+
+import io.smallrye.config.ConfigMapping;
+import java.util.Set;
+
+@ConfigMapping(prefix = "stargate.jsonapi.vector")
+public interface VectorConstant {
+  /*
+  Supported Source Models for Vector Index in Cassandra
+   */
+  Set<String> SUPPORTED_SOURCES =
+      Set.of(
+          "ada002", "openai_v3_small", "openai_v3_large", "bert", "gecko", "nv_qa_4", "cohere_v3");
+}

src/main/java/io/stargate/sgv2/jsonapi/exception/SchemaException.java

@@ -18,6 +18,7 @@ public enum Code implements ErrorCode<SchemaException> {
     COLUMN_TYPE_INCORRECT,
     COLUMN_TYPE_UNSUPPORTED,
     INVALID_CONFIGURATION,
+    INVALID_INDEX_DEFINITION,
     INVALID_VECTORIZE_CONFIGURATION,
     LIST_TYPE_INCORRECT_DEFINITION,
     MAP_TYPE_INCORRECT_DEFINITION,

src/main/java/io/stargate/sgv2/jsonapi/service/cqldriver/override/ExtendedCreateIndex.java

@@ -0,0 +1,90 @@
+package io.stargate.sgv2.jsonapi.service.cqldriver.override;
+
+import com.datastax.oss.driver.api.core.CqlIdentifier;
+import com.datastax.oss.driver.internal.querybuilder.CqlHelper;
+import com.datastax.oss.driver.internal.querybuilder.schema.DefaultCreateIndex;
+import com.datastax.oss.driver.internal.querybuilder.schema.OptionsUtils;
+import com.datastax.oss.driver.shaded.guava.common.collect.ImmutableMap;
+import com.datastax.oss.driver.shaded.guava.common.collect.UnmodifiableIterator;
+import java.util.Map;
+
+/**
+ * An extension of the {@link DefaultCreateIndex} class, This is needed because the column name
+ * appended to the builder needs to use `asCql(true)` to keep the quotes.
+ */
+public class ExtendedCreateIndex extends DefaultCreateIndex {
+
+  public ExtendedCreateIndex(DefaultCreateIndex defaultCreateIndex) {
+
+    super(
+        defaultCreateIndex.getIndex(),
+        defaultCreateIndex.isIfNotExists(),
+        defaultCreateIndex.getKeyspace(),
+        defaultCreateIndex.getTable(),
+        defaultCreateIndex.getColumnToIndexType(),
+        defaultCreateIndex.getUsingClass(),
+        // This is fine as the internal options object is ImmutableMap
+        (ImmutableMap<String, Object>) defaultCreateIndex.getOptions());
+  }
+
+  @Override
+  public String asCql() {
+    StringBuilder builder = new StringBuilder("CREATE ");
+    if (this.getUsingClass() != null) {
+      builder.append("CUSTOM ");
+    }
+
+    builder.append("INDEX");
+    if (this.isIfNotExists()) {
+      builder.append(" IF NOT EXISTS");
+    }
+
+    if (this.getIndex() != null) {
+      builder.append(' ').append(this.getIndex().asCql(true));
+    }
+
+    if (this.getTable() == null) {
+      return builder.toString();
+    } else {
+      builder.append(" ON ");
+      CqlHelper.qualify(this.getKeyspace(), this.getTable(), builder);
+      if (this.getColumnToIndexType().isEmpty()) {
+        return builder.toString();
+      } else {
+        builder.append(" (");
+        boolean firstColumn = true;
+        UnmodifiableIterator var3 = this.getColumnToIndexType().entrySet().iterator();
+
+        while (var3.hasNext()) {
+          Map.Entry<CqlIdentifier, String> entry = (Map.Entry) var3.next();
+          if (firstColumn) {
+            firstColumn = false;
+          } else {
+            builder.append(",");
+          }
+
+          if (((String) entry.getValue()).equals("__NO_INDEX_TYPE")) {
+            builder.append(entry.getKey().asCql(true));
+          } else {
+            builder
+                .append((String) entry.getValue())
+                .append("(")
+                .append(entry.getKey().asCql(true))
+                .append(")");
+          }
+        }
+
+        builder.append(")");
+        if (this.getUsingClass() != null) {
+          builder.append(" USING '").append(this.getUsingClass()).append('\'');
+        }
+
+        if (!this.getOptions().isEmpty()) {
+          builder.append(OptionsUtils.buildOptions(this.getOptions(), true));
+        }
+
+        return builder.toString();
+      }
+    }
+  }
+}

src/main/java/io/stargate/sgv2/jsonapi/service/cqldriver/override/ExtendedVectorType.java

@@ -0,0 +1,19 @@
+package io.stargate.sgv2.jsonapi.service.cqldriver.override;
+
+import com.datastax.oss.driver.api.core.type.DataType;
+import com.datastax.oss.driver.internal.core.type.DefaultVectorType;
+
+/**
+ * Extended vector type to support vector size This is needed because java drivers
+ * DataTypes.vectorOf() method has a bug
+ */
+public class ExtendedVectorType extends DefaultVectorType {
+  public ExtendedVectorType(DataType subtype, int vectorSize) {
+    super(subtype, vectorSize);
+  }
+
+  @Override
+  public String asCql(boolean includeFrozen, boolean pretty) {
+    return "VECTOR<" + getElementType().asCql(includeFrozen, pretty) + "," + getDimensions() + ">";
+  }
+}