Support for File Search Customizations

README.md

@@ -649,7 +649,7 @@ public class ConversationV2Demo {
                         .model("gpt-4o")
                         .instructions("You are a skilled tutor on geo-politic topics.")
                         .tools(functionExecutor.getToolFunctions())
-                        .tool(AssistantTool.FILE_SEARCH)
+                        .tool(AssistantTool.fileSearch())
                         .toolResources(ToolResourceFull.builder()
                                 .fileSearch(FileSearch.builder()
                                         .vectorStoreId(vectorStoreId)

pom.xml

@@ -6,7 +6,7 @@
 
   <groupId>io.github.sashirestela</groupId>
   <artifactId>simple-openai</artifactId>
-  <version>3.3.0</version>
+  <version>3.4.0</version>
   <packaging>jar</packaging>
 
   <name>simple-openai</name>
@@ -61,16 +61,16 @@
     <mockito.version>5.12.0</mockito.version>
     <!-- Plugins Versions -->
     <compiler.version>3.13.0</compiler.version>
-    <enforcer.version>3.4.1</enforcer.version>
+    <enforcer.version>3.5.0</enforcer.version>
     <surefire.version>3.2.5</surefire.version>
     <exec.version>3.3.0</exec.version>
     <jacoco.version>0.8.12</jacoco.version>
     <helper.version>3.6.0</helper.version>
     <versions.version>2.16.2</versions.version>
     <source.version>3.3.1</source.version>
-    <javadoc.version>3.6.3</javadoc.version>
+    <javadoc.version>3.7.0</javadoc.version>
     <gpg.version>3.1.0</gpg.version>
-    <sonatype.version>1.6.13</sonatype.version>
+    <sonatype.version>1.7.0</sonatype.version>
     <spotless.version>2.43.0</spotless.version>
   </properties>
 

src/demo/java/io/github/sashirestela/openai/demo/AssistantV2Demo.java

@@ -4,13 +4,27 @@
 import io.github.sashirestela.openai.domain.assistant.AssistantModifyRequest;
 import io.github.sashirestela.openai.domain.assistant.AssistantRequest;
 import io.github.sashirestela.openai.domain.assistant.AssistantTool;
+import io.github.sashirestela.openai.domain.assistant.ChunkingStrategy;
+import io.github.sashirestela.openai.domain.assistant.ChunkingStrategy.StaticChunking;
+import io.github.sashirestela.openai.domain.assistant.ToolResourceFull;
+import io.github.sashirestela.openai.domain.assistant.ToolResourceFull.FileSearch;
+import io.github.sashirestela.openai.domain.assistant.ToolResourceFull.FileSearch.VectorStore;
+import io.github.sashirestela.openai.domain.file.FileRequest.PurposeType;
 
 import java.util.Map;
 
 public class AssistantV2Demo extends AbstractDemo {
 
+    private FileDemo fileDemo;
+    private String fileId;
     private String assistantId;
 
+    public AssistantV2Demo() {
+        fileDemo = new FileDemo();
+        var file = fileDemo.createFile("src/demo/resources/mistral-ai.txt", PurposeType.ASSISTANTS);
+        fileId = file.getId();
+    }
+
     public void createAssistant() {
         var assistantRequest = AssistantRequest.builder()
                 .model("gpt-4-turbo")
@@ -19,7 +33,18 @@ public void createAssistant() {
                 .instructions("You are a very kind assistant. If you cannot find correct facts to answer the "
                         + "questions, you have to refer to the attached files or use the functions provided. "
                         + "Finally, if you receive math questions, you must write and run code to answer them.")
-                .tool(AssistantTool.FILE_SEARCH)
+                .tool(AssistantTool.fileSearch(10))
+                .toolResources(ToolResourceFull.builder()
+                        .fileSearch(FileSearch.builder()
+                                .vectorStore(VectorStore.builder()
+                                        .fileId(fileId)
+                                        .chunkingStrategy(ChunkingStrategy.staticType(StaticChunking.builder()
+                                                .maxChunkSizeTokens(100)
+                                                .chunkOverlapTokens(50)
+                                                .build()))
+                                        .build())
+                                .build())
+                        .build())
                 .metadata(Map.of("user", "tester"))
                 .temperature(0.2)
                 .responseFormat("auto")
@@ -50,6 +75,15 @@ public void listAssistants() {
     }
 
     public void deleteAssistant() {
+        var assistant = openAI.assistants().getOne(assistantId).join();
+        var vectorStoreId = assistant.getToolResources().getFileSearch().getVectorStoreIds().get(0);
+
+        var deletedFile = fileDemo.deleteFile(fileId);
+        System.out.println(deletedFile);
+
+        var deletedVectorStore = openAI.vectorStores().delete(vectorStoreId).join();
+        System.out.println(deletedVectorStore);
+
         var deletedAssistant = openAI.assistants().delete(assistantId).join();
         System.out.println(deletedAssistant);
     }

src/demo/java/io/github/sashirestela/openai/demo/ConversationV2Demo.java

@@ -80,7 +80,7 @@ public void prepareConversation() {
                         .model("gpt-4o")
                         .instructions("You are a skilled tutor on geo-politic topics.")
                         .tools(functionExecutor.getToolFunctions())
-                        .tool(AssistantTool.FILE_SEARCH)
+                        .tool(AssistantTool.fileSearch())
                         .toolResources(ToolResourceFull.builder()
                                 .fileSearch(FileSearch.builder()
                                         .vectorStoreId(vectorStoreId)

src/demo/java/io/github/sashirestela/openai/demo/ThreadRunV2Demo.java

@@ -72,7 +72,7 @@ private void prepareDemo() {
                                 + "questions, you have to refer to the attached files or use the functions provided. "
                                 + "Finally, if you receive math questions, you must write and run code to answer them.")
                         .tools(functionExecutor.getToolFunctions())
-                        .tool(AssistantTool.FILE_SEARCH)
+                        .tool(AssistantTool.fileSearch())
                         .toolResources(ToolResourceFull.builder()
                                 .fileSearch(FileSearch.builder()
                                         .vectorStoreId(vectorStoreId)

src/main/java/io/github/sashirestela/openai/domain/assistant/AssistantTool.java

@@ -1,8 +1,11 @@
 package io.github.sashirestela.openai.domain.assistant;
 
 import com.fasterxml.jackson.annotation.JsonInclude;
+import com.fasterxml.jackson.databind.PropertyNamingStrategies;
+import com.fasterxml.jackson.databind.annotation.JsonNaming;
 import io.github.sashirestela.openai.common.tool.Tool;
 import io.github.sashirestela.openai.common.tool.ToolType;
+import io.github.sashirestela.slimvalidator.constraints.Range;
 import lombok.Getter;
 import lombok.NoArgsConstructor;
 import lombok.ToString;
@@ -11,13 +14,42 @@
 @Getter
 @ToString
 @JsonInclude(JsonInclude.Include.NON_NULL)
+@JsonNaming(PropertyNamingStrategies.SnakeCaseStrategy.class)
 public class AssistantTool extends Tool {
 
-    public static final AssistantTool CODE_INTERPRETER = new AssistantTool(ToolType.CODE_INTERPRETER, null);
-    public static final AssistantTool FILE_SEARCH = new AssistantTool(ToolType.FILE_SEARCH, null);
+    private FileSearch fileSearch;
+
+    private AssistantTool(ToolType type, FileSearch fileSearch) {
+        super(type, null);
+        this.fileSearch = fileSearch;
+    }
+
+    public static AssistantTool codeInterpreter() {
+        return new AssistantTool(ToolType.CODE_INTERPRETER, null);
+    }
+
+    public static AssistantTool fileSearch() {
+        return new AssistantTool(ToolType.FILE_SEARCH, null);
+    }
+
+    public static AssistantTool fileSearch(Integer maxNumResults) {
+        return new AssistantTool(ToolType.FILE_SEARCH, new FileSearch(maxNumResults));
+    }
+
+    @NoArgsConstructor
+    @Getter
+    @ToString
+    @JsonInclude(JsonInclude.Include.NON_NULL)
+    @JsonNaming(PropertyNamingStrategies.SnakeCaseStrategy.class)
+    public static class FileSearch {
+
+        @Range(min = 1, max = 50)
+        private Integer maxNumResults;
+
+        public FileSearch(Integer maxNumResults) {
+            this.maxNumResults = maxNumResults;
+        }
 
-    private AssistantTool(ToolType type, ToolFunctionDef function) {
-        super(type, function);
     }
 
 }

src/main/java/io/github/sashirestela/openai/domain/assistant/ChunkingStrategy.java

@@ -0,0 +1,66 @@
+package io.github.sashirestela.openai.domain.assistant;
+
+import com.fasterxml.jackson.annotation.JsonInclude;
+import com.fasterxml.jackson.annotation.JsonInclude.Include;
+import com.fasterxml.jackson.annotation.JsonProperty;
+import com.fasterxml.jackson.databind.PropertyNamingStrategies;
+import com.fasterxml.jackson.databind.annotation.JsonNaming;
+import io.github.sashirestela.slimvalidator.constraints.Range;
+import io.github.sashirestela.slimvalidator.constraints.Required;
+import lombok.Builder;
+import lombok.Getter;
+import lombok.NoArgsConstructor;
+
+@NoArgsConstructor
+@Getter
+@JsonInclude(Include.NON_EMPTY)
+@JsonNaming(PropertyNamingStrategies.SnakeCaseStrategy.class)
+public class ChunkingStrategy {
+
+    private ChunkingStrategyType type;
+
+    @JsonProperty("static")
+    private StaticChunking staticChunking;
+
+    private ChunkingStrategy(ChunkingStrategyType type, StaticChunking staticChunking) {
+        this.type = type;
+        this.staticChunking = staticChunking;
+    }
+
+    public static ChunkingStrategy autoType() {
+        return new ChunkingStrategy(ChunkingStrategyType.AUTO, null);
+    }
+
+    public static ChunkingStrategy staticType(StaticChunking staticChunking) {
+        return new ChunkingStrategy(ChunkingStrategyType.STATIC, staticChunking);
+    }
+
+    public enum ChunkingStrategyType {
+
+        @JsonProperty("auto")
+        AUTO,
+
+        @JsonProperty("static")
+        STATIC,
+
+        @JsonProperty("other")
+        OTHER;
+    }
+
+    @Getter
+    @Builder
+    @JsonInclude(Include.NON_EMPTY)
+    @JsonNaming(PropertyNamingStrategies.SnakeCaseStrategy.class)
+    public static class StaticChunking {
+
+        @Required
+        @Range(min = 100, max = 4_096)
+        private Integer maxChunkSizeTokens;
+
+        @Required
+        @Range(max = 2_048)
+        private Integer chunkOverlapTokens;
+
+    }
+
+}

src/main/java/io/github/sashirestela/openai/domain/assistant/ToolResource.java

@@ -11,6 +11,7 @@
 import lombok.NoArgsConstructor;
 import lombok.Singular;
 import lombok.ToString;
+import lombok.experimental.SuperBuilder;
 
 import java.util.List;
 
@@ -45,7 +46,7 @@ public static class CodeInterpreter {
     @NoArgsConstructor
     @Getter
     @ToString
-    @Builder
+    @SuperBuilder
     @JsonInclude(Include.NON_EMPTY)
     @JsonNaming(PropertyNamingStrategies.SnakeCaseStrategy.class)
     public static class FileSearch {

src/main/java/io/github/sashirestela/openai/domain/assistant/ToolResourceFull.java

@@ -8,6 +8,7 @@
 import lombok.Builder;
 import lombok.Getter;
 import lombok.Singular;
+import lombok.experimental.SuperBuilder;
 
 import java.util.List;
 import java.util.Map;
@@ -18,30 +19,14 @@
 @JsonNaming(PropertyNamingStrategies.SnakeCaseStrategy.class)
 public class ToolResourceFull {
 
-    private CodeInterpreter codeInterpreter;
+    private ToolResource.CodeInterpreter codeInterpreter;
     private FileSearch fileSearch;
 
     @Getter
-    @Builder
+    @SuperBuilder
     @JsonInclude(Include.NON_EMPTY)
     @JsonNaming(PropertyNamingStrategies.SnakeCaseStrategy.class)
-    public static class CodeInterpreter {
-
-        @Singular
-        @Size(max = 20)
-        private List<String> fileIds;
-
-    }
-
-    @Getter
-    @Builder
-    @JsonInclude(Include.NON_EMPTY)
-    @JsonNaming(PropertyNamingStrategies.SnakeCaseStrategy.class)
-    public static class FileSearch {
-
-        @Singular
-        @Size(max = 1)
-        private List<String> vectorStoreIds;
+    public static class FileSearch extends ToolResource.FileSearch {
 
         @Singular
         @Size(max = 1)
@@ -57,6 +42,8 @@ public static class VectorStore {
             @Size(max = 10_000)
             private List<String> fileIds;
 
+            private ChunkingStrategy chunkingStrategy;
+
             @Size(max = 16)
             private Map<String, String> metadata;
 

src/main/java/io/github/sashirestela/openai/domain/assistant/VectorStoreFile.java

@@ -19,5 +19,6 @@ public class VectorStoreFile {
     private String vectorStoreId;
     private FileStatus status;
     private LastError lastError;
+    private ChunkingStrategy chunkingStrategy;
 
 }

src/main/java/io/github/sashirestela/openai/domain/assistant/VectorStoreFileBatchRequest.java

@@ -23,4 +23,6 @@ public class VectorStoreFileBatchRequest {
     @Size(min = 1, max = 500)
     private List<String> fileIds;
 
+    private ChunkingStrategy chunkingStrategy;
+
 }

src/main/java/io/github/sashirestela/openai/domain/assistant/VectorStoreFileRequest.java

@@ -17,4 +17,6 @@ public class VectorStoreFileRequest {
     @Required
     private String fileId;
 
+    private ChunkingStrategy chunkingStrategy;
+
 }

src/main/java/io/github/sashirestela/openai/domain/assistant/VectorStoreRequest.java

@@ -26,6 +26,8 @@ public class VectorStoreRequest {
 
     private ExpiresAfter expiresAfter;
 
+    private ChunkingStrategy chunkingStrategy;
+
     @Size(max = 16)
     private Map<String, String> metadata;
 

src/test/java/io/github/sashirestela/openai/domain/assistant/AssistantDomainTest.java

@@ -3,6 +3,9 @@
 import io.github.sashirestela.openai.SimpleOpenAI;
 import io.github.sashirestela.openai.common.ResponseFormat;
 import io.github.sashirestela.openai.domain.DomainTestingHelper;
+import io.github.sashirestela.openai.domain.assistant.ChunkingStrategy.StaticChunking;
+import io.github.sashirestela.openai.domain.assistant.ToolResourceFull.FileSearch;
+import io.github.sashirestela.openai.domain.assistant.ToolResourceFull.FileSearch.VectorStore;
 import org.junit.jupiter.api.BeforeAll;
 import org.junit.jupiter.api.Test;
 
@@ -37,7 +40,18 @@ void testCreateAssistant() throws IOException {
                 .instructions("You are a very kind assistant. If you cannot find correct facts to answer the "
                         + "questions, you have to refer to the attached files or use the functions provided. "
                         + "Finally, if you receive math questions, you must write and run code to answer them.")
-                .tool(AssistantTool.FILE_SEARCH)
+                .tool(AssistantTool.fileSearch(10))
+                .toolResources(ToolResourceFull.builder()
+                        .fileSearch(FileSearch.builder()
+                                .vectorStore(VectorStore.builder()
+                                        .fileId("fileId")
+                                        .chunkingStrategy(ChunkingStrategy.staticType(StaticChunking.builder()
+                                                .maxChunkSizeTokens(100)
+                                                .chunkOverlapTokens(50)
+                                                .build()))
+                                        .build())
+                                .build())
+                        .build())
                 .metadata(Map.of("user", "tester"))
                 .temperature(0.2)
                 .responseFormat("auto")

src/test/resources/assistants_create.json

@@ -1 +1 @@
-{"id":"asst_aCEgMriCKzNDqvkhl1TmrJaK","object":"assistant","created_at":1714499627,"name":"Demo Assistant","description":"This is an assistant for demonstration purposes.","model":"gpt-4-turbo","instructions":"You are a very kind assistant. If you cannot find correct facts to answer the questions, you have to refer to the attached files or use the functions provided. Finally, if you receive math questions, you must write and run code to answer them.","tools":[{"type":"file_search"}],"top_p":1.0,"temperature":0.2,"tool_resources":{"file_search":{"vector_store_ids":[]}},"metadata":{"user":"tester"},"response_format":"auto"}
+{"id":"asst_pTVcc7eVjgnf347eVIoEJHqG","object":"assistant","created_at":1717562128,"name":"Demo Assistant","description":"This is an assistant for demonstration purposes.","model":"gpt-4-turbo","instructions":"You are a very kind assistant. If you cannot find correct facts to answer the questions, you have to refer to the attached files or use the functions provided. Finally, if you receive math questions, you must write and run code to answer them.","tools":[{"type":"file_search","file_search":{"max_num_results":10}}],"top_p":1.0,"temperature":0.2,"tool_resources":{"file_search":{"vector_store_ids":["vs_LdYNXS4dMnllTRG5t5Kp1OSg"]}},"metadata":{"user":"tester"},"response_format":"auto"}