release: 0.32.0 (#278)

* feat(client): accept `InputStream` and `Path` for file params (#277)

* feat(client): accept `InputStream` and `Path` for file params

* docs: document file uploads in readme (#279)

* release: 0.32.0

---------

Co-authored-by: stainless-app[bot] <142633134+stainless-app[bot]@users.noreply.github.com>
Co-authored-by: Stainless Bot <[email protected]>

Author: stainless-app[bot]

.release-please-manifest.json

@@ -1,3 +1,3 @@
 {
-  ".": "0.31.1"
+  ".": "0.32.0"
 }

CHANGELOG.md

@@ -1,5 +1,19 @@
 # Changelog
 
+## 0.32.0 (2025-03-06)
+
+Full Changelog: [v0.31.1...v0.32.0](https://github.com/openai/openai-java/compare/v0.31.1...v0.32.0)
+
+### Features
+
+* **client:** accept `InputStream` and `Path` for file params ([3fda365](https://github.com/openai/openai-java/commit/3fda365a9d1b228ec4799a57137a288c25441dc7))
+* **client:** accept `InputStream` and `Path` for file params ([#277](https://github.com/openai/openai-java/issues/277)) ([04f2d3c](https://github.com/openai/openai-java/commit/04f2d3c1042187fe2b52ec132ff3a0529f5b1316))
+
+
+### Documentation
+
+* document file uploads in readme ([#279](https://github.com/openai/openai-java/issues/279)) ([96e4969](https://github.com/openai/openai-java/commit/96e4969a322da35f6ac13324f69d9170c7280aff))
+
 ## 0.31.1 (2025-03-05)
 
 Full Changelog: [v0.31.0...v0.31.1](https://github.com/openai/openai-java/compare/v0.31.0...v0.31.1)

README.md

@@ -9,8 +9,8 @@
 
 <!-- x-release-please-start-version -->
 
-[![Maven Central](https://img.shields.io/maven-central/v/com.openai/openai-java)](https://central.sonatype.com/artifact/com.openai/openai-java/0.31.1)
-[![javadoc](https://javadoc.io/badge2/com.openai/openai-java/0.31.1/javadoc.svg)](https://javadoc.io/doc/com.openai/openai-java/0.31.1)
+[![Maven Central](https://img.shields.io/maven-central/v/com.openai/openai-java)](https://central.sonatype.com/artifact/com.openai/openai-java/0.32.0)
+[![javadoc](https://javadoc.io/badge2/com.openai/openai-java/0.32.0/javadoc.svg)](https://javadoc.io/doc/com.openai/openai-java/0.32.0)
 
 <!-- x-release-please-end -->
 
@@ -25,7 +25,7 @@ The REST API documentation can be found on [platform.openai.com](https://platfor
 ### Gradle
 
 ```kotlin
-implementation("com.openai:openai-java:0.31.1")
+implementation("com.openai:openai-java:0.32.0")
 ```
 
 ### Maven
@@ -34,7 +34,7 @@ implementation("com.openai:openai-java:0.31.1")
 <dependency>
     <groupId>com.openai</groupId>
     <artifactId>openai-java</artifactId>
-    <version>0.31.1</version>
+    <version>0.32.0</version>
 </dependency>
 ```
 
@@ -263,6 +263,74 @@ OpenAIClient client = OpenAIOkHttpClient.builder()
     .build();
 ```
 
+## File uploads
+
+The SDK defines methods that accept files.
+
+To upload a file, pass a [`Path`](https://docs.oracle.com/javase/8/docs/api/java/nio/file/Path.html):
+
+```java
+import com.openai.models.FileCreateParams;
+import com.openai.models.FileObject;
+import com.openai.models.FilePurpose;
+import java.nio.file.Paths;
+
+FileCreateParams params = FileCreateParams.builder()
+    .purpose(FilePurpose.FINE_TUNE)
+    .file(Paths.get("input.jsonl"))
+    .build();
+FileObject fileObject = client.files().create(params);
+```
+
+Or an arbitrary [`InputStream`](https://docs.oracle.com/javase/8/docs/api/java/io/InputStream.html):
+
+```java
+import com.openai.models.FileCreateParams;
+import com.openai.models.FileObject;
+import com.openai.models.FilePurpose;
+import java.net.URL;
+
+FileCreateParams params = FileCreateParams.builder()
+    .purpose(FilePurpose.FINE_TUNE)
+    .file(new URL("https://example.com").openStream())
+    .build();
+FileObject fileObject = client.files().create(params);
+```
+
+Or a `byte[]` array:
+
+```java
+import com.openai.models.FileCreateParams;
+import com.openai.models.FileObject;
+import com.openai.models.FilePurpose;
+
+FileCreateParams params = FileCreateParams.builder()
+    .purpose(FilePurpose.FINE_TUNE)
+    .file("content".getBytes())
+    .build();
+FileObject fileObject = client.files().create(params);
+```
+
+Note that when passing a non-`Path` its filename is unknown so it will not be included in the request. To manually set a filename, pass a `MultipartField`:
+
+```java
+import com.openai.core.MultipartField;
+import com.openai.models.FileCreateParams;
+import com.openai.models.FileObject;
+import com.openai.models.FilePurpose;
+import java.io.InputStream;
+import java.net.URL;
+
+FileCreateParams params = FileCreateParams.builder()
+    .purpose(FilePurpose.FINE_TUNE)
+    .file(MultipartField.<InputStream>builder()
+        .value(new URL("https://example.com").openStream())
+        .filename("input.jsonl")
+        .build())
+    .build();
+FileObject fileObject = client.files().create(params);
+```
+
 ## Binary responses
 
 The SDK defines methods that return binary responses, which are used for API responses that shouldn't necessarily be parsed, like non-JSON data.

build.gradle.kts

@@ -8,7 +8,7 @@ repositories {
 
 allprojects {
     group = "com.openai"
-    version = "0.31.1" // x-release-please-version
+    version = "0.32.0" // x-release-please-version
 }
 
 subprojects {

openai-java-core/src/main/kotlin/com/openai/core/ObjectMappers.kt

@@ -3,23 +3,43 @@
 package com.openai.core
 
 import com.fasterxml.jackson.annotation.JsonInclude
+import com.fasterxml.jackson.core.JsonGenerator
 import com.fasterxml.jackson.databind.DeserializationFeature
 import com.fasterxml.jackson.databind.SerializationFeature
+import com.fasterxml.jackson.databind.SerializerProvider
 import com.fasterxml.jackson.databind.cfg.CoercionAction.Fail
 import com.fasterxml.jackson.databind.cfg.CoercionInputShape.Integer
 import com.fasterxml.jackson.databind.json.JsonMapper
+import com.fasterxml.jackson.databind.module.SimpleModule
 import com.fasterxml.jackson.datatype.jdk8.Jdk8Module
 import com.fasterxml.jackson.datatype.jsr310.JavaTimeModule
 import com.fasterxml.jackson.module.kotlin.jacksonMapperBuilder
+import java.io.InputStream
 
 fun jsonMapper(): JsonMapper =
     jacksonMapperBuilder()
         .addModule(Jdk8Module())
         .addModule(JavaTimeModule())
+        .addModule(SimpleModule().addSerializer(InputStreamJsonSerializer))
         .serializationInclusion(JsonInclude.Include.NON_ABSENT)
         .disable(DeserializationFeature.ADJUST_DATES_TO_CONTEXT_TIME_ZONE)
         .disable(SerializationFeature.FLUSH_AFTER_WRITE_VALUE)
         .disable(SerializationFeature.WRITE_DATES_AS_TIMESTAMPS)
         .disable(SerializationFeature.WRITE_DURATIONS_AS_TIMESTAMPS)
         .withCoercionConfig(String::class.java) { it.setCoercion(Integer, Fail) }
         .build()
+
+private object InputStreamJsonSerializer : BaseSerializer<InputStream>(InputStream::class) {
+
+    override fun serialize(
+        value: InputStream?,
+        gen: JsonGenerator?,
+        serializers: SerializerProvider?,
+    ) {
+        if (value == null) {
+            gen?.writeNull()
+        } else {
+            value.use { gen?.writeBinary(it.readBytes()) }
+        }
+    }
+}

openai-java-core/src/main/kotlin/com/openai/core/Values.kt

@@ -27,6 +27,7 @@ import com.fasterxml.jackson.databind.node.JsonNodeType.POJO
 import com.fasterxml.jackson.databind.node.JsonNodeType.STRING
 import com.fasterxml.jackson.databind.ser.std.NullSerializer
 import com.openai.errors.OpenAIInvalidDataException
+import java.io.InputStream
 import java.util.Objects
 import java.util.Optional
 
@@ -508,7 +509,10 @@ private constructor(
             return MultipartField(
                 value,
                 contentType
-                    ?: if (value is KnownValue && value.value is ByteArray)
+                    ?: if (
+                        value is KnownValue &&
+                            (value.value is InputStream || value.value is ByteArray)
+                    )
                         "application/octet-stream"
                     else "text/plain; charset=utf-8",
                 filename,

openai-java-core/src/main/kotlin/com/openai/core/http/HttpRequestBodies.kt

@@ -9,6 +9,8 @@ import com.fasterxml.jackson.databind.json.JsonMapper
 import com.fasterxml.jackson.databind.node.JsonNodeType
 import com.openai.core.MultipartField
 import com.openai.errors.OpenAIInvalidDataException
+import java.io.ByteArrayInputStream
+import java.io.InputStream
 import java.io.OutputStream
 import kotlin.jvm.optionals.getOrNull
 import org.apache.hc.client5.http.entity.mime.MultipartEntityBuilder
@@ -41,8 +43,18 @@ internal fun multipartFormData(
             MultipartEntityBuilder.create()
                 .apply {
                     fields.forEach { (name, field) ->
-                        val node = jsonMapper.valueToTree<JsonNode>(field.value)
-                        serializePart(name, node).forEach { (name, bytes) ->
+                        val knownValue = field.value.asKnown().getOrNull()
+                        val parts =
+                            if (knownValue is InputStream) {
+                                // Read directly from the `InputStream` instead of reading it all
+                                // into memory due to the `jsonMapper` serialization below.
+                                sequenceOf(name to knownValue)
+                            } else {
+                                val node = jsonMapper.valueToTree<JsonNode>(field.value)
+                                serializePart(name, node)
+                            }
+
+                        parts.forEach { (name, bytes) ->
                             addBinaryBody(
                                 name,
                                 bytes,
@@ -55,16 +67,19 @@ internal fun multipartFormData(
                 .build()
         }
 
-        private fun serializePart(name: String, node: JsonNode): Sequence<Pair<String, ByteArray>> =
+        private fun serializePart(
+            name: String,
+            node: JsonNode,
+        ): Sequence<Pair<String, InputStream>> =
             when (node.nodeType) {
                 JsonNodeType.MISSING,
                 JsonNodeType.NULL -> emptySequence()
-                JsonNodeType.BINARY -> sequenceOf(name to node.binaryValue())
-                JsonNodeType.STRING -> sequenceOf(name to node.textValue().toByteArray())
+                JsonNodeType.BINARY -> sequenceOf(name to ByteArrayInputStream(node.binaryValue()))
+                JsonNodeType.STRING -> sequenceOf(name to node.textValue().toInputStream())
                 JsonNodeType.BOOLEAN ->
-                    sequenceOf(name to node.booleanValue().toString().toByteArray())
+                    sequenceOf(name to node.booleanValue().toString().toInputStream())
                 JsonNodeType.NUMBER ->
-                    sequenceOf(name to node.numberValue().toString().toByteArray())
+                    sequenceOf(name to node.numberValue().toString().toInputStream())
                 JsonNodeType.ARRAY ->
                     node.elements().asSequence().flatMap { element ->
                         serializePart("$name[]", element)
@@ -78,6 +93,8 @@ internal fun multipartFormData(
                     throw OpenAIInvalidDataException("Unexpected JsonNode type: ${node.nodeType}")
             }
 
+        private fun String.toInputStream(): InputStream = ByteArrayInputStream(toByteArray())
+
         override fun writeTo(outputStream: OutputStream) = entity.writeTo(outputStream)
 
         override fun contentType(): String = entity.contentType

openai-java-core/src/main/kotlin/com/openai/models/AudioTranscriptionCreateParams.kt

@@ -14,8 +14,13 @@ import com.openai.core.http.Headers
 import com.openai.core.http.QueryParams
 import com.openai.core.toImmutable
 import com.openai.errors.OpenAIInvalidDataException
+import java.io.ByteArrayInputStream
+import java.io.InputStream
+import java.nio.file.Path
 import java.util.Objects
 import java.util.Optional
+import kotlin.io.path.inputStream
+import kotlin.io.path.name
 
 /** Transcribes audio into the input language. */
 class AudioTranscriptionCreateParams
@@ -29,7 +34,7 @@ private constructor(
      * The audio file object (not file name) to transcribe, in one of these formats: flac, mp3, mp4,
      * mpeg, mpga, m4a, ogg, wav, or webm.
      */
-    fun file(): ByteArray = body.file()
+    fun file(): InputStream = body.file()
 
     /**
      * ID of the model to use. Only `whisper-1` (which is powered by our open source Whisper V2
@@ -78,7 +83,7 @@ private constructor(
      * The audio file object (not file name) to transcribe, in one of these formats: flac, mp3, mp4,
      * mpeg, mpga, m4a, ogg, wav, or webm.
      */
-    fun _file(): MultipartField<ByteArray> = body._file()
+    fun _file(): MultipartField<InputStream> = body._file()
 
     /**
      * ID of the model to use. Only `whisper-1` (which is powered by our open source Whisper V2
@@ -148,7 +153,7 @@ private constructor(
     class Body
     @JsonCreator
     private constructor(
-        private val file: MultipartField<ByteArray>,
+        private val file: MultipartField<InputStream>,
         private val model: MultipartField<AudioModel>,
         private val language: MultipartField<String>,
         private val prompt: MultipartField<String>,
@@ -161,7 +166,7 @@ private constructor(
          * The audio file object (not file name) to transcribe, in one of these formats: flac, mp3,
          * mp4, mpeg, mpga, m4a, ogg, wav, or webm.
          */
-        fun file(): ByteArray = file.value.getRequired("file")
+        fun file(): InputStream = file.value.getRequired("file")
 
         /**
          * ID of the model to use. Only `whisper-1` (which is powered by our open source Whisper V2
@@ -214,7 +219,7 @@ private constructor(
          * The audio file object (not file name) to transcribe, in one of these formats: flac, mp3,
          * mp4, mpeg, mpga, m4a, ogg, wav, or webm.
          */
-        fun _file(): MultipartField<ByteArray> = file
+        fun _file(): MultipartField<InputStream> = file
 
         /**
          * ID of the model to use. Only `whisper-1` (which is powered by our open source Whisper V2
@@ -296,7 +301,7 @@ private constructor(
         /** A builder for [Body]. */
         class Builder internal constructor() {
 
-            private var file: MultipartField<ByteArray>? = null
+            private var file: MultipartField<InputStream>? = null
             private var model: MultipartField<AudioModel>? = null
             private var language: MultipartField<String> = MultipartField.of(null)
             private var prompt: MultipartField<String> = MultipartField.of(null)
@@ -321,13 +326,31 @@ private constructor(
              * The audio file object (not file name) to transcribe, in one of these formats: flac,
              * mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.
              */
-            fun file(file: ByteArray) = file(MultipartField.of(file))
+            fun file(file: InputStream) = file(MultipartField.of(file))
 
             /**
              * The audio file object (not file name) to transcribe, in one of these formats: flac,
              * mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.
              */
-            fun file(file: MultipartField<ByteArray>) = apply { this.file = file }
+            fun file(file: MultipartField<InputStream>) = apply { this.file = file }
+
+            /**
+             * The audio file object (not file name) to transcribe, in one of these formats: flac,
+             * mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.
+             */
+            fun file(file: ByteArray) = file(ByteArrayInputStream(file))
+
+            /**
+             * The audio file object (not file name) to transcribe, in one of these formats: flac,
+             * mp3, mp4, mpeg, mpga, m4a, ogg, wav, or webm.
+             */
+            fun file(file: Path) =
+                file(
+                    MultipartField.builder<InputStream>()
+                        .value(file.inputStream())
+                        .filename(file.name)
+                        .build()
+                )
 
             /**
              * ID of the model to use. Only `whisper-1` (which is powered by our open source Whisper
@@ -506,6 +529,18 @@ private constructor(
             additionalQueryParams = audioTranscriptionCreateParams.additionalQueryParams.toBuilder()
         }
 
+        /**
+         * The audio file object (not file name) to transcribe, in one of these formats: flac, mp3,
+         * mp4, mpeg, mpga, m4a, ogg, wav, or webm.
+         */
+        fun file(file: InputStream) = apply { body.file(file) }
+
+        /**
+         * The audio file object (not file name) to transcribe, in one of these formats: flac, mp3,
+         * mp4, mpeg, mpga, m4a, ogg, wav, or webm.
+         */
+        fun file(file: MultipartField<InputStream>) = apply { body.file(file) }
+
         /**
          * The audio file object (not file name) to transcribe, in one of these formats: flac, mp3,
          * mp4, mpeg, mpga, m4a, ogg, wav, or webm.
@@ -516,7 +551,7 @@ private constructor(
          * The audio file object (not file name) to transcribe, in one of these formats: flac, mp3,
          * mp4, mpeg, mpga, m4a, ogg, wav, or webm.
          */
-        fun file(file: MultipartField<ByteArray>) = apply { body.file(file) }
+        fun file(file: Path) = apply { body.file(file) }
 
         /**
          * ID of the model to use. Only `whisper-1` (which is powered by our open source Whisper V2