Skip to content

fix: allow missing task and duration in diarized transcriptions #803

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wskr00 wants to merge 4 commits into
base: main
Choose a base branch
from
Open

fix: allow missing task and duration in diarized transcriptions #803

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
d7cd8d9
fix: turn diarized duration into optional
wskr00 Jul 19, 2026
b170336
fix: allow missing task in diarized transcriptions
wskr00 Jul 19, 2026
7c8d0d2
fix: validate optional diarized duration
wskr00 Jul 19, 2026
83a861f
fix: preserve transcription response variants with empty segments
wskr00 Jul 19, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
27 changes: 18 additions & 9 deletions ...java-core/src/main/kotlin/com/openai/models/audio/transcriptions/TranscriptionDiarized.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -58,10 +58,10 @@ private constructor(
/**
* Duration of the input audio in seconds.
*
* @throws OpenAIInvalidDataException if the JSON field has an unexpected type or is
* unexpectedly missing or null (e.g. if the server responded with an unexpected value).
* @throws OpenAIInvalidDataException if the JSON field has an unexpected type (e.g. if the
* server responded with an unexpected value).
*/
fun duration(): Double = duration.getRequired("duration")
fun duration(): Optional<Double> = duration.getOptional("duration")

@chatgpt-codex-connector chatgpt-codex-connector Bot Jul 19, 2026

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

P2 Badge Preserve the duration getter's JVM descriptor

When an application or library compiled against the previous SDK invokes this getter after the SDK jar is upgraded without recompiling that caller, its bytecode still resolves duration()D, while this class now only provides duration()Ljava/util/Optional;; the invocation therefore fails at runtime with NoSuchMethodError, even when the response contains a duration. The primitive builder overload preserves the setter ABI, but the public getter needs equivalent compatibility handling (for example, retaining it and exposing the optional value under a new method name) or this change must be treated as a breaking release.

Useful? React with 👍 / 👎.


/**
* Segments of the transcript annotated with timestamps and speaker labels.
Expand Down Expand Up @@ -149,7 +149,6 @@ private constructor(
*
* The following fields are required:
* ```java
* .duration()
* .segments()
* .text()
* ```
Expand All @@ -160,7 +159,7 @@ private constructor(
/** A builder for [TranscriptionDiarized]. */
class Builder internal constructor() {

private var duration: JsonField<Double>? = null
private var duration: JsonField<Double> = JsonMissing.of()
private var segments: JsonField<MutableList<TranscriptionDiarizedSegment>>? = null
private var task: JsonValue = JsonValue.from("transcribe")
private var text: JsonField<String>? = null
Expand All @@ -178,7 +177,17 @@ private constructor(
}

/** Duration of the input audio in seconds. */
fun duration(duration: Double) = duration(JsonField.of(duration))
fun duration(duration: Double?) = duration(JsonField.ofNullable(duration))

/**
* Alias for [Builder.duration].
*
* This unboxed primitive overload exists for backwards compatibility.
*/
fun duration(duration: Double) = duration(duration as Double?)

/** Alias for calling [Builder.duration] with `duration.orElse(null)`. */
fun duration(duration: Optional<Double>) = duration(duration.getOrNull())

/**
* Sets [Builder.duration] to an arbitrary JSON value.
Expand Down Expand Up @@ -294,7 +303,6 @@ private constructor(
*
* The following fields are required:
* ```java
* .duration()
* .segments()
* .text()
* ```
Expand All @@ -303,7 +311,7 @@ private constructor(
*/
fun build(): TranscriptionDiarized =
TranscriptionDiarized(
checkRequired("duration", duration),
duration,
checkRequired("segments", segments).map { it.toImmutable() },
task,
checkRequired("text", text),
Expand All @@ -330,7 +338,7 @@ private constructor(
duration()
segments().forEach { it.validate() }
Comment thread
wskr00 marked this conversation as resolved.
_task().let {
if (it != JsonValue.from("transcribe")) {
if (it != JsonMissing.of() && it != JsonValue.from("transcribe")) {
Comment thread
wskr00 marked this conversation as resolved.
throw OpenAIInvalidDataException("'task' is invalid, received $it")
}
}
Expand All @@ -355,6 +363,7 @@ private constructor(
@JvmSynthetic
internal fun validity(): Int =
(if (duration.asKnown().isPresent) 1 else 0) +
(if (segments.asKnown().isPresent) 1 else 0) +
Comment thread
wskr00 marked this conversation as resolved.
(segments.asKnown().getOrNull()?.sumOf { it.validity().toInt() } ?: 0) +
task.let { if (it == JsonValue.from("transcribe")) 1 else 0 } +
(if (text.asKnown().isPresent) 1 else 0) +
Expand Down
1 change: 1 addition & 0 deletions ...-java-core/src/main/kotlin/com/openai/models/audio/transcriptions/TranscriptionVerbose.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -370,6 +370,7 @@ private constructor(
(if (duration.asKnown().isPresent) 1 else 0) +
(if (language.asKnown().isPresent) 1 else 0) +
(if (text.asKnown().isPresent) 1 else 0) +
(if (segments.asKnown().isPresent) 1 else 0) +
(segments.asKnown().getOrNull()?.sumOf { it.validity().toInt() } ?: 0) +
(usage.asKnown().getOrNull()?.validity() ?: 0) +
(words.asKnown().getOrNull()?.sumOf { it.validity().toInt() } ?: 0)
Expand Down
38 changes: 38 additions & 0 deletions ...src/test/kotlin/com/openai/models/audio/transcriptions/TranscriptionCreateResponseTest.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -162,6 +162,44 @@ internal class TranscriptionCreateResponseTest {
assertThat(roundtrippedTranscriptionCreateResponse).isEqualTo(transcriptionCreateResponse)
}

@Test
fun deserializeDiarizedWithEmptySegments() {
val transcriptionCreateResponse =
jsonMapper()
.readValue(
"""
{
"text": "",
"segments": []
}
"""
.trimIndent(),
jacksonTypeRef<TranscriptionCreateResponse>(),
)

assertThat(transcriptionCreateResponse.isDiarized()).isTrue()
}

@Test
fun deserializeVerboseWithEmptySegments() {
val transcriptionCreateResponse =
jsonMapper()
.readValue(
"""
{
"duration": 0.0,
"language": "en",
"text": "",
"segments": []
}
"""
.trimIndent(),
jacksonTypeRef<TranscriptionCreateResponse>(),
)

assertThat(transcriptionCreateResponse.isVerbose()).isTrue()
}

@Test
fun ofVerbose() {
val verbose =
Expand Down
81 changes: 79 additions & 2 deletions ...-core/src/test/kotlin/com/openai/models/audio/transcriptions/TranscriptionDiarizedTest.kt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3,6 +3,7 @@
package com.openai.models.audio.transcriptions

import com.fasterxml.jackson.module.kotlin.jacksonTypeRef
import com.openai.core.JsonValue
import com.openai.core.jsonMapper
import org.assertj.core.api.Assertions.assertThat
import org.junit.jupiter.api.Test
Expand All @@ -13,7 +14,6 @@ internal class TranscriptionDiarizedTest {
fun create() {
val transcriptionDiarized =
TranscriptionDiarized.builder()
.duration(0.0)
.addSegment(
TranscriptionDiarizedSegment.builder()
.id("id")
Expand All @@ -39,7 +39,7 @@ internal class TranscriptionDiarizedTest {
)
.build()

assertThat(transcriptionDiarized.duration()).isEqualTo(0.0)
assertThat(transcriptionDiarized.duration()).isEmpty
assertThat(transcriptionDiarized.segments())
.containsExactly(
TranscriptionDiarizedSegment.builder()
Expand Down Expand Up @@ -69,6 +69,82 @@ internal class TranscriptionDiarizedTest {
)
}

@Test
fun deserializeWithoutOptionalFields() {
val transcriptionDiarized =
jsonMapper()
.readValue(
"""
{
"text": "text",
"segments": [
{
"type": "transcript.text.segment",
"id": "id",
"end": 1.0,
"speaker": "speaker",
"start": 0.0,
"text": "text"
}
]
}
"""
.trimIndent(),
jacksonTypeRef<TranscriptionDiarized>(),
)

assertThat(transcriptionDiarized.duration()).isEmpty
assertThat(transcriptionDiarized.isValid()).isTrue()
}

@Test
fun rejectsInvalidDuration() {
val transcriptionDiarized =
jsonMapper()
.readValue(
"""
{
"duration": "invalid",
"text": "text",
"segments": [
{
"type": "transcript.text.segment",
"id": "id",
"end": 1.0,
"speaker": "speaker",
"start": 0.0,
"text": "text"
}
]
}
"""
.trimIndent(),
jacksonTypeRef<TranscriptionDiarized>(),
)

assertThat(transcriptionDiarized.isValid()).isFalse()
}

@Test
fun rejectsUnexpectedTask() {
val transcriptionDiarized =
TranscriptionDiarized.builder()
.addSegment(
TranscriptionDiarizedSegment.builder()
.id("id")
.end(1.0)
.speaker("speaker")
.start(0.0)
.text("text")
.build()
)
.task(JsonValue.from("translate"))
.text("text")
.build()

assertThat(transcriptionDiarized.isValid()).isFalse()
}

@Test
fun roundtrip() {
val jsonMapper = jsonMapper()
Expand Down Expand Up @@ -107,5 +183,6 @@ internal class TranscriptionDiarizedTest {
)

assertThat(roundtrippedTranscriptionDiarized).isEqualTo(transcriptionDiarized)
assertThat(roundtrippedTranscriptionDiarized.duration()).contains(0.0)
}
}