SDK Update and Refactor

CHANGELOG.md

@@ -4,6 +4,90 @@ All notable changes to this project will be documented in this file.
 The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/),
 and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
+## [0.6.0] - 2024-07-23
+
+### Migrating from 0.4.0 -> 0.6.0
+
+Module Imports
+
+```
+use deepgram::{
+---    transcription::prerecorded::{
++++    common::{
+        audio_source::AudioSource,
+        options::{Language, Options},
+    },
+    Deepgram, DeepgramError,
+};
+```
+
+Streaming Changes
+
+Now you can pass Options using stream_request_with_options
+```
+let options = Options::builder()
+    .smart_format(true)
+    .language(Language::en_US)
+    .build();
+
+let mut results = dg
+    .transcription()
+    .stream_request_with_options(Some(&options))
+    .file(PATH_TO_FILE, AUDIO_CHUNK_SIZE, Duration::from_millis(16))
+    .await?
+    .start()
+    .await?;
+```
+
+### Changed
+
+- Add streaming features
+- Add support for pre-recorded features when streaming
+- Add Speech to Text
+- Refactor Code
+
+Note that this modifies some public enum's which are used internally but might need to be taken into account
+- Bumped the minor version but it might need a major bump?
+- Maybe a 1.0.0 release?
+
+### Streaming Features
+- endpointing
+- utterance_end_ms
+- interim_results
+- no_delay
+- vad_events
+
+### Streaming Functions
+- keep_alive
+
+### New Streaming Message Types
+- Utterance End
+- Speech Started
+
+### Pre-Recorded Features
+- encoding
+- smart_format
+- callback
+- callback_method
+- filler_words
+- paragraphs
+- diarize_version
+- dictation
+- measurements
+- extra
+
+### Pre-Recorded Audio Intelligence Features
+- detect_entities
+- sentiment
+- topics
+- summarize
+- intents
+- custom_intents
+- custom_intent_mode
+- topics
+- custom_topics
+- custom_topic_mode
+
 ## [Unreleased]
 - Deprecate tiers and add explicit support for all currently available models.
 - Expand language enum to include all currently-supported languages.

Cargo.toml

@@ -47,40 +47,40 @@ speak = ["dep:rodio", "dep:pkg-config"]
 
 [[example]]
 name = "prerecorded_from_file"
-path = "examples/transcription/prerecorded/prerecorded_from_file.rs"
+path = "examples/transcription/rest/prerecorded_from_file.rs"
 required-features = ["listen"]
 
 [[example]]
 name = "callback"
-path = "examples/transcription/prerecorded/callback.rs"
+path = "examples/transcription/rest/callback.rs"
 required-features = ["listen"]
 
 [[example]]
 name = "make_prerecorded_request_builder"
-path = "examples/transcription/prerecorded/make_prerecorded_request_builder.rs"
+path = "examples/transcription/rest/make_prerecorded_request_builder.rs"
 required-features = ["listen"]
 
 [[example]]
 name = "prerecorded_from_url"
-path = "examples/transcription/prerecorded/prerecorded_from_url.rs"
+path = "examples/transcription/rest/prerecorded_from_url.rs"
 required-features = ["listen"]
 
 [[example]]
 name = "simple_stream"
-path = "examples/transcription/live/simple_stream.rs"
+path = "examples/transcription/websocket/simple_stream.rs"
 required-features = ["listen"]
 
 [[example]]
 name = "microphone_stream"
-path = "examples/transcription/live/microphone_stream.rs"
+path = "examples/transcription/websocket/microphone_stream.rs"
 required-features = ["listen"]
 
 [[example]]
 name = "text_to_speech_to_file"
-path = "examples/speak/text_to_speech_to_file.rs"
+path = "examples/speak/rest/text_to_speech_to_file.rs"
 required-features = ["speak"]
 
 [[example]]
 name = "text_to_speech_to_stream"
-path = "examples/speak/text_to_speech_to_stream.rs"
+path = "examples/speak/rest/text_to_speech_to_stream.rs"
 required-features = ["speak"]

examples/transcription/live/simple_stream.rs → examples/transcription/websocket/simple_stream.rs

@@ -27,8 +27,8 @@ async fn main() -> Result<(), DeepgramError> {
         .encoding(Encoding::Linear16)
         .sample_rate(44100)
         .channels(2)
-        // .endpointing(Endpointing::Enabled())
-        .endpointing(Endpointing::CustomValue(300))
+        // .endpointing(Endpointing::Enabled)
+        .endpointing(Endpointing::CustomDurationMs(300))
         .interim_results(true)
         .utterance_end_ms(1000)
         .vad_events(true)

src/common/audio_source.rs

@@ -7,7 +7,7 @@
 use reqwest::{header::CONTENT_TYPE, RequestBuilder};
 use serde::Serialize;
 
-/// Used as a parameter for [`Transcription::prerecorded`](crate::listen::speech_to_text::Transcription::prerecorded) and similar functions.
+/// Used as a parameter for [`Transcription::prerecorded`](crate::Transcription::prerecorded) and similar functions.
 #[derive(Debug)]
 pub struct AudioSource(InternalAudioSource);
 
@@ -55,7 +55,7 @@ impl AudioSource {
         })
     }
 
-    /// Fill body
+    #[allow(missing_docs)]
     pub fn fill_body(self, request_builder: RequestBuilder) -> RequestBuilder {
         match self.0 {
             InternalAudioSource::Url(url) => {

src/common/batch_response.rs

@@ -7,7 +7,7 @@
 use serde::{Deserialize, Serialize};
 use uuid::Uuid;
 
-/// Returned by [`Transcription::prerecorded`](crate::listen::speech_to_text::Transcription::prerecorded).
+/// Returned by [`Transcription::prerecorded`](crate::Transcription::prerecorded).
 ///
 /// See the [Deepgram API Reference][api] for more info.
 ///
@@ -22,7 +22,7 @@ pub struct Response {
     pub results: ListenResults,
 }
 
-/// Returned by [`Transcription::prerecorded_callback`](crate::listen::speech_to_text::Transcription::prerecorded_callback).
+/// Returned by [`Transcription::prerecorded_callback`](crate::Transcription::prerecorded_callback).
 ///
 /// See the [Deepgram Callback feature docs][docs] for more info.
 ///

src/common/options.rs

@@ -8,7 +8,7 @@ use std::collections::HashMap;
 
 use serde::{ser::SerializeSeq, Deserialize, Serialize};
 
-/// Used as a parameter for [`Transcription::prerecorded`](crate::listen::speech_to_text::Transcription::prerecorded) and similar functions.
+/// Used as a parameter for [`Transcription::prerecorded`](crate::Transcription::prerecorded) and similar functions.
 #[derive(Debug, PartialEq, Clone)]
 pub struct Options {
     model: Option<Model>,
@@ -165,7 +165,7 @@ pub enum Endpointing {
     Disabled,
 
     #[allow(missing_docs)]
-    CustomValue(u32),
+    CustomDurationMs(u32),
 }
 
 /// Endpointing impl
@@ -175,7 +175,7 @@ impl Endpointing {
         match self {
             Endpointing::Enabled => "true".to_string(),
             Endpointing::Disabled => "false".to_string(),
-            Endpointing::CustomValue(value) => value.to_string(),
+            Endpointing::CustomDurationMs(value) => value.to_string(),
         }
     }
 }
@@ -603,13 +603,16 @@ pub struct Keyword {
 #[non_exhaustive]
 pub enum Utterances {
     #[allow(missing_docs)]
-    Enabled {
-        #[allow(missing_docs)]
-        utt_split: Option<f64>,
-    },
+    Enabled,
 
     #[allow(missing_docs)]
     Disabled,
+
+    #[allow(missing_docs)]
+    CustomSplit {
+        #[allow(missing_docs)]
+        utt_split: Option<f64>,
+    },
 }
 
 /// Used as a parameter for [`OptionsBuilder::multichannel`].
@@ -621,21 +624,22 @@ pub enum Utterances {
 #[non_exhaustive]
 pub enum Multichannel {
     #[allow(missing_docs)]
-    Enabled {
-        #[allow(missing_docs)]
-        models: Option<Vec<Model>>,
-    },
+    Enabled,
 
     #[allow(missing_docs)]
     Disabled,
+
+    #[allow(missing_docs)]
+    ModelPerChannel {
+        #[allow(missing_docs)]
+        models: Option<Vec<Model>>,
+    },
 }
 
 /// Builds an [`Options`] object using [the Builder pattern][builder].
 ///
 /// Use it to set of Deepgram's features, excluding the Callback feature.
-/// The Callback feature can be set when making the request by calling [`Transcription::prerecorded_callback`](crate::listen::speech_to_text::Transcription::prerecorded_callback).
-///
-/// [builder]: https://rust-unofficial.github.io/patterns/patterns/creational/builder.html
+/// The Callback feature can be set when making the request by calling [`Transcription::prerecorded_callback`](crate::Transcription::prerecorded_callback).
 #[derive(Debug, PartialEq, Clone)]
 pub struct OptionsBuilder(Options);
 
@@ -717,25 +721,10 @@ impl OptionsBuilder {
     ///     .build();
     /// ```
     ///
-    /// ```
-    /// # use deepgram::common::options::{Model, Options};
-    /// #
-    /// let options1 = Options::builder()
-    ///     .multichannel_with_models([Model::Nova2Meeting, Model::Nova2Phonecall])
-    ///     .model(Model::Nova2)
-    ///     .build();
-    ///
-    /// let options2 = Options::builder()
-    ///     .multichannel(true)
-    ///     .model(Model::Nova2)
-    ///     .build();
-    ///
-    /// assert_eq!(options1, options2);
-    /// ```
     pub fn model(mut self, model: Model) -> Self {
         self.0.model = Some(model);
 
-        if let Some(Multichannel::Enabled { models }) = &mut self.0.multichannel {
+        if let Some(Multichannel::ModelPerChannel { models }) = &mut self.0.multichannel {
             *models = None;
         }
 
@@ -963,7 +952,7 @@ impl OptionsBuilder {
     /// ```
     pub fn multichannel(mut self, multichannel: bool) -> Self {
         self.0.multichannel = Some(if multichannel {
-            Multichannel::Enabled { models: None }
+            Multichannel::Enabled
         } else {
             Multichannel::Disabled
         });
@@ -1073,7 +1062,7 @@ impl OptionsBuilder {
     /// assert_eq!(options1, options2);
     /// ```
     pub fn multichannel_with_models(mut self, models: impl IntoIterator<Item = Model>) -> Self {
-        if let Some(Multichannel::Enabled {
+        if let Some(Multichannel::ModelPerChannel {
             models: Some(old_models),
         }) = &mut self.0.multichannel
         {
@@ -1082,7 +1071,7 @@ impl OptionsBuilder {
             old_models.extend(models);
         } else {
             // Multichannel with models already enabled
-            self.0.multichannel = Some(Multichannel::Enabled {
+            self.0.multichannel = Some(Multichannel::ModelPerChannel {
                 models: Some(models.into_iter().collect()),
             });
         }
@@ -1385,7 +1374,7 @@ impl OptionsBuilder {
     /// ```
     pub fn utterances(mut self, utterances: bool) -> Self {
         self.0.utterances = Some(if utterances {
-            Utterances::Enabled { utt_split: None }
+            Utterances::Enabled
         } else {
             Utterances::Disabled
         });
@@ -1413,7 +1402,7 @@ impl OptionsBuilder {
     ///     .build();
     /// ```
     pub fn utterances_with_utt_split(mut self, utt_split: f64) -> Self {
-        self.0.utterances = Some(Utterances::Enabled {
+        self.0.utterances = Some(Utterances::CustomSplit {
             utt_split: Some(utt_split),
         });
         self
@@ -1961,15 +1950,20 @@ impl Serialize for SerializableOptions<'_> {
         match multichannel {
             // Multichannels with models is enabled
             // Ignore self.model field
-            Some(Multichannel::Enabled {
+            Some(Multichannel::ModelPerChannel {
                 models: Some(models),
             }) => {
                 seq.serialize_element(&("model", models_to_string(models)))?;
             }
 
             // Multichannel with models is not enabled
             // Use self.model field
-            Some(Multichannel::Enabled { models: None } | Multichannel::Disabled) | None => {
+            Some(
+                Multichannel::ModelPerChannel { models: None }
+                | Multichannel::Enabled
+                | Multichannel::Disabled,
+            )
+            | None => {
                 if let Some(model) = model {
                     seq.serialize_element(&("model", model.as_ref()))?;
                 }
@@ -2016,7 +2010,8 @@ impl Serialize for SerializableOptions<'_> {
 
         match multichannel {
             Some(Multichannel::Disabled) => seq.serialize_element(&("multichannel", false))?,
-            Some(Multichannel::Enabled { models: _ }) => {
+            Some(Multichannel::Enabled) => seq.serialize_element(&("multichannel", true))?,
+            Some(Multichannel::ModelPerChannel { models: _ }) => {
                 // Multichannel models are serialized above if they exist
                 // This is done instead of serializing the self.model field
                 seq.serialize_element(&("multichannel", true))?;
@@ -2061,7 +2056,8 @@ impl Serialize for SerializableOptions<'_> {
 
         match utterances {
             Some(Utterances::Disabled) => seq.serialize_element(&("utterances", false))?,
-            Some(Utterances::Enabled { utt_split }) => {
+            Some(Utterances::Enabled) => seq.serialize_element(&("utterances", true))?,
+            Some(Utterances::CustomSplit { utt_split }) => {
                 seq.serialize_element(&("utterances", true))?;
 
                 if let Some(utt_split) = utt_split {