@wovin/tranz 0.0.26

package/dist/utils/audio/merge-results.d.ts

@@ -0,0 +1,47 @@
+/**
+ * Utilities for merging transcription results from split audio segments
+ */
+import type { TranscriptionResult } from '../transcription/providers.js';
+import type { AudioSegment } from './split.js';
+/**
+ * Word-level data with timing information
+ */
+export interface WordData {
+    word: string;
+    start: number;
+    end: number;
+    confidence?: number;
+    speaker?: string;
+}
+/**
+ * Merged transcription result with segment information
+ */
+export interface MergedTranscriptionResult extends TranscriptionResult {
+    /** Segment metadata for reference */
+    segments?: {
+        index: number;
+        startSec: number;
+        endSec: number;
+        text: string;
+    }[];
+    /** Total segments that were merged */
+    totalSegments?: number;
+}
+/**
+ * Merge multiple transcription results from audio segments into one
+ * Adjusts word timestamps to be relative to the original audio
+ *
+ * @param results - Array of transcription results from each segment
+ * @param segments - Array of audio segment metadata
+ * @returns Merged transcription result
+ */
+export declare function mergeTranscriptionResults(results: TranscriptionResult[], segments: AudioSegment[]): MergedTranscriptionResult;
+/**
+ * Format merged results with optional segment markers in the text
+ *
+ * @param result - Merged transcription result
+ * @param includeMarkers - Whether to include [Segment N] markers
+ * @returns Formatted text
+ */
+export declare function formatMergedText(result: MergedTranscriptionResult, includeMarkers?: boolean): string;
+//# sourceMappingURL=merge-results.d.ts.map

package/dist/utils/audio/merge-results.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"merge-results.d.ts","sourceRoot":"","sources":["../../../src/utils/audio/merge-results.ts"],"names":[],"mappings":"AAAA;;GAEG;AAEH,OAAO,KAAK,EAAE,mBAAmB,EAAE,MAAM,+BAA+B,CAAA;AACxE,OAAO,KAAK,EAAE,YAAY,EAAE,MAAM,YAAY,CAAA;AAE9C;;GAEG;AACH,MAAM,WAAW,QAAQ;IACvB,IAAI,EAAE,MAAM,CAAA;IACZ,KAAK,EAAE,MAAM,CAAA;IACb,GAAG,EAAE,MAAM,CAAA;IACX,UAAU,CAAC,EAAE,MAAM,CAAA;IACnB,OAAO,CAAC,EAAE,MAAM,CAAA;CACjB;AAED;;GAEG;AACH,MAAM,WAAW,yBAA0B,SAAQ,mBAAmB;IACpE,qCAAqC;IACrC,QAAQ,CAAC,EAAE;QACT,KAAK,EAAE,MAAM,CAAA;QACb,QAAQ,EAAE,MAAM,CAAA;QAChB,MAAM,EAAE,MAAM,CAAA;QACd,IAAI,EAAE,MAAM,CAAA;KACb,EAAE,CAAA;IACH,sCAAsC;IACtC,aAAa,CAAC,EAAE,MAAM,CAAA;CACvB;AAED;;;;;;;GAOG;AACH,wBAAgB,yBAAyB,CACvC,OAAO,EAAE,mBAAmB,EAAE,EAC9B,QAAQ,EAAE,YAAY,EAAE,GACvB,yBAAyB,CAqF3B;AAED;;;;;;GAMG;AACH,wBAAgB,gBAAgB,CAC9B,MAAM,EAAE,yBAAyB,EACjC,cAAc,GAAE,OAAe,GAC9B,MAAM,CAeR"}

package/dist/utils/audio/split.d.ts

@@ -0,0 +1,106 @@
+/**
+ * Audio splitting utilities for tranz-cli
+ * Provides silence detection and optimal split point calculation
+ */
+/**
+ * Configuration for audio splitting
+ */
+export interface SplitConfig {
+    /** Maximum segment duration in seconds (default: 600 = 10min) */
+    maxDurationSec: number;
+    /** Minimum silence duration to consider for split (default: 1.0s) */
+    minSilenceDurSec: number;
+    /** FFmpeg silence threshold (default: '-35dB') */
+    silenceThreshold: string;
+    /** Prefer longer silences for splits (default: true) */
+    preferLongerSilence: boolean;
+    /** Buffer to leave at silence edges (default: 0.2s) */
+    silenceBuffer: number;
+}
+/**
+ * A detected silence region in the audio
+ */
+export interface SilenceRegion {
+    startSec: number;
+    endSec: number;
+    durationSec: number;
+}
+/**
+ * A calculated split point
+ */
+export interface SplitPoint {
+    /** Time in seconds where to split (middle of silence) */
+    timeSec: number;
+    /** Duration of the silence at this split point */
+    silenceDuration: number;
+}
+/**
+ * An audio segment after splitting
+ */
+export interface AudioSegment {
+    index: number;
+    startSec: number;
+    endSec: number;
+    durationSec: number;
+    outputPath: string;
+}
+/**
+ * Default split configuration
+ */
+export declare const DEFAULT_SPLIT_CONFIG: SplitConfig;
+/**
+ * Get the duration of an audio file in seconds
+ */
+export declare function getAudioDuration(audioPath: string): Promise<number>;
+/**
+ * Detect silence regions in an audio file using FFmpeg
+ * Uses spawn directly for better compatibility
+ */
+export declare function detectSilenceRegions(audioPath: string, config?: Partial<SplitConfig>): Promise<SilenceRegion[]>;
+/**
+ * Find optimal split points in audio based on silence regions
+ * Prefers splitting at longer silences when possible
+ *
+ * @param silenceRegions - Detected silence regions
+ * @param totalDuration - Total audio duration in seconds
+ * @param config - Split configuration
+ * @returns Array of optimal split points
+ */
+export declare function findOptimalSplitPoints(silenceRegions: SilenceRegion[], totalDuration: number, config?: Partial<SplitConfig>): SplitPoint[];
+/**
+ * Split audio file at specified points using FFmpeg
+ *
+ * @param audioPath - Path to source audio file
+ * @param splitPoints - Where to split the audio
+ * @param totalDuration - Total duration of source audio
+ * @param outputDir - Directory to write segments
+ * @param baseName - Base name for output files
+ * @returns Array of created audio segments
+ */
+export declare function splitAudioAtPoints(audioPath: string, splitPoints: SplitPoint[], totalDuration: number, outputDir: string, baseName: string): Promise<AudioSegment[]>;
+/**
+ * Auto-split an audio file if it exceeds the maximum duration
+ * Returns the original file path if no split is needed
+ *
+ * @param audioPath - Path to source audio file
+ * @param outputDir - Directory for split segments
+ * @param config - Split configuration
+ * @returns Array of audio segment paths (single element if no split needed)
+ */
+export declare function autoSplitAudio(audioPath: string, outputDir: string, config?: Partial<SplitConfig>): Promise<AudioSegment[]>;
+/**
+ * Information about split points for logging/debugging
+ */
+export interface SplitAnalysis {
+    totalDuration: number;
+    numSegments: number;
+    splitPoints: SplitPoint[];
+    silenceRegions: SilenceRegion[];
+    needsSplit: boolean;
+}
+/**
+ * Analyze audio file and return split information without actually splitting
+ * Useful for preview/dry-run functionality
+ */
+export declare function analyzeSplitPoints(audioPath: string, config?: Partial<SplitConfig>): Promise<SplitAnalysis>;
+//# sourceMappingURL=split.d.ts.map

package/dist/utils/audio/split.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"split.d.ts","sourceRoot":"","sources":["../../../src/utils/audio/split.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAOH;;GAEG;AACH,MAAM,WAAW,WAAW;IAC1B,iEAAiE;IACjE,cAAc,EAAE,MAAM,CAAA;IACtB,qEAAqE;IACrE,gBAAgB,EAAE,MAAM,CAAA;IACxB,kDAAkD;IAClD,gBAAgB,EAAE,MAAM,CAAA;IACxB,wDAAwD;IACxD,mBAAmB,EAAE,OAAO,CAAA;IAC5B,uDAAuD;IACvD,aAAa,EAAE,MAAM,CAAA;CACtB;AAED;;GAEG;AACH,MAAM,WAAW,aAAa;IAC5B,QAAQ,EAAE,MAAM,CAAA;IAChB,MAAM,EAAE,MAAM,CAAA;IACd,WAAW,EAAE,MAAM,CAAA;CACpB;AAED;;GAEG;AACH,MAAM,WAAW,UAAU;IACzB,yDAAyD;IACzD,OAAO,EAAE,MAAM,CAAA;IACf,kDAAkD;IAClD,eAAe,EAAE,MAAM,CAAA;CACxB;AAED;;GAEG;AACH,MAAM,WAAW,YAAY;IAC3B,KAAK,EAAE,MAAM,CAAA;IACb,QAAQ,EAAE,MAAM,CAAA;IAChB,MAAM,EAAE,MAAM,CAAA;IACd,WAAW,EAAE,MAAM,CAAA;IACnB,UAAU,EAAE,MAAM,CAAA;CACnB;AAED;;GAEG;AACH,eAAO,MAAM,oBAAoB,EAAE,WAMlC,CAAA;AAED;;GAEG;AACH,wBAAsB,gBAAgB,CAAC,SAAS,EAAE,MAAM,GAAG,OAAO,CAAC,MAAM,CAAC,CAezE;AAED;;;GAGG;AACH,wBAAsB,oBAAoB,CACxC,SAAS,EAAE,MAAM,EACjB,MAAM,GAAE,OAAO,CAAC,WAAW,CAAM,GAChC,OAAO,CAAC,aAAa,EAAE,CAAC,CAsD1B;AAED;;;;;;;;GAQG;AACH,wBAAgB,sBAAsB,CACpC,cAAc,EAAE,aAAa,EAAE,EAC/B,aAAa,EAAE,MAAM,EACrB,MAAM,GAAE,OAAO,CAAC,WAAW,CAAM,GAChC,UAAU,EAAE,CAwEd;AAED;;;;;;;;;GASG;AACH,wBAAsB,kBAAkB,CACtC,SAAS,EAAE,MAAM,EACjB,WAAW,EAAE,UAAU,EAAE,EACzB,aAAa,EAAE,MAAM,EACrB,SAAS,EAAE,MAAM,EACjB,QAAQ,EAAE,MAAM,GACf,OAAO,CAAC,YAAY,EAAE,CAAC,CA8CzB;AAED;;;;;;;;GAQG;AACH,wBAAsB,cAAc,CAClC,SAAS,EAAE,MAAM,EACjB,SAAS,EAAE,MAAM,EACjB,MAAM,GAAE,OAAO,CAAC,WAAW,CAAM,GAChC,OAAO,CAAC,YAAY,EAAE,CAAC,CAuCzB;AAED;;GAEG;AACH,MAAM,WAAW,aAAa;IAC5B,aAAa,EAAE,MAAM,CAAA;IACrB,WAAW,EAAE,MAAM,CAAA;IACnB,WAAW,EAAE,UAAU,EAAE,CAAA;IACzB,cAAc,EAAE,aAAa,EAAE,CAAA;IAC/B,UAAU,EAAE,OAAO,CAAA;CACpB;AAED;;;GAGG;AACH,wBAAsB,kBAAkB,CACtC,SAAS,EAAE,MAAM,EACjB,MAAM,GAAE,OAAO,CAAC,WAAW,CAAM,GAChC,OAAO,CAAC,aAAa,CAAC,CA0BxB"}

package/dist/utils/file-utils.d.ts

@@ -0,0 +1,6 @@
+import path from 'path';
+export declare const getExt: (filePath: string) => string;
+export declare const getName: (filePath: string) => string;
+export declare const getNameWithExt: (filePath: string) => string;
+export declare const getFileInfo: (filePath: string) => path.ParsedPath;
+//# sourceMappingURL=file-utils.d.ts.map

package/dist/utils/file-utils.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"file-utils.d.ts","sourceRoot":"","sources":["../../src/utils/file-utils.ts"],"names":[],"mappings":"AAAA,OAAO,IAAI,MAAM,MAAM,CAAA;AAEvB,eAAO,MAAM,MAAM,GAAI,UAAU,MAAM,WAEtC,CAAA;AACD,eAAO,MAAM,OAAO,GAAI,UAAU,MAAM,WAEvC,CAAA;AACD,eAAO,MAAM,cAAc,GAAI,UAAU,MAAM,WAG9C,CAAA;AACD,eAAO,MAAM,WAAW,GAAI,UAAU,MAAM,oBAG3C,CAAA"}

package/dist/utils/transcription/format.d.ts

@@ -0,0 +1,14 @@
+/**
+ * Format transcript with line breaks based on pauses between words
+ * @param transcript The punctuated transcript text from the API
+ * @param words Array of word objects with timing data
+ * @param shortPauseThreshold Threshold for single line break (default: 1.0s)
+ * @param longPauseThreshold Threshold for paragraph break/double newline (default: 5.0s)
+ */
+export declare function formatTranscriptWithPauses(transcript: string, words: Array<{
+    word: string;
+    start: number;
+    end: number;
+    confidence: number;
+}>, shortPauseThreshold?: number, longPauseThreshold?: number): string;
+//# sourceMappingURL=format.d.ts.map

package/dist/utils/transcription/format.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"format.d.ts","sourceRoot":"","sources":["../../../src/utils/transcription/format.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AACH,wBAAgB,0BAA0B,CACzC,UAAU,EAAE,MAAM,EAClB,KAAK,EAAE,KAAK,CAAC;IAAE,IAAI,EAAE,MAAM,CAAC;IAAC,KAAK,EAAE,MAAM,CAAC;IAAC,GAAG,EAAE,MAAM,CAAC;IAAC,UAAU,EAAE,MAAM,CAAA;CAAE,CAAC,EAC9E,mBAAmB,SAAM,EACzB,kBAAkB,SAAM,GACtB,MAAM,CAgDR"}

package/dist/utils/transcription/mime-detection.d.ts

@@ -0,0 +1,25 @@
+import { Buffer } from 'node:buffer';
+/**
+ * Detect audio format from buffer magic bytes
+ *
+ * @param buffer - The audio buffer to analyze
+ * @returns The detected MIME type string ('audio/mpeg', 'audio/ogg', 'audio/wav', 'audio/flac')
+ *
+ * @description
+ * This function analyzes the magic bytes at the beginning of an audio buffer to determine
+ * its format. It handles MP3 files with ID3 tags by skipping over them before checking
+ * for the actual audio frame header.
+ *
+ * Supported formats:
+ * - MP3 (MPEG Audio): FF FB, FF FA, FF F3, FF F2 with optional ID3 tag
+ * - OGG: "OggS" (4F 67 67 53)
+ * - WAV (RIFF): "RIFF" (52 49 46 46)
+ * - FLAC: "fLaC" (66 4C 61 43)
+ *
+ * @example
+ * const buffer = readFileSync('audio.mp3')
+ * const mimeType = detectAudioMimeType(buffer)
+ * console.log(mimeType) // 'audio/mpeg'
+ */
+export declare function detectAudioMimeType(buffer: Buffer): string;
+//# sourceMappingURL=mime-detection.d.ts.map

package/dist/utils/transcription/mime-detection.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"mime-detection.d.ts","sourceRoot":"","sources":["../../../src/utils/transcription/mime-detection.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,MAAM,EAAE,MAAM,aAAa,CAAA;AAEpC;;;;;;;;;;;;;;;;;;;;;GAqBG;AACH,wBAAgB,mBAAmB,CAAC,MAAM,EAAE,MAAM,GAAG,MAAM,CAuD1D"}

package/dist/utils/transcription/providers.d.ts

@@ -0,0 +1,146 @@
+/**
+ * Transcription provider types and interfaces
+ * Defines the contract for all transcription providers
+ */
+/**
+ * Result object returned from transcription operations
+ * Contains the transcribed text and optional provider-specific metadata
+ */
+export interface TranscriptionResult {
+    /** The transcribed text content */
+    text: string;
+    /** Raw response object from the provider (optional, for debugging) */
+    rawResponse?: any;
+    /** Error message if transcription failed */
+    error?: string;
+    /** Confidence score of the transcription (0-1) */
+    confidence?: number;
+    /** Word-level data (timing, confidence, etc.) */
+    words?: any[];
+    /** Duration of audio in seconds */
+    duration?: number;
+    /** Detected or specified language code */
+    language?: string;
+    /** Model used for transcription (as returned by provider) */
+    model?: string;
+}
+/**
+ * Interface that all transcription providers must implement
+ * Defines the standard contract for transcription functionality
+ */
+export interface TranscriptionProvider {
+    /** Provider name/identifier */
+    name: string;
+    /** Maximum audio duration in seconds (undefined = no limit) */
+    maxAudioDurationSec?: number;
+    /**
+     * Transcribe audio from the given parameters
+     * @param params - Transcription parameters
+     * @returns Promise resolving to transcription result
+     */
+    transcribe(params: TranscribeParams): Promise<TranscriptionResult>;
+}
+/**
+ * Parameters for transcription operations
+ * Supports both common and provider-specific options
+ */
+export interface TranscribeParams {
+    /** Path to the audio file to transcribe */
+    audioPath?: string;
+    /** Audio buffer to transcribe */
+    audioBuffer?: Buffer;
+    /** MIME type for audioBuffer (auto-detected if not provided) */
+    mimeType?: string;
+    /** URL to audio file (e.g., IPFS gateway URL) */
+    audioUrl?: string;
+    /** Model to use for transcription (provider-specific) */
+    model?: string;
+    /** Language code for transcription (e.g., 'en', 'fr') */
+    language?: string;
+    /** API key for authentication (provider-specific) */
+    apiKey?: string;
+    /** Enable speaker diarization (Whisper-specific) */
+    diarize?: boolean;
+    /** Timestamp granularity for transcription (Mistral-specific) */
+    timestampGranularity?: 'segment' | 'word';
+    /** Path to model file (Whisper-specific) */
+    modelPath?: string;
+    /** Output directory for results (Whisper-specific) */
+    outputDir?: string;
+    /** Provider configuration object */
+    config?: any;
+}
+/**
+ * Union type for supported provider names
+ */
+export type ProviderName = 'whisper' | 'mistral' | 'greenpt';
+/**
+ * Factory function to create a transcription provider instance
+ * @param providerName - Name of the provider to create
+ * @param config - Optional configuration object for the provider
+ * @returns Transcription provider instance
+ */
+export declare function createProvider(providerName: ProviderName, config?: any): TranscriptionProvider;
+/**
+ * Whisper provider for local whisper.cpp transcription
+ * Manages model caching and local transcription execution
+ */
+export declare class WhisperProvider implements TranscriptionProvider {
+    name: string;
+    private cacheDir;
+    static DEFAULTS: {
+        DIARIZE: boolean;
+        SILDUR: string;
+        SILBUF: number;
+        SILTHR: string;
+        MODEL_KEYS: {
+            tinyd: string;
+            small: string;
+            medium: string;
+        };
+        MODELS: {
+            tinyd: string;
+            small: string;
+            medium: string;
+        };
+    };
+    constructor(config?: any);
+    transcribe(params: TranscribeParams): Promise<TranscriptionResult>;
+    private ensureRequestedModelIsCached;
+}
+/**
+ * Voxtral/Mistral API limits
+ * These may need adjustment based on actual API constraints
+ */
+export declare const VOXTRAL_LIMITS: {
+    /** Maximum audio duration in seconds (3 hours for Voxtral Transcribe 2) */
+    maxAudioDurationSec: number;
+    /** Recommended max duration before splitting (for reliability) */
+    recommendedMaxDurationSec: number;
+    /** Maximum context biasing words/phrases */
+    maxContextBiasingTerms: number;
+    /** Maximum file size in bytes (1GB) */
+    maxFileSizeBytes: number;
+};
+export declare class MistralProvider implements TranscriptionProvider {
+    name: string;
+    maxAudioDurationSec: number;
+    /**
+     * Check if audio duration exceeds recommended limits
+     */
+    static shouldSplit(durationSec: number): boolean;
+    /**
+     * Get the recommended max segment duration for splitting
+     */
+    static getRecommendedMaxSegment(): number;
+    transcribe(params: TranscribeParams): Promise<TranscriptionResult>;
+}
+/**
+ * GreenPT transcription provider
+ * Uses GreenPT API for audio transcription with Deepgram-compatible response format
+ */
+export declare class GreenPTProvider implements TranscriptionProvider {
+    name: string;
+    transcribe(params: TranscribeParams): Promise<TranscriptionResult>;
+}
+//# sourceMappingURL=providers.d.ts.map

package/dist/utils/transcription/providers.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"providers.d.ts","sourceRoot":"","sources":["../../../src/utils/transcription/providers.ts"],"names":[],"mappings":"AAAA;;;GAGG;AAUH;;;GAGG;AACH,MAAM,WAAW,mBAAmB;IAClC,mCAAmC;IACnC,IAAI,EAAE,MAAM,CAAA;IACZ,sEAAsE;IACtE,WAAW,CAAC,EAAE,GAAG,CAAA;IACjB,4CAA4C;IAC5C,KAAK,CAAC,EAAE,MAAM,CAAA;IACd,kDAAkD;IAClD,UAAU,CAAC,EAAE,MAAM,CAAA;IACnB,iDAAiD;IACjD,KAAK,CAAC,EAAE,GAAG,EAAE,CAAA;IACb,mCAAmC;IACnC,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB,0CAA0C;IAC1C,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB,6DAA6D;IAC7D,KAAK,CAAC,EAAE,MAAM,CAAA;CACf;AAED;;;GAGG;AACH,MAAM,WAAW,qBAAqB;IACpC,+BAA+B;IAC/B,IAAI,EAAE,MAAM,CAAA;IACZ,+DAA+D;IAC/D,mBAAmB,CAAC,EAAE,MAAM,CAAA;IAC5B;;;;OAIG;IACH,UAAU,CAAC,MAAM,EAAE,gBAAgB,GAAG,OAAO,CAAC,mBAAmB,CAAC,CAAA;CACnE;AAED;;;GAGG;AACH,MAAM,WAAW,gBAAgB;IAC/B,2CAA2C;IAC3C,SAAS,CAAC,EAAE,MAAM,CAAA;IAClB,iCAAiC;IACjC,WAAW,CAAC,EAAE,MAAM,CAAA;IACpB,gEAAgE;IAChE,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB,iDAAiD;IACjD,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB,yDAAyD;IACzD,KAAK,CAAC,EAAE,MAAM,CAAA;IACd,yDAAyD;IACzD,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB,qDAAqD;IACrD,MAAM,CAAC,EAAE,MAAM,CAAA;IACf,oDAAoD;IACpD,OAAO,CAAC,EAAE,OAAO,CAAA;IACjB,iEAAiE;IACjE,oBAAoB,CAAC,EAAE,SAAS,GAAG,MAAM,CAAA;IACzC,4CAA4C;IAC5C,SAAS,CAAC,EAAE,MAAM,CAAA;IAClB,sDAAsD;IACtD,SAAS,CAAC,EAAE,MAAM,CAAA;IAClB,oCAAoC;IACpC,MAAM,CAAC,EAAE,GAAG,CAAA;CACb;AAED;;GAEG;AACH,MAAM,MAAM,YAAY,GAAG,SAAS,GAAG,SAAS,GAAG,SAAS,CAAA;AAE5D;;;;;GAKG;AACH,wBAAgB,cAAc,CAAC,YAAY,EAAE,YAAY,EAAE,MAAM,CAAC,EAAE,GAAG,GAAG,qBAAqB,CAW9F;AASD;;;GAGG;AACH,qBAAa,eAAgB,YAAW,qBAAqB;IAC3D,IAAI,SAAY;IAEhB,OAAO,CAAC,QAAQ,CAAQ;IAExB,MAAM,CAAC,QAAQ;;;;;;;;;;;;;;;MAkBd;gBAEW,MAAM,CAAC,EAAE,GAAG;IAKlB,UAAU,CAAC,MAAM,EAAE,gBAAgB,GAAG,OAAO,CAAC,mBAAmB,CAAC;YA+F1D,4BAA4B;CAkC3C;AAED;;;GAGG;AACH,eAAO,MAAM,cAAc;IACzB,2EAA2E;;IAE3E,kEAAkE;;IAElE,4CAA4C;;IAE5C,uCAAuC;;CAExC,CAAA;AAED,qBAAa,eAAgB,YAAW,qBAAqB;IAC3D,IAAI,SAAY;IAChB,mBAAmB,SAAqC;IAExD;;OAEG;IACH,MAAM,CAAC,WAAW,CAAC,WAAW,EAAE,MAAM,GAAG,OAAO;IAIhD;;OAEG;IACH,MAAM,CAAC,wBAAwB,IAAI,MAAM;IAInC,UAAU,CAAC,MAAM,EAAE,gBAAgB,GAAG,OAAO,CAAC,mBAAmB,CAAC;CA8FzE;AAED;;;GAGG;AACH,qBAAa,eAAgB,YAAW,qBAAqB;IAC3D,IAAI,SAAY;IAEV,UAAU,CAAC,MAAM,EAAE,gBAAgB,GAAG,OAAO,CAAC,mBAAmB,CAAC;CAkGzE"}

package/dist/utils/transcription/transcribe.d.ts

@@ -0,0 +1,59 @@
+/**
+ * Simple high-level transcription API with good defaults
+ */
+import { type MergedTranscriptionResult } from '../audio/merge-results.js';
+export interface TranscribeOptions {
+    /** Path to audio file */
+    audioPath?: string;
+    /** Audio buffer to transcribe */
+    audioBuffer?: Buffer;
+    /** MIME type for audioBuffer (auto-detected if not provided) */
+    mimeType?: string;
+    /** URL to audio file (e.g., IPFS gateway URL) */
+    audioUrl?: string;
+    /** Language code (e.g. 'en', 'fr') - note: disables word timestamps for Mistral */
+    language?: string;
+    /** Model to use (default: voxtral-mini-latest) */
+    model?: string;
+    /** Enable speaker diarization (default: true) */
+    diarize?: boolean;
+    /** Timestamp granularity: 'word' | 'segment' (default: 'word', disabled if language set) */
+    timestamps?: 'word' | 'segment';
+    /** Auto-split long audio (default: true, only works with audioPath) */
+    autoSplit?: boolean;
+    /** Output directory for split segments (default: system temp) */
+    splitOutputDir?: string;
+}
+export interface MistralTranscriberConfig {
+    /** Mistral API key */
+    apiKey: string;
+    /** Default model (default: voxtral-mini-latest) */
+    model?: string;
+}
+/**
+ * Simple Mistral transcriber with auto-splitting and good defaults
+ *
+ * @example
+ * ```ts
+ * const transcriber = createMistralTranscriber({ apiKey: process.env.MISTRAL_API_KEY })
+ *
+ * // From file (supports auto-split for long audio)
+ * const result = await transcriber.transcribe({ audioPath: './interview.mp3' })
+ *
+ * // From URL (e.g., IPFS gateway)
+ * const result = await transcriber.transcribe({ audioUrl: 'https://gateway.ipfs.io/ipfs/Qm...' })
+ *
+ * // From buffer
+ * const result = await transcriber.transcribe({ audioBuffer: buffer, mimeType: 'audio/mpeg' })
+ * ```
+ */
+export declare function createMistralTranscriber(config: MistralTranscriberConfig): {
+    /**
+     * Transcribe audio with auto-splitting for long files (file path only)
+     * Diarization and word timestamps enabled by default
+     */
+    transcribe(options: TranscribeOptions): Promise<MergedTranscriptionResult>;
+};
+/** Alias for simpler import */
+export declare const transcribe: typeof createMistralTranscriber;
+//# sourceMappingURL=transcribe.d.ts.map

package/dist/utils/transcription/transcribe.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"transcribe.d.ts","sourceRoot":"","sources":["../../../src/utils/transcription/transcribe.ts"],"names":[],"mappings":"AAAA;;GAEG;AAOH,OAAO,EAA6B,KAAK,yBAAyB,EAAE,MAAM,2BAA2B,CAAA;AAErG,MAAM,WAAW,iBAAiB;IAChC,yBAAyB;IACzB,SAAS,CAAC,EAAE,MAAM,CAAA;IAClB,iCAAiC;IACjC,WAAW,CAAC,EAAE,MAAM,CAAA;IACpB,gEAAgE;IAChE,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB,iDAAiD;IACjD,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB,mFAAmF;IACnF,QAAQ,CAAC,EAAE,MAAM,CAAA;IACjB,kDAAkD;IAClD,KAAK,CAAC,EAAE,MAAM,CAAA;IACd,iDAAiD;IACjD,OAAO,CAAC,EAAE,OAAO,CAAA;IACjB,4FAA4F;IAC5F,UAAU,CAAC,EAAE,MAAM,GAAG,SAAS,CAAA;IAC/B,uEAAuE;IACvE,SAAS,CAAC,EAAE,OAAO,CAAA;IACnB,iEAAiE;IACjE,cAAc,CAAC,EAAE,MAAM,CAAA;CACxB;AAED,MAAM,WAAW,wBAAwB;IACvC,sBAAsB;IACtB,MAAM,EAAE,MAAM,CAAA;IACd,mDAAmD;IACnD,KAAK,CAAC,EAAE,MAAM,CAAA;CACf;AAED;;;;;;;;;;;;;;;;GAgBG;AACH,wBAAgB,wBAAwB,CAAC,MAAM,EAAE,wBAAwB;IAKrE;;;OAGG;wBACuB,iBAAiB,GAAG,OAAO,CAAC,yBAAyB,CAAC;EA4EnF;AAED,+BAA+B;AAC/B,eAAO,MAAM,UAAU,iCAA2B,CAAA"}

package/package.json

@@ -0,0 +1,62 @@
+{
+  "name": "@wovin/tranz",
+  "version": "0.0.26",
+  "type": "module",
+  "description": "Audio transcription library with provider support and auto-splitting",
+  "author": "gotjoshua @gotjoshua",
+  "license": "ISC",
+  "homepage": "https://gitlab.com/onezoomin/ztax/tranz",
+  "repository": "https://gitlab.com/onezoomin/ztax/tranz",
+  "bugs": "https://gitlab.com/onezoomin/ztax/tranz/-/issues",
+  "main": "./dist/index.min.js",
+  "module": "./dist/index.min.js",
+  "types": "./dist/index.d.ts",
+  "exports": {
+    ".": {
+      "import": "./dist/index.min.js",
+      "types": "./dist/index.d.ts"
+    },
+    "./providers": {
+      "import": "./dist/providers.min.js",
+      "types": "./dist/providers.d.ts"
+    },
+    "./audio": {
+      "import": "./dist/audio.min.js",
+      "types": "./dist/audio.d.ts"
+    }
+  },
+  "files": [
+    "./dist/"
+  ],
+  "publishConfig": {
+    "access": "public"
+  },
+  "dependencies": {
+    "fluent-ffmpeg": "^2.1.2"
+  },
+  "devDependencies": {
+    "@types/fluent-ffmpeg": "^2.1.21",
+    "@types/node": "^24.10.1",
+    "concurrently": "^8.2.2",
+    "tsup": "^8.5.0",
+    "typescript": "^5.9.3",
+    "tsupconfig": "^0.0.0"
+  },
+  "keywords": [
+    "transcription",
+    "audio",
+    "mistral",
+    "whisper",
+    "voxtral"
+  ],
+  "scripts": {
+    "build": "rm -rf dist/ && concurrently \"pnpm build:code\" \"pnpm build:types\"",
+    "build:code": "tsup",
+    "build:types": "tsc --emitDeclarationOnly --declaration",
+    "dev": "concurrently \"pnpm dev:code\" \"pnpm dev:types\"",
+    "dev:code": "tsup --watch",
+    "dev:types": "tsc --emitDeclarationOnly --declaration --watch",
+    "clean": "rm -rf .turbo && rm -rf node_modules && rm -rf dist",
+    "pub": "npm publish --tag latest --access=public"
+  }
+}