Move shared type contracts from `contracts` to new `api-contracts` module in `compute/core` for clearer API boundaries. Remove legacy `contracts`, `pdf`, and `whisper` index files. Update all imports to use `api-contracts`. Refactor server job logic by replacing `parsePdfJob.ts` with modular `user-pdf-layout-job.ts` and `user-whisper-align-job.ts`, updating all relevant API routes and compute integration. This improves maintainability and separation of concerns across compute and server layers.
Move and re-export all PDF parsing and TTS-related types from
@openreader/compute-core/types, eliminating local type duplication.
Update imports throughout the codebase to use the new centralized type
module. Remove obsolete worker-contract file and update tsconfig paths
and compute-core exports for the new types entry point. Refactor API
routes and job logic to use the new type imports and shared parse-state
healing utility. This streamlines type management and improves
consistency between compute and app layers.
Introduce automatic healing for stale PDF parse states in API routes. Add
`isDocumentParseStateStale` utility and use a shared `COMPUTE_OP_STALE_MS`
environment variable (with config getter) to control the stale window for both
worker op replacement and app-side parse-state healing. Update documentation and
environment variable references to reflect new config.
Enables automatic marking of stuck parses as failed, improving reliability and
retry behavior in distributed compute environments.
Update compute worker to use separate environment variable for PDF job attempts
(`COMPUTE_PDF_JOB_ATTEMPTS`) and set Whisper job max deliveries to 1, clarifying
and separating retry logic between job types. Adjust `.env.example` and deployment
docs to comment out defaults and document new/renamed variables for advanced tuning.
In TTS segment ensure route, introduce explicit timeouts for Whisper alignment
operations, with configurable durations for local and worker modes, improving
robustness and error handling for long-running alignments.
Introduce detailed progress reporting for PDF layout parsing, exposing phase and page-level updates to clients. Replace legacy parseStatus with a structured parseState field in the database schema, updating all relevant backend and API logic. Add SSE endpoint for real-time parse progress updates. Update client hooks and UI to reflect granular progress and improve user feedback during document parsing. Includes migration scripts and new parse-state utility module.
Migrate worker and backend contract to an operation-centric model, introducing opKey-based deduplication for Whisper alignment and PDF layout parsing. Add new operation request/state types and update worker endpoints to accept generic operation requests and expose operation status via SSE. Refactor client and server logic to support operation flows. Update deployment docs, environment examples, and Dockerfile to reflect new architecture and dependency management.
Add ability to fetch parsed documents using a stored S3 key, enabling more flexible blob access patterns. Update API route to prefer `parsedJsonKey` if present, falling back to legacy lookup. Introduce `COMPUTE_WORKER_WAIT_TIMEOUT_MS` environment variable for customizing worker job wait time, with a new default of 120 seconds. Update documentation and examples to reflect these changes.
Eliminate the smartSentenceSplitting configuration flag from application state, UI, and all TTS and audiobook adapter logic. Consolidate TTS segment planning to always include context units, streamlining code and reducing conditional branches. Update tests and type definitions to reflect the removal of this feature toggle.
Refactor compute backend initialization to use asynchronous loading and promise caching, replacing previous synchronous singleton pattern. Update all consumers—including TTS segment alignment and PDF layout parsing jobs—to await compute backend resolution before invoking methods. This change improves compatibility with dynamic imports and future-proofs backend selection logic.
Eliminate unused PDF text extraction logic and margin-based filtering from the client library. Refactor API routes to generalize parse status normalization, removing explicit references to deprecated "unsupported" status and type variants. Update usages to accept string parse statuses for improved flexibility. Simplify document state dependency tracking in the PDF document hook.
Eliminate the "none" compute mode and all related code paths, including the NoneComputeBackend, "unsupported" parse status, and PDF margin settings. Parsing is now always available if the app starts successfully, and configuration is limited to "local" or "worker" compute modes. Update types, API routes, client adapters, and documentation to reflect this simplification.
BREAKING CHANGE: "none" is no longer a valid COMPUTE_MODE; only "local" and "worker" are supported. "unsupported" parse status and PDF margin settings are removed.
Introduce support for external compute worker mode (`COMPUTE_MODE=worker`) using a new `WorkerComputeBackend`. This enables offloading heavy ONNX Whisper alignment and PDF layout parsing to a standalone worker service (Redis + BullMQ), improving scalability and compatibility with serverless/limited environments.
- Add `@openreader/compute-core` as a shared package for ONNX inference and PDF parsing logic.
- Implement `WorkerComputeBackend` and worker contract/types for remote job execution.
- Update compute backend selection logic and remove previous worker mode guards.
- Extend `WhisperAlignInput` and `PdfLayoutInput` types to support object keys for remote data access.
- Refactor local compute backend to use `@openreader/compute-core` and support both buffer and object key inputs.
- Update job runner, TTS segment alignment, and PDF layout parsing flows to use new compute backend APIs.
- Add scripts, Docker workflow, and documentation for deploying and running the compute worker.
- Update environment variable docs and examples for worker mode, including storage requirements and configuration.
- Document published images and stack changes to reflect the new compute worker architecture.
BREAKING CHANGE: `COMPUTE_MODE=worker` now requires an external compute worker service and S3-compatible object storage. Embedded SeaweedFS (`weed mini`) is not supported in worker mode. See the new documentation for deployment and configuration details.
Replace the previous whisper.cpp-based word alignment with a fully ONNX-based
implementation using onnxruntime-node and @huggingface/tokenizers. Add new
Whisper ONNX model management, alignment mapping, and spectral analysis modules.
Remove all code and documentation referencing whisper.cpp, update environment
variables, Dockerfile, and docs to reflect ONNX-only alignment. Add unit tests
for alignment and ONNX model logic.
Migrate PDF parsing to use ONNX Docling layout model for structured block extraction, enabling improved TTS segmentation and chaptering. Add compute backend abstraction for heavy tasks (alignment, layout parsing) with configuration via `OPENREADER_COMPUTE_MODE`. Integrate block-level locators and document settings for per-document PDF parsing options. Update S3 storage for parsed PDF JSON, add migration and schema changes, and extend client and server APIs for parsed data and settings. Remove legacy word highlight flag in favor of compute capability detection.
- Add ONNX model download, local/none compute modes, and `onnxruntime-node` dependency
- Update PDF viewer and TTS pipeline to use parsed blocks and block-level locators
- Add document settings storage and APIs for per-document PDF options
- Update S3 storage layout for parsed PDF JSON
- Add admin/config/docs updates for new compute and parsing features
- Remove obsolete word highlight runtime flag and UI
BREAKING CHANGE: PDF parsing and word highlighting now require `OPENREADER_COMPUTE_MODE=local` and ONNX model; document settings and S3 layout updated; legacy word highlight flag removed.
Introduce sentence and word highlighting for HTML/TXT/MD documents, including new config options (`htmlHighlightEnabled`, `htmlWordHighlightEnabled`) and synced user preferences. Update the HTML viewer to support these highlights using custom spans, and add corresponding CSS styles. Extend document settings to allow toggling these features. Implement block-level locator sorting and display improvements. Add audiobook export support for HTML documents, including adapter and pipeline integration.
- Add `htmlHighlightEnabled` and `htmlWordHighlightEnabled` to config, user state, and Dexie storage
- Update HTML viewer to apply and manage highlights for sentences and words
- Add CSS for HTML highlight classes and scroll margin
- Extend document settings UI for HTML highlight toggles
- Support HTML audiobook export in modal and pipeline
- Improve block locator sorting and sidebar display for HTML
- Add HTML audiobook adapter and block parsing utilities
Update GitHub Actions workflows to latest action versions and improve pnpm usage by introducing per-package workspace configs. Add docs-site/pnpm-workspace.yaml and update Dockerfile/package.json for better monorepo support. Upgrade dependencies across main and docs-site packages for compatibility and security. Refine TypeScript types and hooks for improved type safety and maintainability.
Refactor chapter API to use new coerceAudiobookGenerationSettings utility for
validating and migrating audiobook metadata. Add src/lib/server/audiobooks/settings.ts
with shared logic and introduce corresponding unit tests to ensure correct
settings migration and validation.
Integrate changelog version check on settings open, triggering the changelog
panel when a new app version is detected. Refactor user preferences storage to
support metadata via payload utilities, enabling version tracking and future
extensibility. Add API endpoint for changelog version check and related tests
for changelog and preferences payload logic.
Transition all TTS-related logic, types, and UI to use the new providerRef/providerType model in place of legacy ttsProvider fields. Introduce a centralized tts-provider-policy module to encapsulate provider/model capability checks, default value resolution, and compatibility logic. Update all API routes, contexts, hooks, components, and tests to use providerRef and providerType, ensuring consistent handling of built-in and shared TTS providers. Remove legacy defaultTtsModel config in favor of per-provider defaults and shared provider admin control. Add the showAllProviderModels runtime flag to restrict users to provider default models when desired.
BREAKING CHANGE: ttsProvider fields are replaced by providerRef/providerType throughout the codebase; defaultTtsModel config is removed in favor of per-provider defaults.
Add database-backed runtime configuration for feature flags and TTS provider credentials, editable via a new admin UI panel. Replace static NEXT_PUBLIC_* and TTS provider env vars with admin-managed settings stored in the database and injected at SSR for client access. Implement admin-only panels for managing shared TTS provider credentials (with encrypted API keys) and live site feature flags. Add schema migrations, API routes, React contexts, and hooks for SSR-injected runtime config and live updates. Update client and server logic to resolve configuration from the database at runtime, enforcing admin restrictions and supporting migration from legacy env-based config.
BREAKING CHANGE: Feature flags and TTS provider credentials are now managed at runtime via the admin UI. Environment variables are only used for initial seeding and are ignored after first boot. Existing deployments must migrate configuration to the admin panel.
Eliminate the deprecated `/api/tts` route and its client usage, consolidating TTS generation logic under chapter and segment endpoints. Update client API utilities to remove direct TTS buffer fetching and improve error propagation for chapter generation failures. Extend TTS segment error metadata to include upstream rate limit and retry details, and standardize error typing in shared types. Adjust tests and helpers to reflect the new TTS mocking approach and endpoint structure. This streamlines TTS API surface and enhances error reporting consistency.
Move daily quota exceeded problem response logic to a dedicated utility in `lib/server/rate-limit/problem-response`. Refactor API routes to use the new builder and remove duplicated formatting code. Update all relevant imports for consistency. This improves maintainability and ensures uniform error responses across TTS endpoints.
Redesign TTS segment storage by decomposing the `tts_segments` table into two normalized tables: `tts_segment_entries` for unique segment identity and locator projection, and `tts_segment_variants` for per-settings audio variants. Update schema, queries, and API routes to use the new structure, including manifest pagination and segment resolution logic. Refactor segment locator handling to use stable projections and manifest cursors. Migrate tests and data cleanup utilities to support the new model.
BREAKING CHANGE: Database schema for TTS segments is now split into entry and variant tables; all code and consumers must use the new structure. Persisted data and APIs relying on the old `tts_segments` table are incompatible.
Implement strict validation for EPUB TTS segment locators, requiring stable spine coordinates (`spineHref`, `spineIndex`, `charOffset`) and rejecting legacy CFI-only locators. Introduce `segmentKey` as a normalized identity derived from segment text, used for robust merging and deduplication of synthesized and persisted segments. Update database schemas, API routes, and manifest logic to support the new locator format and segmentKey. Add helpers for locator resolution and segmentKey construction, with comprehensive unit tests for identity, normalization, and manifest grouping.
BREAKING CHANGE: Persisted EPUB TTS segment locators must now include stable spine coordinates; legacy CFI-only locators are no longer accepted. Consumers must handle the new `segmentKey` field for segment identity and merging.
Refactor TTS segment identification to support canonical segment keys, decoupling segment IDs from locator and index for more robust deduplication and handoff. Add `segmentKey` to segment types, manifest, and ensure route. Implement natural sorting for EPUB CFI locations and segment locators using new comparison utilities. Update sidebar and manifest helpers to use locator-aware sorting. Add foundational modules for EPUB word highlighting, TTS segment planning, and EPUB handoff logic. Expand types and tests to cover new canonical segment and locator behaviors.
BREAKING CHANGE: TTS segment IDs now support canonical segment keys; APIs and manifest consumers must handle `segmentKey` and updated locator comparison logic.
Introduce configurable segment preloading depth, sentence lookahead, and TTS segment max block length for both PDF and EPUB readers. Add new settings to user preferences, config context, and document settings UI. Refactor TTS segment splitting logic to support per-user max block length and propagate these options through PDF/EPUB adapters and TTS segment generation. Update API, manifest pagination, and segment deduplication for improved performance and correctness. Add tests for new NLP options and manifest logic.
BREAKING CHANGE: TTS segment splitting and manifest APIs now require max block length and preloading parameters; user config schema updated.
Update segment manifest API to group segments by both index and locator
attributes, ensuring distinct segment variants are not merged incorrectly.
Adjust SegmentsSidebar logic to select the most relevant segment per index
based on locator match and recency, improving sidebar accuracy when
multiple locator variants exist for a segment.
Introduce SegmentsSidebar component and supporting reader UI for segment-level
navigation and inspection. Add TTS segments manifest and clear API endpoints to
enable efficient retrieval and management of segment variants. Update TTS
segment schema to include settings_hash for improved lookup and indexing,
and adjust indexes to use settings_hash instead of settings_json. Update
contexts and types to support segment variant display and interaction.
This enhances document navigation and TTS segment management for end users.
Remove legacy TTS segment manifest and audio endpoints in favor of new presigned and fallback audio URL approach. Update segment manifest structure to include `audioPresignUrl` and `audioFallbackUrl` fields, deprecating the previous `audioUrl`. Refactor segment ensure logic and TTS context to utilize the new playback source model. Add server utilities for generating presigned S3 URLs and introduce new API routes for presign and fallback audio retrieval. Clean up obsolete client API functions and types to reflect the new manifest format.
Introduce TTS segment support with new database tables and migrations for both Postgres and SQLite. Implement segment ensure, audio, and manifest API endpoints, segment storage in S3, and segment authorization logic. Update TTS context and client API to use segment-based synthesis and retrieval. Add server-side helpers for segment ID, settings hash, locator normalization, and audio probing. Extend user data cleanup and teardown logic to remove TTS segment objects. Add unit tests for segment helpers.
BREAKING CHANGE: TTS synthesis now uses segment-based storage and APIs; database migrations required.
Move resolveVoices and resolveReplicateVoiceInputKey to a new server-side
voice-resolution module to improve separation of concerns and reduce shared
bundle size. Update all imports and related tests to use the new module.
Remove unused LRUMap and related caches from shared catalog. Update Replicate
cooldown logic to use per-scope LRU cache for improved concurrency handling.
Move getUpstreamStatus and getUpstreamRetryAfterSeconds to a shared utility
module for consistent upstream error handling across TTS endpoints. Strengthen
audiobook chapter API by introducing runtime validation for incoming settings
payloads, ensuring type safety and error reporting for malformed requests.
Replace in-memory Map caches with LRUMap for Replicate voice and schema
lookups, improving memory management and eviction logic.
Align default Replicate model to the versioned Kokoro model across UI, docs,
and server logic. Refactor TTS settings merging for audiobooks to ensure
consistent normalization. Improve Retry-After header handling for upstream
rate limits. Expand Replicate voice resolution to use model schemas when
available, with test coverage for custom and built-in models.
Update documentation to reflect new Replicate defaults, model selection,
and configuration guidance.
- Updated environment variables documentation to include Replicate as a TTS provider option.
- Added Replicate to the sidebar for TTS provider guides.
- Included Replicate as a dependency in package.json and pnpm-lock.yaml.
- Enhanced audiobook chapter generation to normalize native speed settings based on the TTS provider.
- Improved error handling in TTS API routes to provide retry information for rate-limited responses.
- Updated AudiobookExportModal to reflect native speed support for Replicate models.
- Modified SettingsModal to set default model for Replicate.
- Enhanced SpeedControl component to conditionally render native speed controls based on provider support.
- Updated TTSContext to utilize effective native speed for TTS requests.
- Implemented Replicate request handling in the TTS generation logic.
- Added new documentation for configuring Replicate as a TTS provider.
- Move TTS provider resolution and voice-fetch logic into shared catalog and make
remote fetches more resilient (timeouts, abort handling, baseUrl normalization,
treat explicit empty custom-openai responses as valid).
- Add continuation guard to TTS feature set and update API voices endpoint error
logging to better surface auth/session failures.
- Extract and stabilize ebook pipeline and epub adapter:
- Cache prepared EPUB chapters to avoid repeated parsing and handle missing/empty
chapter titles safely.
- Fail early when created chapter lacks a bookId.
- Improve config/preferences handling:
- Replace JSON stringify comparisons with a deepEqual utility to compare defaults.
- Tighten ConfigContext effect dependencies and consolidate synced preference
queuing logic to include additional keys.
- Harden end-to-end test helpers and unit tests:
- Make settings dialog dismissal deterministic (use Enter/Escape and explicit
visibility/enabled checks) and simplify upload retry behavior.
- Add unit test to ensure custom-openai empty voice lists are preserved.
Why: Reduce runtime and test flakiness, centralize TTS logic for easier maintenance,
and make EPUB processing and preference diffing more robust. Breaking changes: none.
- Introduce a shared tts-provider-catalog to centralize provider/model/voice defaults
and feature flags (supportsTtsInstructions, providerSupportsCustomModel, resolveProviderModels,
getDefaultVoices, etc.). Replace ad-hoc provider logic across server and client with catalog calls.
- Replace inline voice/model UI logic in SettingsModal, AudiobookExportModal, TTSContext,
generate server code, and voice hook to use catalog helpers (resolveTtsSettingsViewModel,
supportsTtsInstructions, getDefaultVoices).
- Extract audiobook generation responsibilities into a small client-side pipeline:
- Add createEpubAudiobookSourceAdapter and createPdfAudiobookSourceAdapter adapter modules
to prepare chapters from EPUB/PDF sources.
- Add runAudiobookGeneration and regenerateAudiobookChapter pipeline functions to handle
chapter preparation, progress tracking, request header construction, and retry/abort behavior.
- Wire adapters/pipeline into EPUBContext and PDFContext to simplify and centralize audiobook flow.
- Move config preference helpers out of ConfigContext into dedicated client modules:
- buildSyncedPreferencePatch (lib/client/config/preferences.ts)
- applyConfigUpdate and getVoicePreferenceKey (lib/client/config/updates.ts)
- Use these helpers in ConfigContext to keep update logic small and testable.
- Add TTS settings view model resolver (lib/client/settings/tts-settings.ts) and tests
for the tts-provider catalog and config helpers (tests/unit/tts-provider-catalog.spec.ts).
- Minor usages updated: AudiobookExportModal, SettingsModal, useVoiceManagement, TTSContext,
server tts generation, and various context files to consume the new modules.
Why: Reduce duplication, improve testability, and separate concerns for provider metadata,
UI view-model logic, and audiobook generation flow. Breaking changes: none.
