Adds helper functions to retrieve historical chat memory context (`_chat_memory_context`) and merge enriched messages. Updates both the main chat endpoint and regeneration logic to utilize this context, ensuring conversation continuity and reliable source deduplication.
Introduce optional web search enrichment flow for chat and regenerate requests. New /websearch endpoint calls enrich_prompt via SearXNG and returns enriched_prompt + citation sources.
DB: add sources_json column to chat_messages via ensure_sources_column migration helper.
Backend: persist sources_json for assistant replies (streaming and non-streaming); extend ChatRequest/RegenerateRequest to accept enriched_message and sources; history endpoint returns sources.
Frontend: add toggle for web search, settings for SearXNG URL + engines, and optional enrichment calls in sendMessage/regenerate. Render citation sources as rounded chips labeled with base domain under assistant replies.
Dependencies: add beautifulsoup4, httpx[http2], numpy for enrichment pipeline.
Add PUT /sessions/{session_id}/messages/{index} to edit a user message and prune all subsequent messages.
Add POST /sessions/{session_id}/regenerate to regenerate an assistant response from a given index (supports optional streaming).
Introduce EditMessageRequest and RegenerateRequest schemas.
Update the frontend to: • allow in‑place editing of user messages with an auto‑focusing textarea; • copy messages to clipboard; • regenerate assistant responses via a button.
Add new CSS rules for message options bars and the editing textarea.
Backend: HTML-unescape and strip <think>/<thinking> from generated titles; trim and return cleaned value; add debug logs.
Electron: send 'window-focused' from main; expose onWindowFocus in preload.
Frontend: stream-safe AssistantMessageContent with collapsible Thoughts; switch streaming render to React state updates; autofocus textarea on window focus/new chat and when clicking empty chat area; sanitize session title client-side.
Markdown: support blockquotes; allow '-' or '*' bullets; simplify <think> removal to handle streaming; drop table wrapper div (emit <table class='nice'>); theme-aware code block headers/borders.
CSS: rounded 'nice' tables with light inner grid; blockquote styling; Thoughts toggle/panel styles.
Color: brighten Grayscale --accent.
Follow-ups: add IPC listener cleanup; ensure single source of truth for colorSchemes.
Backend
- /chat: support streaming via StreamingResponse; save full reply after stream ends. Non-stream path unchanged.
- ChatRequest: add stream flag (default false).
- GenerateTitleRequest: add model and use it instead of hardcoded llama3.
- ollama_client.chat_stream(): new async generator parsing Ollama streaming JSON (both formats).
- Remove response_model from /chat to allow streaming; non-stream still returns { reply }.
Electron
- Open external links in system browser (setWindowOpenHandler, shell.openExternal).
- New IPC: update-settings, open-external-link.
- Set minimum window size; preload exposes updateSettings and openExternalLink.
Frontend (React)
- Streaming UI with live chunking; sticky-bottom only when user at bottom.
- Per-session scroll persistence and robust restore.
- New message tip to jump to latest reply when scrolled up.
- Disable Send while sending; spinner.
- General Settings: stream output toggle; propagate model/stream changes.
- Apply color scheme at boot; extract colorSchemes helper.
- Sidebar UX tweaks and unread badges.
Markdown/rendering
- Code blocks: language title bar and wrapper.
- Tables: GitHub-style parsing, per-cell borders, rounded wrapper, spacing, alignment.
- Headings: remove blank line after h1-h4.
- <hr>: handle after tables; strip following whitespace.
- Links: target=_blank with icon and URL tooltip.
Styles
- Add styles for code/table wrappers, new-message tip, toggle, spinner; hover/active vars; narrower sidebar.
API notes / breaking changes
- /chat accepts stream=true and returns text/plain streamed chunks.
- generate-title now requires a model.
- Non-stream /chat response shape unchanged.
