|
Add Firefox built-in PDF Viewer support using only public WebExtensions
APIs: a selection context menu ("選択したテキストを翻訳") starts a
translation and the toolbar Action Popup shows the streaming result.
No content-script injection into the privileged PDF Viewer, no PDF.js
internals, no separate windows, no custom viewer (Specifications.md §4.6).
Specification alignment (Phase 1)
- Specifications.md: new requirements PDF-01..04 (PDF selection
translation), CTX-01..03 (context menu translation), POP-01..06
(popup two modes); SEL-06 updated to route PDF through the menus API;
section 13.4 adds a 14-item manual PDF verification checklist; section
15 removes "Firefox内蔵PDF Viewerへの対応" from out-of-scope
- Design.md: architecture diagram with the menus path, menu-translation
component, menus permission, popup translation-result state machine
(idle/starting/streaming/completed/error/cancelled), shared translation
state (§9.2), context-menu integration phase, test plan and API audit
checklist items (menus permission scope, no PDF Viewer internals)
- Security.md: menus permission restricted to selection-context
translation; PDF Viewer treated as privileged browser UI (no injection,
no wrappedJSObject/cloneInto/exportFunction, no internal DOM access);
menus.OnClickData.selectionText treated as untrusted input passing the
same validation pipeline; popup result rendering stays textContent-only
Background (Phase 2)
- translation-service.js (new): the single translation pipeline shared by
every entry point - selection validation (type/trim/empty/control
chars/max length), prompt building, Ollama request, session/abort, and
the shared translation state (requestId, sourceText, translatedText,
status, model, error, startedAt, source). "source" is metadata only; a
PDF-specific translation engine does not exist
- menu-translation.js (new): validates menus.OnClickData.selectionText
and starts a sink-less translation observed via GET_POPUP_STATE
- message-router.js: routes port translations through the shared service;
new runtime messages GET_POPUP_STATE and REGENERATE_LAST; popup-state
port handler broadcasts live state (initial push + every chunk/status
change) so an open popup shows streaming progress in real time
- index.js: registers the selection context menu (contexts: ["selection"]
only, no menus.onShown) and calls browser.action.openPopup()
synchronously inside the click user gesture, catching failures so a
popup-open problem never breaks the translation itself
- manifest.json: permissions now ["storage", "menus"]
- settings-service.js: __resetCacheForTests hook for test isolation
Popup two-mode UI (Phase 3)
- popup.html/js/css: mode nav (設定 / 翻訳). Translation mode shows
status, model, streaming text (textContent only), user-safe error, and
copy (Clipboard API only, no execCommand fallback), regenerate (new
requestId reusing the stored source text) and a settings shortcut.
Closing the popup never cancels the request; reopening shows the
latest text and continues receiving chunks live. Initial mode: active
(starting/streaming) translation wins, otherwise the last tab the user
picked is restored (storage.local.popupMode, default settings)
Fresh-install model auto-selection
- translation-service.js ensureModelConfigured(): when storage.model is
empty, pick the first model from the running Ollama and persist it
before translating, mirroring the popup/options auto-select. A user who
never opened any settings UI can translate immediately; when Ollama is
unreachable a clear OLLAMA_UNAVAILABLE message is shown
User-safe error details
- detailForUser(): strips URLs (no internal host names echoed to the UI),
caps length at 120 chars, drops full JSON response bodies, and hides
meaningless low-level network errors so the popover/popup shows the
base message plus at most a short, safe detail
Tests: 85 passed, 0 failed
- selectionText validation (untrusted input pipeline)
- menu selection -> translation -> popup state streaming/completed
- empty selection rejected without starting a request
- streaming status transitions observed through GET_POPUP_STATE
- Ollama unreachable surfaces a user-safe error into popup state
- newer request replaces older state (abort semantics)
- regenerateLast reuses source text with a fresh requestId
- popup-state port receives live broadcasts; wrong port names ignored
- fresh-install auto-select persists the first Ollama model
- error detail sanitization (URL stripping, length cap, JSON drop)
- popup/options smoke tests updated for the two-mode markup
- setupRouter resets the shared settings cache for test isolation
Docs: README documents the two-mode popup and PDF flow; Testing-Results
records automated results and the manual PDF checklist (to be executed).
No git-side surprises: this commit was created explicitly on request
(requirement 18 of the PDF task had deferred commits until now).
|
|
version)
Implement the complete initial version of a Firefox WebExtension (Manifest V3)
that translates selected text on web pages using a local Ollama LLM and shows
the result in a small popover near the selection.
Specification alignment (Phase 0)
- Unified Specifications.md / Design.md / Security.md on the same approach:
Ollama REST API + Fetch API + NDJSON streaming (no ollama-js SDK dependency)
- Fixed host_permissions to http://localhost/* and http://127.0.0.1/* because
Firefox match patterns do not support port numbers (bug 1362809/1468162);
the actual connection target is restricted to localhost/127.0.0.1 + any port
by application-side validation (defense in depth, OLL-09)
- Clipboard: use navigator.clipboard.writeText() only inside the copy button
click (user activation); no deprecated execCommand fallback
Implementation (zero dependencies, vanilla JS, no npm/Node required)
- manifest.json: MV3, background.scripts + type:module (event page, no
service_worker), action popup, options_ui, content_scripts for http/https,
permissions: storage + localhost host permissions only
- src/shared: constants (ports, limits, presets, error codes), message
validation, settings schema (defaults / merge / sanitize / buildOptions /
toApiThinking), system prompt builder (fixed prompt first + user prompt)
- src/background: message-router (runtime.Port streaming + runtime.onMessage
for model list), translation-session (single active request, abort previous
on new request, stale chunk suppression), ollama-client (fetch against
/api/tags and /api/chat with buffering NDJSON line parser tolerant to chunk
boundaries inside JSON objects, AbortController, host re-validation before
every request, HTTP status to error-code mapping), settings-service
(storage.local load/save with validation and fallback to defaults)
- src/content/content.js: single classic script (Firefox does not support ES
modules in content scripts) with selection detection (selectionchange +
mouseup + keyup, debounced), Shadow DOM action button, Shadow DOM popover
with HIDDEN/LOADING/STREAMING/COMPLETED/ERROR state machine, port client
with requestId-based stale chunk filtering, enabled-state sync via
storage.onChanged; all dynamic text rendered via textContent only
- src/popup + src/options: full settings UI (enable toggle, Ollama host,
model list from Ollama with saved-model warning, target language, thinking
mode, temperature slider 0.0-2.0 step 0.1, context length presets
4k/8k/16k/32k/128k/256k, seed random/fixed with integer value, user system
prompt, advanced numeric options with steppers, stop sequences textarea)
- icons generated by tools/make-icons.ps1 (System.Drawing, no external tools)
- build.ps1 packages manifest.json + icons/ + src/ into dist zip, excluding
tests/, doc/, ollama-js/ and tools/
Tests (71 passed, 0 failed)
- In-browser test runner (tests/test.html) with custom describe/it/assert
framework; no npm, no Node.js
- Unit: constants sync between content mirror and shared module, Ollama host
validation (localhost/127.0.0.1 + any port only; rejects external hosts,
LAN IPs, https, credentials, paths), settings schema defaults and fallback,
temperature 0.1-step validation, context length presets, seed modes,
advanced option types, buildOptions omission of unset values and absence of
tfs_z/mirostat fields, system prompt composition order, message validation
with selection length boundary values
- Integration: NDJSON streaming across awkward byte boundaries, thinking
chunk forwarding, malformed line tolerance, error payload mapping, think/
options inclusion rules, model list, TranslationSession abort semantics,
settings round-trip with storage.onChanged cache refresh, full port flow
(started/chunk/completed), newer request aborts older with stale chunk
discard, malformed message rejection, OLLAMA_UNAVAILABLE and
INVALID_SETTINGS error paths, untrusted sender rejection
Security audit (Design.md section 24 checklist: all pass)
- No innerHTML/outerHTML/document.write/eval/new Function/localStorage
- No remote JS/CSS/fonts/CDN resources; no inline scripts in extension pages
- Ollama access limited to background; content scripts never fetch Ollama
- Background validates message type/length/requestId; single concurrent
translation enforced
- Clipboard API only (no execCommand fallback)
- Minimal permissions: storage + localhost host permissions
Docs
- README.md with build/load/usage instructions, API list, permission rationale
- doc/Testing-Results.md with automated test results, security audit results,
and a 14-item manual Firefox verification checklist (to be executed on a
real Firefox + Ollama environment)
|
