fix(chat): keep in-flight answers alive across in-app view switches

Root cause (owner repro, verified in a real browser 2026-09-06): the
five navbar views (Chat, RAG, Sources, Tuning, History) were separate
HTML documents, so a navbar click was a REAL cross-document navigation
— the chat page unloaded, the in-flight SSE fetch was aborted, and the
phase-48 teardown (app/api/chat.py `finally`, "chat: turn cancelled")
stopped the model. Observed: send question -> click RAG mid-stream ->
click Chat -> the answer never finished: no `query_log` row, and on
return a dangling question with no brain record (the pre-token pagehide
partial persist skips because `acc` is empty).

Phase-48 LOCKED-DECISION REFINEMENT (owner-confirmed 2026-09-06,
flagged per AGENTS.md rule 3, not silently deviated): "real navigation
cancels the fetch" now means LEAVING THE APP — tab close,
external/other-document navigation, the Stop button. In-app navbar
switches are client-side view switches and no longer cancel.

Fix — Option A (SPA shell), chosen over B (Service Worker owns the
stream) and C (server-side turn registry + resume):
- frontend/index.html is the shell: ONE `<main id="main">` holds the
  five `<section class="view">` blocks; hidden views carry BOTH
  `hidden` and `inert` (WCAG — no focus/keyboard traversal). The
  shared header, the single `doc-modal-*` skeleton, and the
  `#app-version` footer each exist exactly once; the per-view copies
  from the four folded pages are dropped.
- New frontend/assets/router.js (vanilla module — no framework, no
  bundler, No-CDN rule intact): lazy-imports a view module on FIRST
  show only (mount-once, hide-forever — the chat view's in-flight SSE
  reader persists across switches; that persistence IS the fix);
  intercepts same-shell navbar links with preventDefault +
  history.pushState (never a document load); handles popstate; single
  writer of `.nav-link` active state (is-active + aria-current),
  document.title, and the per-view meta description (values carried
  over from the old pages' heads, brand-resolved at write time).
- Each folded page's JS becomes `export async function mount(root)` —
  root-scoped queries; `initSharedHeader()` dropped (the header boots
  once in the shell via the chat module; the admin flag comes from the
  same cached `fetchIsAdmin()` promise — zero extra requests).
- app/main.py: a small list-driven route factory serves the shell for
  /tuning.html, /sources.html, /git-sources.html, /history.html —
  registered AFTER the API routers and BEFORE the static catch-all
  (routes-first). The phase-33 caching middleware applies no-cache +
  `?v=` rewriting unchanged; app/core/caching.py needed NO change
  (the view paths did not change — pinned by the integration tests).
- The four old view .html files are DELETED (one source of truth);
  deep links to the old URLs keep working (the router picks the view
  from the pathname); `/?chat=<id>` is unaffected; the Containerfile
  bundles router.js (inlining the lazy view modules) and drops the
  folded page files.
- app/schemas.py: HistoryTurn.text cap 4000 -> 32000 — the shell
  keeps long saved answers in the chat, and the old cap (stricter than
  the 24_000-char total history budget) 422-rejected any second turn
  in such a chat (found by the phase-42 E2E suite on the shell).

Boundaries: login.html, shared.html, doc-edit.html, document.html
REMAIN separate documents (flow pages, not navbar tabs); a mid-stream
navigation to doc-edit/document.html still cancels per phase 48
(follow-up candidate, out of scope). The SSE API is unchanged. Real
departures still cancel the turn — phase 48 intact (pinned by
tests/e2e/test_stop_generation.py, unchanged, and by the new suite's
real-departure control).

Tests:
- Phase-20 suite REWRITTEN to the new semantics
  (tests/e2e/test_sources_midstream_bug.py): a navbar switch no longer
  cancels — the stream survives the switch and the FULL answer
  settles; the pagehide partial persist REMAINS for real departures
  (the partial's exact shape — first streamed chunk prefix, no done
  metadata — is still pinned there).
- NEW story suite tests/e2e/test_nav_switch_keeps_stream.py (mock
  LLM): the owner repro (send -> RAG mid-stream -> Chat: window
  sentinel survives = same document, FULL answer, exactly one brain
  turn in bor.chat.v1, exactly one settled query_log row, auto-saved
  row matches) + the same mid-stream switch against the other three
  views + the real-departure-still-cancels control + the no-switch
  baseline.
- tests/unit/test_frontend_router.py: source-level pins of the router
  invariants (click interceptor targets ONLY same-shell view paths,
  pushState-only switches, mount-once guard, hidden+inert pair,
  single-writer active state/title); shell-route integration tests
  (each folded path serves the shell with no-cache + `?v=` body; a
  non-view path still 404s); the file-reading unit pins re-pointed at
  the shell (the four view files are gone — the shell is the source
  of truth).

Verification (this commit): full suite green — 1565 unit+integration
tests, app/ coverage 99% (>90% floor); ruff + pyright clean; the
phase's E2E suites green in isolation (house protocol, AGENTS.md rule
9). Owner repro verified in a real browser against the real LLM
(dev server :8010, headful Chromium): "tell me about everquest" ->
RAG mid-stream -> Chat — the answer completed with one brain bubble
and no error banner, `query_log` gained exactly one settled row
(deflected=True: the dev KB holds no EverQuest docs — the settle, not
the topic, is the proof), zero "chat: turn cancelled" lines for that
turn; the control (real navigation to /shared.html mid-stream) still
cancelled (no settled row, the cancel line logged, the partial
persisted on return). Screenshots: .agents/screenshots/76_manual_*.

Phase 76 (76_spa_nav_shell) complete — moved to
.agents/phases/complete/.
This commit is contained in:
2026-09-06 06:31:31 -04:00
parent 7e567bddf3
commit ffa919b8bf
78 changed files with 5548 additions and 3244 deletions
+45 -39
View File
@@ -41,8 +41,7 @@ SOURCES_JS = FRONTEND / "assets" / "sources.js"
DOCUMENT_JS = FRONTEND / "assets" / "document.js"
MARKDOWN_JS = FRONTEND / "assets" / "markdown.js"
MODAL_JS = FRONTEND / "assets" / "document-modal.js" # phase 26: the modal owner
INDEX_HTML = FRONTEND / "index.html"
SOURCES_HTML = FRONTEND / "sources.html"
INDEX_HTML = FRONTEND / "index.html" # the ONE-document shell (phase 76: sources.html folded in)
HAVE_NODE = shutil.which("node") is not None
@@ -294,10 +293,12 @@ def test_markdown_renderer_stays_xss_safe_and_unchanged() -> None:
def test_render_document_exported_and_modal_imports_it() -> None:
"""Phase 26: document.js EXPORTS renderDocument(doc, { … }) — the
exact renderer the standalone page and the modal share (no drift).
The modal module imports it relatively, and BOTH page scripts import
the modal module relatively — no direct <script> tag (the header.js
single-evaluation design: esbuild inlines it into the page bundle,
one module instance per page)."""
The modal module imports it relatively, and BOTH view scripts
(app.js — the chat view at shell boot — and sources.js — the RAG
view module) import the modal module relatively — no direct
<script> tag (the header.js single-evaluation design: esbuild
inlines it into the bundle, one module instance per document).
"""
doc_js = _read(DOCUMENT_JS)
# Task 03 signature: the page passes its #doc-title / #doc-meta /
# #doc-content elements under exactly these names.
@@ -315,12 +316,11 @@ def test_render_document_exported_and_modal_imports_it() -> None:
assert 'from "./document-modal.js"' in js, (
f"{name}: must import the modal module relatively"
)
for page in (INDEX_HTML, SOURCES_HTML):
text = _read(page)
assert not re.search(r"<script[^>]*document-modal\.js", text), (
f"{page.name}: no direct document-modal.js <script> tag "
"(single-evaluation design — the page script imports it)"
)
text = _read(INDEX_HTML)
assert not re.search(r"<script[^>]*document-modal\.js", text), (
"index.html: no direct document-modal.js <script> tag "
"(single-evaluation design — the view scripts import it)"
)
def test_document_js_page_init_is_import_safe() -> None:
@@ -338,37 +338,43 @@ def test_document_js_page_init_is_import_safe() -> None:
)
def test_both_pages_carry_the_modal_skeleton() -> None:
"""Phase 26: chat AND Sources ship the same modal skeleton (the
task-01 markup) — the a11y frame included: role=dialog +
aria-modal, a labelled close control, a focusable content target
(tabindex=-1), and a role=status announcer. Hidden by default —
inert until JS opens it."""
for page in (INDEX_HTML, SOURCES_HTML):
text = _read(page)
assert '<div class="doc-modal" id="doc-modal" hidden>' in text, page.name
assert 'id="doc-modal-backdrop"' in text, page.name
assert 'id="doc-modal-panel"' in text, page.name
assert 'role="dialog"' in text and 'aria-modal="true"' in text, page.name
assert 'id="doc-modal-title"' in text, page.name
assert 'id="doc-modal-meta"' in text, page.name
assert 'id="doc-modal-desc"' in text, page.name
assert 'id="doc-modal-open"' in text, page.name
assert re.search(r'id="doc-modal-content"[^>]*tabindex="-1"', text), page.name
assert re.search(
r'id="doc-modal-close"[^>]*aria-label="Close document"', text
), page.name
def test_shell_carries_the_single_modal_skeleton() -> None:
"""Phase 26 + phase 76 (task 02) dedup pin: the shell ships the
modal skeleton EXACTLY ONCE (the chat's, body level) — the RAG
view's second copy was dropped in the fold; BOTH view scripts
(app.js chat chips, sources.js RAG row links) open documents
through openDocumentModal(...) against that single instance
(document-modal.js resolves it by document-level querySelector at
import). The a11y frame is included: role=dialog + aria-modal, a
labelled close control, a focusable content target (tabindex=-1),
and a role=status announcer. Hidden by default — inert until JS
opens it."""
text = _read(INDEX_HTML)
assert text.count('id="doc-modal"') == 1, (
"the shell must carry exactly ONE modal skeleton (the fold dedup)"
)
assert '<div class="doc-modal" id="doc-modal" hidden>' in text
assert 'id="doc-modal-backdrop"' in text
assert 'id="doc-modal-panel"' in text
assert 'role="dialog"' in text and 'aria-modal="true"' in text
assert 'id="doc-modal-title"' in text
assert 'id="doc-modal-meta"' in text
assert 'id="doc-modal-desc"' in text
assert 'id="doc-modal-open"' in text
assert re.search(r'id="doc-modal-content"[^>]*tabindex="-1"', text)
assert re.search(r'id="doc-modal-close"[^>]*aria-label="Close document"', text)
def test_sources_page_loads_markdown_before_its_module() -> None:
"""Phase 26: the modal renders md documents on the Sources page too —
so sources.html loads the classic markdown.js (global renderMarkdown)
via a relative <script src> BEFORE its module script, exactly like
index.html does."""
html = _read(SOURCES_HTML)
def test_shell_loads_markdown_before_its_modules() -> None:
"""Phase 26 + phase 76 (task 02): the modal renders md documents
in the RAG view too — so the shell loads the classic markdown.js
(global renderMarkdown) via a relative <script src> BEFORE its
module scripts (app.js — the chat view — and the lazy view
modules' modal imports), exactly as the old sources.html did."""
html = _read(INDEX_HTML)
assert re.search(r'<script src="assets/markdown\.js"></script>', html)
assert html.index('src="assets/markdown.js"') < html.index('type="module"'), (
"sources.html: markdown.js must load before the module script"
"index.html: markdown.js must load before the module scripts"
)