Visiting `/Project-1/archive/PartyA/mdl` (no trailing slash) errored with
`Unrecognized table URL` because tableNameFromUrl only matched
`…/<rowsdir>/table.html`. The cascade declares `default_tool: tables` at
`archive/<party>/mdl`, so the server serves the tables HTML at the bare
directory URL — a shape the client didn't recognize.
Two coordinated fixes:
- shared/zddc-source.js `pathToDir`: was over-eagerly stripping the last
segment when the URL didn't end in `/`. Now checks whether the last
segment contains a dot — file URLs strip to parent (original behavior
preserved), bare-directory URLs append the missing slash. Only call
site is detectServerRoot, so blast radius is contained.
- tables/js/context.js `tableNameFromUrl` + `rowEditUrl`: accept both
legacy `…/<rowsdir>/table.html` and the new bare-directory shape.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds a UI checkbox next to the existing Sort dropdown that surfaces
hidden entries when ACL would otherwise allow read. Default off
(matches today's filtered behavior). On toggle, browse re-fetches
the current directory with ?hidden=1 and re-renders.
┌─ browse toolbar ─────────────────────────────────────────────┐
│ Sort: [Name (A→Z) ▾] ☐ Show hidden │
└──────────────────────────────────────────────────────────────┘
Server-side surface:
- internal/fs/tree.go ListDirectory gains an `includeHidden bool`
parameter. The .-prefix filter (previously hard-coded) now also
drops _-prefix entries (matches dispatch's reserved-prefix guard)
and honors the new flag.
- internal/handler/directory.go reads `?hidden=1` from the request
and threads it through.
- cmd/zddc-server/main.go dispatcher relaxes its dot-prefix and
_-prefix guards for GET/HEAD when `?hidden=1` is set, so clicking
a hidden entry's link works. `_app/` (apps cache) stays
unconditionally reserved — those bytes must go through the apps
resolver. Writes to hidden paths stay blocked (the file API has
its own segment check that the flag does NOT relax).
- internal/listing/listing.go: signature parity (the lower-level
helper that's used by tests + non-cascade listing paths).
Security model unchanged: the ACL chain on the parent dir is the only
real gate. Whoever can read the dir can see its contents — toggling
"Show hidden" just stops the client-side filter from masking
.-prefixed and _-prefixed entries. Hidden paths today:
• <dir>/.zddc ACL YAML — already exposed via /.profile/zddc
• <dir>/.converted/<base> cached MD→DOCX/HTML/PDF, same sensitivity as source
• <root>/.zddc.d/tokens/ per-token metadata; filename = sha256(token)
so not bearer-usable. Default root ACL
restricts to admins; matches /.tokens UI.
• <root>/.zddc.d/logs/ access logs; same admins-only audience
• <root>/_app/ cached upstream tool HTML (public)
• <root>/_template/ install.zip scaffolding (public)
None of these contain bearer credentials or secret material that the
existing ACL doesn't already gate. The walls are still the cascade.
The HTML→PDF path produced PDFs where content extended past the
right margin of each letter page. Two contributing causes in
viewer-template.html's @media print rules:
1. .content-wrapper carries max-width: min(900px, 100%) from the
screen layout. The print override set width: 100% but didn't reset
max-width. Chromium's --print-to-pdf renders at the full page
width (816px for letter at 96dpi) and only clips at print time,
so without max-width: none the element actually extends past the
~624px printable area.
2. Tables, preformatted blocks, and long URLs had no print
containment. A wide <pre> or a <table> with many columns would
blow out the right edge even when the parent constraints held.
Fixes applied to @media print:
- html, body, .app-container: explicit width: 100% + max-width: 100%
to be sure the print viewport flows top-down with no horizontal
creep at the layout root.
- .content-wrapper: max-width: none + width: 100% (was just width).
- .content-page: width: 100% added (was just max-width: none).
- .document-content: max-width: 100% + box-sizing: border-box so
the existing 0.5in horizontal padding stays inside the page.
- pre/code/table/blockquote/img/video: max-width: 100% +
overflow-wrap: break-word; <pre> additionally white-space:
pre-wrap + word-break: break-word so unbreakable token runs
(URLs, paths, command lines) wrap instead of overflowing.
- table: table-layout: fixed so columns shrink to fit rather than
forcing horizontal scroll/overflow.
Both source files (pandoc/viewer-template.html and the embed copy at
zddc/internal/convert/viewer-template.html) updated and verified
identical with diff -q.
The HTML→PDF stage failed with:
Creating shared memory in /dev/shm/.org.chromium.Chromium.XXXXXX
failed: Read-only file system (30)
Unable to access(W_OK|X_OK) /dev/shm: Read-only file system (30)
Chromium tries to put its IPC shared-memory segments under /dev/shm
by default. Our container runs --read-only with /dev/shm inherited
from the image (which makes it read-only too). The well-known fix is
the --disable-dev-shm-usage chromium flag, which routes those
allocations to /tmp instead.
/tmp is a writable tmpfs we already set up. Bump its size from
128 MiB to 256 MiB so chromium has room for both its user-data-dir
and the redirected shared-memory segments. A small PDF flow used
~64 MiB free of 128 MiB available; doubling gives headroom without
materially changing the pod's memory footprint (tmpfs only consumes
RAM for bytes actually written).
The discardable_shared_memory_manager warning ("Less than 64MB of
free space in temporary directory") in the prior chromium log was a
symptom of this same /tmp-too-small condition; the bump quiets it
too.
Other warnings in the log (dbus connect failures) are not load-
bearing — chromium falls back gracefully when dbus is absent. No fix
needed there.
Drops the `download` attribute and adds `target="_blank"`. Click now
opens the conversion URL in a new tab; the server's
Content-Disposition: inline routes each format to a sensible
behavior in that tab:
• HTML → renders as a web page
• PDF → opens in the browser's PDF viewer
• DOCX → auto-downloads (browser can't render Office Open XML),
tab is transient
Right-click → "Save Link As" still works for explicit save-to-disk.
Side benefit for debugging: when the conversion endpoint returns
422 or 503, the response body appears as a plain-text page in the
new tab, which is easier to triage than a transient toast.
Right-click → "Copy Link Address" / "Open in New Tab" / "Save Link
As" now work natively, so users can share the conversion URL or pick
their own download path. The buttons are styled <a href> elements
instead of <button>, with the `download` attribute set to the
expected filename (foo.docx etc.) so a plain click still downloads.
Click handler simplifies a lot: on clean buffer, the handler returns
immediately and the browser handles the navigation. On dirty buffer,
the handler intercepts, auto-saves, then re-fires the click — which
re-enters the handler with dirty=false and falls through to the
native navigation. No more JS fetch + blob + objectURL plumbing for
the common path.
Side effect: if the server returns 422 or 503, the browser shows the
response body in the target tab. That's less polished than the
previous toast, but it's also a more direct view of what the server
actually said. The toast path stays in shared/zddc-source.js's
downloadConverted helper for tools that prefer the JS-driven flow.
zddc-server can now invoke podman as a CLIENT against a remote socket
instead of creating containers in its own process. The sidecar pattern
in tnd-zddc-chart will use this so zddc-server's own pod stays
unprivileged (only the podman-system-service sidecar runs privileged).
New surface:
--convert-podman-socket / ZDDC_CONVERT_PODMAN_SOCKET
e.g. unix:///var/run/podman/podman.sock
Empty (default) → local mode (podman creates containers in
zddc-server's own filesystem namespace).
Non-empty → remote mode: `podman --remote --url=<this> run …`
dispatches each container request to whatever process owns the
socket. Typically a `podman system service` sidecar in the same
Kubernetes pod.
--convert-scratch-dir / ZDDC_CONVERT_SCRATCH_DIR
Host-side directory for per-conversion intermediates (template,
HTML, PDF). In remote mode this MUST be a path the sidecar sees
at the same mountpoint — typically a shared emptyDir at /work
in both containers. Empty = $TMPDIR (local-mode default).
Runner behaviour:
local mode → unchanged. `podman run --userns=host --rm --pull=missing
--network=none --read-only …`. `--userns=host` stays so nested-podman
on a privileged host (the previous chart shape) keeps working for
anyone still using it.
remote mode → `podman --remote --url=<sock> run --rm --pull=missing
--network=none --read-only …`. `--userns=host` is dropped because
the sidecar is rootful inside its own privileged container and
doesn't need userns juggling.
Health probe gains a Mode field ("local" | "remote") and, in remote
mode, runs `podman --remote --url=<sock> version` to confirm the
sidecar's socket is reachable. Unreachable-socket → 503 with a clear
reason (sidecar may still be starting up); reachable → ready.
Capabilities log now includes engine_version + mode + remote_url for
easier debugging of "which podman is actually doing the work".
No tests removed — the existing fake-runner table covers both modes
since the runner's args are uniform (remote prefix is the only thing
that differs).
When zddc-server runs inside a Kubernetes pod and shells out to
`podman run`, the inner podman tries to set up its own user namespace
via /usr/bin/newuidmap. The mapping fails inside the pod's namespace
even with privileged: true:
newuidmap: write to uid_map failed: Invalid argument
Error: cannot set up namespace using "/usr/bin/newuidmap": exit status 1
Adding --userns=host to the inner `podman run` tells it to reuse the
caller's user namespace instead of creating a new one — newuidmap
isn't invoked. The chart already runs the pod privileged so reusing
its userns adds no new privilege; --cap-drop=ALL + --network=none +
--read-only + --tmpfs continue to isolate the inner container.
On a bare-metal host invocation, --userns=host means "no userns
remapping at all", which is the default for rootful podman and works
identically to the prior behavior — the bitnest test setup and any
laptop dev runs are unaffected.
Smoke-tested locally with the exact flag set: pandoc/latex:latest in
a --userns=host --read-only container produces valid HTML from
`# Hello world` on stdin.
The on-page label for `--release alpha|beta` cuts used to read
`vX.Y.Z-channel · YYYY-MM-DD · word-word-word`. The three-word slug
(derived deterministically from the source SHA via shared/build-words.txt)
was meant to make "is this the same build you emailed about?" a glance
check, but the cute-words layer turned out to be more confusing than
clarifying — testers prefer a real timestamp + short SHA.
New label shape, identical to plain dev builds:
vX.Y.Z-channel · YYYY-MM-DD HH:MM:SS · <short-source-sha>[-dirty]
Helper renamed from _source_commit_slug to _source_commit_short_sha,
returning the short SHA of the source commit (walking past any
`chore(embedded): cut …` commit at HEAD so a re-cut on unchanged
source produces the same SHA). The wordlist file is no longer
referenced and is removed; tests/build-label.spec.js's regex
simplified to require the full timestamp + SHA form.
The markdown editor's DOCX/HTML/PDF download buttons silently no-op'd
because the gate `typeof window.zddc.source.downloadConverted ===
'function'` always failed: browse rolls its own server-mode detection
(state.source === 'server' + node.url) and never needed the shared/
zddc-source.js polyfill before. The new download helper lives on
window.zddc.source, so browse needs to bundle it.
Adds ../shared/zddc-source.js to browse/build.sh's concat_files list,
right after preview-lib.js. Bundle gains ~12 KB; tools that don't
need the polyfill (browse doesn't use HttpDirectoryHandle directly)
pay a small footprint cost in exchange for getting the helper.
mdedit/ is gone. Its functionality moved into browse's preview plugin
(browse/js/preview-markdown.js) — YAML front matter editing, outline,
and on-demand DOCX/HTML/PDF download all happen there. Browse is the
default_tool for working/ + reviewing/ as of the previous commit, so
existing URLs of the form /<project>/working land on browse without
operator action.
Removed:
• mdedit/ source tree (Toast UI app, CSS, JS, template, build.sh)
• zddc/internal/apps/embedded/mdedit.html (//go:embed blob)
• tests/mdedit.spec.js + the "mdedit" project in playwright.config.js
• mdedit entries in zddc/internal/apps/embed.go (//go:embed, var,
switch case in EmbeddedBytes)
• "mdedit" in zddc/internal/zddc/validate.go AppNames + the matching
error-message app list
• "mdedit.html" branch in zddc/internal/apps/handler.go MatchAppHTML
• mdedit case in tests (handler_test.go, validate_test.go,
zddchandler_test.go) — test fixtures now use browse/classifier
• mdedit from build (per-tool build.sh loop, tool-list literals,
composer cards) and shared/build-lib.sh ZDDC_RELEASE_TOOLS
• mdedit from freshen-channel's tool list and usage banner
• mdedit-specific paragraphs in AGENTS.md and ARCHITECTURE.md;
Markdown Editor section in ARCHITECTURE.md rewritten to point at
browse/js/preview-markdown.js
• mdedit from CLAUDE.md, README.md, zddc/README.md tool lists
Historical mdedit_v*.html / mdedit_v*.html.sig files in
/srv/zddc/releases/ on the deploy host are immutable history — they
stay where they are. The next ./build release cut will simply not
produce new mdedit_v* artifacts.
Flip default_tool from `mdedit` to `browse` (which now ships a Toast UI
markdown editor plugin in its preview pane) at:
• paths."*".paths.working
• paths."*".paths.working.paths."*" (per-user homes)
• paths."*".paths.reviewing
available_tools at those levels drops `mdedit` and adds `browse` next
to `classifier`. Operator overrides per .zddc cascade still work; only
the embedded baseline changes.
Test fixtures updated:
• lookups_test.go — DefaultToolAt assertions for working/+reviewing/
• availability_test.go — AppAvailableAt + DefaultAppAt for working/+
reviewing/+per-user home
• main_test.go — dispatch route asserts "ZDDC Browse" (was "ZDDC
Markdown"); Apps cascade fixture swaps mdedit
for browse so the live route fetches the right
embedded HTML
New endpoint GET /<path>/foo.md?convert=docx|html|pdf renders a markdown
source on demand. Surfaced as the Download buttons in browse's markdown
editor (separate commit).
Execution model — two upstream container images, lazy-pulled:
• docker.io/pandoc/latex:latest — MD→DOCX, MD→HTML (entrypoint pandoc)
• docker.io/zenika/alpine-chrome — HTML→PDF (entrypoint chromium-browser)
No custom image build. The runner passes --pull=missing on every podman/
docker invocation so the operator only needs the runtime installed —
first request pulls the image, subsequent requests use the local cache.
Overrides: --convert-pandoc-image / --convert-chromium-image (and the
matching ZDDC_CONVERT_* env vars). Engine: --convert-engine (podman
preferred, docker fallback). Resource caps: --convert-mem-mib (512),
--convert-cpus (2), --convert-pids (100), --convert-timeout (30s).
PDF flow is two-stage: pandoc renders the markdown through the embedded
viewer-template.html to standalone HTML, then chromium prints that HTML
via --print-to-pdf. Preserves the print-media CSS already authored in
viewer-template.html rather than going through pandoc's LaTeX template.
Each conversion runs in a throw-away container with --rm --network=none
--read-only --tmpfs=/tmp --cap-drop=ALL --security-opt=no-new-privileges
--env=HOME=/tmp plus a bind-mounted scratch dir for I/O. Pandoc reads
markdown from stdin / writes to stdout; the viewer template lives at
/tpl (ro). Chromium reads HTML from a read-write bind mount at /pdf
and writes the PDF to the same mount; the host reads it back. No shell
wrappers, no shell quoting — argv flows straight into each image's
entrypoint.
On-disk cache at <dir>/.converted/<base>.<ext> with mtime synced to the
source. Fast path is a stat-and-serve with no exec; slow path
singleflights concurrent requests for the same target. PUT/DELETE/MOVE
on the source .md purges the .converted/ sidecars.
Per-project template variables (client/project/contractor/project_number)
come from a new .zddc `convert:` cascade block, walked leaf→root with
per-key latest-wins. Filename-derived variables (title, tracking_number,
revision, status, is_draft) come from a new zddc.ParseFilename helper.
If neither podman nor docker is on PATH, the endpoint serves 503 with
a clear Retry-After. The rest of the server keeps working.
This is the first os/exec site in the codebase. The hardening in
internal/convert/runner.go — context.CancelFunc → process kill,
cmd.WaitDelay, platform-specific SysProcAttr (Setpgid + Pdeathsig on
Linux), minimal env, stdout cap via limitWriter, stderr ring buffer —
sets the pattern for any future shell-outs.
Public surface:
convert.ToDocx(ctx, source, meta) / .ToHTML / .ToPDF
convert.Probe(ctx, engineOverride) → install Runner if engine present
convert.SetImages(pandoc, chromium)
convert.ConfigureLimits(memMiB, cpus, pids, timeout)
convert.Available()
Container handler at internal/handler/converthandler.go; dispatcher
branch in cmd/zddc-server/main.go inserts the convert lookup after the
existing ACL gate, reusing the source file's read policy verbatim.
Two improvements to browse's preview-markdown plugin so it can replace
the standalone mdedit tool:
1. **YAML front-matter editing.** The FM pane above the outline used to
render a read-only <dl> of parsed keys — sparse and unusable when
the file had no envelope yet. It's now a dedicated <textarea> that's
always present. On load, parseFrontMatter() splits the `---\n…\n---`
envelope off the body: the body feeds Toast UI Editor, the envelope
feeds the textarea. On save, assembleContent() recombines them.
Dirty tracking covers both halves via a SHA-256 of the assembled
bytes. The shell mirrors mdedit's old layout (FM textarea top,
outline below) but the FM pane is now always functional, eliminating
the "empty pane over the TOC" problem.
2. **Download as DOCX / HTML / PDF.** When the file handle is HTTP-
backed (server mode) and the file is a .md, three buttons appear in
the info header next to Save. Clicking one fetches the server's
?convert=<fmt> endpoint and triggers a browser download with a
clean filename (foo.md → foo.docx). Auto-saves the buffer first if
dirty so the converted bytes reflect what's on screen.
Helper at window.zddc.source.downloadConverted (shared/zddc-source.js)
so other tools — archive, transmittal — can reuse the same flow later.
Friendly error messages map HTTP 503 / 422 / 504 to actionable toasts.
The .zddc cascade-config migration retired the hardcoded folder-name
predicates (special.go's IsAutoOwnPath/IsWormPath/…) in favour of a
baked-in defaults.zddc.yaml with a recursive paths: tree, but several
top-level docs still described the old model:
- README.md / CLAUDE.md: "tools auto-served at folder-name-driven
paths (classifier in Incoming/Working/Staging, …)" → now: which tool
a URL serves is the cascade's default_tool/dir_tool/available_tools;
added the .zip-as-directory + GET /dir/?zip=1 + show-defaults notes;
CLAUDE.md's shared/ inventory refreshed (zip-source.js, fonts, …).
- ARCHITECTURE.md: the "Cooperating layers" table's "Special folders"
row (referenced special.go, the retired "WORM split") → rewritten as
"Canonical-folder behaviour" driven by the auto_own/worm/virtual/
drop_target .zddc keys; the "ACL cascade" row now mentions the
defaults.zddc.yaml bottom layer + paths: walker.
- zddc/README.md: role resolution was described as "shadows" → it's a
union with reset:true; WORM was "path-based, not cascade-based" → it's
the worm: cascade key; the "Special folders" section rewritten as
"Canonical-folder behaviour via .zddc keys" (a key table + the .zip
/?zip=1 notes), pointing at show-defaults as the authoritative ref.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
`./build alpha|beta` now stamps a date + a three-word slug derived from
the source SHA (e.g. "v0.0.17-beta · 2026-05-12 · candle-mast-pearl")
instead of a raw hex SHA. The build-label spec's channel-label regex
only matched the hex-SHA form (still used by plain dev builds), so it
failed on every release cut. Accept either trailing field.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
A "⤓ Download (zip)" button in the browse toolbar (shown once a
directory is loaded) downloads the directory you're currently
viewing — and everything under it you're allowed to see — as a single
.zip. Navigate into a subfolder first to grab just that subtree.
- Server mode: an <a download> at "<currentPath>?zip=1" — zddc-server
streams the ACL-filtered zip (see the previous commit), nothing held
in the browser.
- Offline (file://) mode: new browse/js/download.js walks the picked
folder with the FS-Access API in two passes — metadata first (so it
can confirm() before loading >~2000 files / ~500 MB into memory),
then bytes — bundles with the already-vendored JSZip, and triggers a
blob download. Hidden entries (".":/"_"-prefixed) are skipped, the
zip's top level is "<folderName>/…" so it unpacks tidily, and the
status bar shows progress.
Wired in browse/js/events.js (button click + show/hide alongside the
refresh button); concatenated into browse/build.sh; ARCHITECTURE.md +
AGENTS.md note the ?zip=1 endpoint and the button.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
zddc-server can now hand back a whole directory subtree as a single
streamed application/zip download: GET /some/dir/?zip=1 (works on both
/dir and /dir/) → Content-Type: application/zip + Content-Disposition:
attachment; filename="<dir>.zip", containing every readable file under
/some/dir/, recursively.
handler.ServeSubtreeZip walks the tree with filepath.WalkDir, ACL-gates
each file by the .zddc chain of its containing directory (per-dir
decision cache, same shape as serveArchiveListing), skips hidden
entries ("." and "_" prefixes — .zddc, _template, _app), and adds a
.zip *file* it encounters as opaque bytes (it does not recurse into it
— that's the navigable-virtual-surface feature, a different thing).
The response is streamed (zip.NewWriter straight onto the
ResponseWriter, Store for already-compressed extensions, Deflate
otherwise), so a fully-ACL-denied or empty subtree just yields a valid
empty zip rather than a 403 (a stream can't change status after the
headers go out; empty leaks no more than 403). HEAD sends the headers
and no body. The dispatch's directory ACL gate still runs first, so a
viewer who can't read the directory gets 403 before the handler.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
- tests/browse.spec.js: expand a .zip in the file tree (offline), drill
into a member subdir, preview a text member — exercises shared/zip-source.js
and the migrated offline path end to end.
- tests/archive.spec.js: a .zip whose name parses as a transmittal folder
is scanned like an uncompressed one — members land in the file list with
tracking numbers parsed, tied to the zip transmittal's folder.
- tests/fixtures/mock-fs-api.js: __setMockDirectoryTree now keeps binary
leaf values (Uint8Array/ArrayBuffer/Blob) intact instead of String()-ing
them — needed to feed real zip bytes through the mock FS.
- tests/data/test-archive.sh: each party gets one transmittal delivered as
a single .zip in received/, so the bitnest fixture exercises the
zip-as-virtual-directory path.
- ARCHITECTURE.md / AGENTS.md: document .zip-as-navigable-directory (server
route + ACL model + shared client adapter + the one-level nesting limit).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
New shared/zip-source.js: a ZipDirectoryHandle / ZipFileHandle pair
that exposes a JSZip instance behind the File-System-Access surface
(values/entries/keys, getDirectoryHandle/getFileHandle, getFile) —
read-only, with a zip-slip guard. Mirrors shared/zddc-source.js's
HTTP polyfill. Wired into archive's and browse's build.sh (both
already bundle JSZip).
archive: a .zip whose name minus ".zip" parses as a transmittal-folder
name is now scanned as that transmittal folder. Offline, the zip is
opened in the browser (ZipDirectoryHandle) and its members enumerated
exactly like an uncompressed folder's files — table/export/hash paths
are unchanged (they go through file.handle.getFile()). Online, the
scanner recurses into the server's "<…>.zip/" virtual-directory
listing, so members come back as "<…>.zip/<member>" URLs the server
extracts on demand — no whole-zip download.
browse: the offline (file://) zip path is migrated onto the shared
adapter — expanding a .zip now opens it as a ZipDirectoryHandle and
its members become ordinary dir/file nodes handled by the normal
fetchFsChildren path (nested zips fall out by recursion). The bespoke
flat-entry walker (loadZipChildren / setZipDirChildren / zipEntries /
zipParentId / zipPath / _zipSyntheticDir) is gone — one zip
implementation repo-wide. Markdown members inside a zip are flagged
read-only (the ZipFileHandle refuses createWritable; server "<…>.zip/"
URLs 405 on PUT).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
When browse runs against zddc-server, expanding a top-level .zip now
fetches "<…>.zip/" as a normal directory listing (the server extracts
members on demand) instead of downloading the entire archive and
parsing it with JSZip in the browser. Members open/preview via their
real server URLs like any file. Nested zips (a .zip inside a .zip) and
FS-API (offline) mode keep the existing JSZip path; offline gets
migrated to a shared adapter in the next change.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
zddc-server can now browse into a .zip file without the client
downloading the whole archive:
- GET …/Foo.zip/ → JSON listing of the zip's members
(Accept: application/json), or the
browse SPA (HTML) — same content
negotiation as ServeDirectory/.archive
- GET …/Foo.zip/sub/doc.pdf → extracts and streams that one member
(Range / ETag / conditional GET via
http.ServeContent)
- GET …/Foo.zip → unchanged: the raw .zip download
- PUT/DELETE/POST …/Foo.zip/… → 405 (zip access is read-only)
New internal/zipfs package reconstructs directory levels from the zip's
flat central directory (synthesising intermediate dirs with no explicit
"<dir>/" entry, mirroring what browse does client-side with JSZip) and
drops zip-slip-unsafe entries ("..", absolute, backslash). New
handler.ServeZip wraps it. The dispatcher gets splitZipPath + an
intercept placed before the file-API branch (so a write to a path under
a .zip is refused, not silently mkdir'd); ACL is the chain of the
directory CONTAINING the zip — a zip carries no .zddc of its own, same
as the .archive virtual surface. The os.Stat-per-segment walk is gated
by a cheap ".zip/" substring check so ordinary requests are unaffected.
Also fixes two pre-existing dispatch-test failures uncovered along the
way: a non-existent top-level "*.html" URL was 302'ing to its slash
form (because the bare "*" project glob makes every first-level segment
"declared") — the cascade-declared no-slash block now requires a
directory-shaped URL (trailing slash, or no file extension); and the
stale TestDispatchSlashRouting expectation that archive/<party>/mdl/
302s to mdl/table.html was updated to match the intended behaviour
(the default-MDL virtual fallback shows the browse listing there; only
a real on-disk tables: + *.table.yaml triggers the bounce).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The trailing-slash directory form was hardcoded to serve `browse`. Add a
`dir_tool` .zddc key (cascades leaf→root, floors at `browse`) so an
operator can point a subtree's slash form at another directory-oriented
tool — the symmetric counterpart to `default_tool` (the no-slash
"specialized app"). handler.ServeDirectory now resolves it via
zddc.DirToolAt; JSON listing requests are unaffected (raw listing
always served, so browse can still enumerate).
Also collapse the no-slash dispatch: the on-disk-directory and the
virtual-declared-path branches in main.go each carried their own copy
of "default_tool → tables-carveout-or-apps.Serve → 302", with
inconsistent ACL checks. Extract one chokepoint, serveSpecializedNoSlash,
that enforces ACL uniformly for every default_tool route.
Updates ARCHITECTURE.md and AGENTS.md: the stale "Special folders" /
hardcoded-availability sections now describe the .zddc-cascade model
(defaults.zddc.yaml, the schema-key table, the slash/no-slash
convention, WORM, standard roles).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The file API's mkdir post-hook still seeded auto-own .zddc files via the
hardcoded IsAutoOwnPath path-segment predicate, while
EnsureCanonicalAncestors had already moved to the cascade's auto_own:
flag. Point the hook at AutoOwnAt / AutoOwnFencedAt so both paths agree
and an operator's .zddc reshaping actually takes effect — fenced when
the new directory's own cascade level declares auto_own_fenced (per-user
working homes), unfenced otherwise.
Retires IsAutoOwnPath and WormMask (the latter already superseded by
WormZoneGrant's & VerbsRC) plus their tests, and the now-unused
path/filepath import in special.go.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Clarify the incoming/ semantics per the workflow: it's the
counterparty's drop zone, not a free-for-all.
- project_team gets read only here (inherited from the project
level — they have no c/w, so they can see what's been dropped
but not touch it). No change in effect; documented explicitly.
- document_controller gets rwcd here (restated at the incoming/
cascade level). The QC + transfer workflow — classifier renames
files in place (w), then they move to received/ (delete here +
worm-create there) — needs the delete bit, which the inherited
project-level `rw` lacked.
- The counterparty's uploader still gets access via a deployment
.zddc (acl: { permissions: { "*@acme.com": cr } } at
archive/Acme/incoming/.zddc) or by mkdir'ing a dated subfolder
under incoming/ and owning it via the existing auto_own — both
flows unchanged.
Test: standardroles_test now asserts the doc controller has rwcd at
incoming/ and a project_team member has only r there.
All Go + Playwright tests green.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Answers "can roles reset as well as add?" — yes, both now.
Role membership UNIONS across the cascade:
- A deeper .zddc that defines an inherited role again with one
extra member ADDS that member (was: deepest definition shadowed
the ancestor's entirely).
- New `reset: true` on a role definition breaks the union — that
level's members are authoritative, ancestor definitions above
are excluded; descendants below still union on top. Use it to
give a project its own team independent of a deployment-wide
default.
- lookupRoleMembers / RoleMembers reworked: walk deep→shallow,
union members, stop at the first reset:true; finally fold in
chain.Embedded.Roles as the baseline so a role declared only in
defaults.zddc.yaml is "defined" (and a deployment's on-disk
redefinition unions on top).
Admin checks are now role-aware:
- IsSubtreeAdmin / CanEditZddc's strict-ancestor scan use
MatchesPrincipal instead of MatchesPattern, so `admins:
[document_controller]` resolves to the role's members. The
strict-ancestor scan resolves roles only up to level i, so a
role defined at the deepest level (= dirPath) never confers
self-edit rights.
Two standard roles ship in defaults.zddc.yaml (empty members — a
fresh deployment grants nothing until they're populated):
document_controller — files into the WORM zones. Gets:
- rw at the project level (read + overwrite-existing; NOT c, so
it can't make arbitrary folders)
- rwc at archive/ (can create party subfolders)
- subtree-admin at working/ and staging/ (full create + manage,
including taking over a fenced per-user home) — scoped HERE,
not at the project root, so the WORM constraint still binds
it in archive/<party>/received|issued
- listed in worm: on received/ and issued/ → write-once-create
survives the WORM mask
project_team — read-only across the project. The per-user
working home's fenced auto-own .zddc (rwcda for the creator)
wins via deepest-match, so "read-only except what I own" falls
out of the cascade with no special rule. Inside received/issued
their r is preserved (worm: doesn't strip read).
archive/<party>/ gains `auto_own: true` (UNFENCED) so whoever
creates a party subtree (normally the doc controller) owns it and
can set up that counterparty's .zddc afterward — without fencing,
project_team:r still cascades through to received/issued.
Tests: roles_test (union + reset), standardroles_test (the
doc-controller scoped-create matrix + project-team read-only-except-
owned), ensure_test updated for the new party-folder auto-own.
fileapi_test's WORM doc-controller test already uses worm: [role].
All Go + 248 Playwright tests green.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Per design feedback: the verb string in a worm: entry was always
effectively "cr" (the key's whole job is to restore write-once-create
inside the locked zone, and you need read to see what you filed), so
spelling it out per-entry was redundant. worm: is now just a list of
principal patterns — email-globs, @role:name, or bare role names —
and every listed principal gets read + write-once-create. An empty
list ([]) still marks the WORM zone with no create-capable
principals.
Changes:
- ZddcFile.Worm: map[string]string → []string
- mergeOverlay: concat-dedupe (a deeper .zddc adds controllers);
mergeStringSlicePreserveEmpty keeps `worm: []` non-nil through
the overlay so it still marks the zone
- WormZoneGrant: walks the list, grants VerbsRC to each matching
principal; result is always ⊆ {r, c}
- ValidateFile: validates each entry as an email-glob (role refs
skipped — validated by the role machinery)
- defaults.zddc.yaml: received/ and issued/ carry `worm: []`
- tests updated to the list form (worm_test.go, fileapi_test.go)
All Go tests green.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
WORM (write-once-read-many) is no longer a special folder type keyed
off the literal names "received"/"issued". It's a cascade key —
`worm:` on any directory's .zddc — with the ACL-shaped semantics the
user described.
Schema:
worm:
"doc-control@example.com": cr # email-glob or @role:name → verbs ⊆ {r, c}
# an empty map ({}) is a WORM zone with no create-capable principals
Effect inside a WORM zone (any cascade level declares worm:), applied
AFTER the normal cascade ACL and BEFORE the admin escape hatch:
- w / d / a stripped for everyone
- c survives only via the worm: map
- r survives via the normal ACL OR the worm: map (so a document
controller who isn't in the project ACL still gets read+create)
- worm: grants UNION across the cascade — deeper .zddc can name
more controllers
- admins (root / subtree) bypass entirely — handler does the
IsAdmin check before the policy evaluator
defaults.zddc.yaml: archive/<party>/received and archive/<party>/issued
carry `worm: {}` (WORM zone, no controllers — the deployment names
its document controller by adding a deeper .zddc with
`worm: {<principal>: cr}`). The canonical convention is unchanged;
the difference is an operator can now mark any directory WORM, or
rename received/issued, without a code change.
Removed (hardcoded path predicates, superseded by the cascade walk):
zddc.IsWormPath
zddc.WormFolderLevelIndex
zddc.splitPathSegments (only IsWormPath used it)
Kept: zddc.WormMask (generic verb-set primitive), zddc.VerbsRC.
New:
zddc.WormZoneGrant(chain, email, mode) → (verbs, inWormZone)
Walks the chain for worm: declarations; unions the principal's
grants masked to {r, c}.
policy.InternalDecider.Allow: WORM block rewritten to consult
WormZoneGrant instead of IsWormPath/WormFolderLevelIndex.
ValidateFile: worm: keys validated as email-glob (or @role:name);
values validated as verb strings ⊆ {r, c}.
Tests:
- new worm_test.go covers the embedded convention, operator-granted
controller, w/d masking, cross-cascade union.
- special_test.go's TestIsWormPath / TestWormFolderLevelIndex
retired; TestWormMaskStripsWDA kept.
- fileapi_test.go's WORM tests updated: the doc-controller grant is
now `worm: { _doc_controller: cr }` at issued/.zddc, not
`acl.permissions: { _doc_controller: cr }`.
- federal-parity and admin-bypass tests unchanged — the WORM mask
still strips w/d/a and admins still bypass.
All Go tests green.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The shared/nav.js stage strip previously hardcoded four stages
(archive/working/staging/reviewing) with their labels and target
URLs baked into the file. Operators couldn't add a fifth stage or
rename "Working" to "In-Progress" without forking shared code.
Now cascade-driven end-to-end:
Server-side:
listing.FileInfo gains a Declared bool field. fs.ListDirectory
stamps Declared=true on every entry whose name matches the
cascade's ChildrenDeclaredAt(parent) — both real on-disk dirs
and virtual canonical injections. Bugfix in the same patch:
virtualCanonicalFolders was passing the relative dirPath to
ChildrenDeclaredAt (which expects absolute); now passes absDir.
Client-side:
shared/nav.js fetches the project root's JSON listing on
DOMContentLoaded, filters to declared+is_dir entries, sorts by
canonical workflow order (archive → working → staging →
reviewing, then any extras alphabetically), and renders the
strip. Labels read e.display_name → falls back to titleCase(name).
Hardcoded FALLBACK_STAGES kicks in only on fetch failure
(offline / file:// / non-zddc-server backend). Rendered
immediately so the strip appears without flicker, then the
cascade-fetched list replaces it once available.
Effect:
Project-3 (which has display: { archive: "Records",
working: "In-Progress", ... } in its .zddc) now shows
"Records · In-Progress · Outbox · Pending Responses" in every
tool's strip. Project-1 still shows "Archive · Working ·
Staging · Reviewing". No code change to render either; the
cascade decides.
Tests:
- tests/nav.spec.js relies on the mock server returning HTML at
every URL, so the fetch fails over to fallback stages — the
test renders the same Archive/Working/Staging/Reviewing labels
it always did, with no test changes needed.
- All 248 Playwright + all Go tests green.
Remaining client-side hardcode: archive/js/source.js +
archive/js/app.js's mode detection. Phase 4d.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The /incoming/ path regex in browse/js/grid.js was the second-most
visible client-side hardcode of the canonical convention. Migrating
it to the cascade:
Header surface:
X-ZDDC-Default-Tool: <name> The cascade-resolved default tool
for the listing's directory. Empty
header = no default declared.
Client wiring:
loader.fetchServerChildren reads the header into
state.scopeDefaultTool on every listing fetch (initial mount,
rescope on dblclick, popstate). grid.classifierAvailableHere
now returns scopeDefaultTool === 'classifier' instead of
regex-matching the URL.
Effect:
Grid mode auto-activates wherever the cascade picks classifier
as the default — currently archive/<party>/incoming per
defaults.zddc.yaml. An operator who sets default_tool: classifier
on a custom directory gets grid mode there too, no code change.
An operator who removes the default at incoming sees grid mode
stop auto-activating there.
Bootstrap timing fix:
The initial events.init() runs applyResolvedViewMode before the
detection fetch completes, so state.scopeDefaultTool is empty
at that point and grid never auto-activates on first paint.
app.js bootstrap now re-applies the resolved view mode after
autoDetectServerMode returns, so a fresh /incoming URL lands
on grid mode immediately.
The /incoming/ regex is gone. Two client hardcodes remaining
(archive source heuristics, shared/nav stage strip) — Phase 4c/d.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
The last hardcoded client-side knowledge of the canonical convention
was the upload-zone regex in browse:
var UPLOAD_SCOPES = /\/(working|staging|incoming)(\/|$)/i;
Now declared in the cascade:
Schema:
drop_target: true|false leaf-only; describes THIS dir
(not propagated to descendants)
Lookup:
zddc.DropTargetAt(root, dir) bool
Surfaced to clients:
Directory listings carry an X-ZDDC-Drop-Target: true response
header when the cascade declares this leaf as an upload zone.
No header = no drop target.
Defaults populated:
working / working/* / staging / archive/<party>/incoming
all carry drop_target: true. Operators can extend (e.g. drop
files on archive/<party>/received via override) or disable
(e.g. drop_target: false at a specific staging subtree) without
touching code.
Browse migration:
loader.fetchServerChildren reads the response header and stamps
state.scopeDropTarget on every listing fetch. upload.js's
currentScopeAllows now reads that flag instead of regex-
matching the URL. Initial value is false in init.js so a
listing failure (offline / server doesn't emit the header)
safely defaults to "no drop zone".
Phase 4a closes the most visible asymmetry between server-side and
client-side cascade knowledge. The remaining client hardcodes
(browse grid-mode regex, archive source heuristics, shared/nav
stage strip) follow the same pattern when needed — Phase 4b/c/d.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Phase 3 retired these symbols by migrating their consumers to the
.zddc cascade lookups. Removing them now that nothing references
them:
- var zddc.ProjectRootFolders
- var zddc.PartyFolders
- var zddc.AutoOwnCanonicalNames
- var zddc.VirtualOnlyCanonicalNames
- func zddc.IsProjectRootFolder
- func zddc.IsArchivePartyFolder
- func zddc.IsArchivePartyMdlDir
- func handler.isArchivePartyDir
The canonical convention is expressed in defaults.zddc.yaml and
consulted via lookups.go's DefaultToolAt / AutoOwnAt / VirtualAt /
IsDeclaredPath / ChildrenDeclaredAt / AvailableToolsAt /
IsToolAvailableAt. Operators override per-directory via on-disk
.zddc files; the embedded layer is the documented baseline.
Test removals:
- TestCanonicalLists (lists no longer exist)
- TestIsProjectRootFolder (function no longer exists)
Equivalent coverage lives in lookups_test.go's
TestDefaultToolAt_FromEmbeddedConvention,
TestIsDeclaredPath_FromEmbeddedConvention, etc. — which assert the
convention via the cascade's actual lookup path rather than the
predicates' return values.
handler.isAtArchivePartyMdlDir is RETAINED — it's still actively
consumed by RecognizeTableRequest's default-MDL fallback in
table.html URL resolution. That's a tighter file-path predicate
than the cascade walker would naturally express; can revisit if it
ever needs to become configurable.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Final consumer migration. The Go-coded lists that previously encoded
the ZDDC convention all defer to the .zddc cascade now.
Schema added:
available_tools: [tool1, tool2, ...] concat-union across cascade;
tools not in the union are
denied auto-route at that path
auto_own_fenced: true|false generated auto-own .zddc
carries inherit:false (private
to creator)
Lookups added:
AvailableToolsAt(root, dir) union of available_tools across cascade
IsToolAvailableAt(root, dir, tool)
AutoOwnFencedAt(root, dir) leaf-only
Cascade semantics finalised (per field):
default_tool → leaf→root walk (parent applies to descendants)
available_tools → leaf→root union (each level adds; baseline at root)
auto_own → leaf-only (creating THIS dir specifically)
auto_own_fenced → leaf-only (same)
virtual → leaf-only (THIS dir is virtual, not subtree)
Consumers migrated:
apps.DefaultAppAt → zddc.DefaultToolAt
apps.AppAvailableAt → zddc.IsToolAvailableAt (+ landing special)
EnsureCanonicalAncestors → AutoOwnAt + AutoOwnFencedAt
fs.ListDirectory empty-list fallback → zddc.IsDeclaredPath
fs.virtualCanonicalFolders → zddc.ChildrenDeclaredAt
dispatcher canonical-folder branches → unified into one
cascade-declared block
Hardcoded helpers REMOVED (dead code):
apps.inAncestorWithName
zddc.autoOwnDepthMatch / isAutoOwnDepthMatch
Hardcoded lists kept as data sources for the cascade walker but
no longer drive routing logic:
ProjectRootFolders / PartyFolders / AutoOwnCanonicalNames /
VirtualOnlyCanonicalNames / IsProjectRootFolder / IsArchivePartyFolder /
IsArchivePartyMdlDir — all still defined; only `ProjectRootFolders`
is used by special.go's IsProjectRootFolder. The rest are dead.
Dispatcher unified: the previously-two branches (per-party folder vs
project-root folder) collapse into one cascade-declared-path block
that handles the slash/no-slash convention uniformly:
- no-slash, default_tool=tables → ServeTable (default-MDL fallback)
- no-slash, default_tool set → apps.Serve(tool)
- no-slash, no default_tool → 302 to slash form
- slash, any → ServeDirectory empty-list fallback
The IsDir branch's switch also un-hardcoded — any cascade tool is
served (not just the legacy 3 names), so e.g. /Project/archive/<party>
/incoming (no slash) now serves classifier directly rather than 302'ing
to the slash form.
defaults.zddc.yaml populated with the canonical convention as the
recipe. Operators edit it (or override per-directory on disk) to
change any behaviour — no Go code changes required.
Browse drag-drop scope (working/staging/incoming) is the one remaining
client-side hardcoded regex; cascading that requires the cascade JSON
to be served to the client, which is its own Phase 4 piece.
Tests updated for the new no-slash mdl URL convention (landing MDL
card test) and no-slash stage URLs (nav strip test). All 248
Playwright + all Go tests green.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Two pieces:
1. Lookup helpers walk chain.Levels from leaf back to root. The
"parent applies to descendants unless overridden" cascade rule
means a working/ default_tool=mdedit propagates to deep paths
like working/alice@example.com/notes/sub/deep without anyone
declaring it at every level. AutoOwnAt and VirtualAt follow the
same walk; explicit false at a descendant can override an
ancestor's true (*bool semantics).
2. apps.DefaultAppAt delegates to zddc.DefaultToolAt. The hardcoded
switch on parts[1] (archive→archive, staging→transmittal,
working→mdedit, reviewing→mdedit, mdl→tables) and its case-
sensitivity quirks now live in defaults.zddc.yaml. Operators can
override any of these per-directory with an on-disk .zddc; no
code change required.
Semantic improvement: archive/<party>/incoming previously defaulted
to "archive" (because parts[1]=archive and the switch didn't look
deeper). The new convention routes it to "classifier" — incoming/ is
the bulk-rename surface, not a record browser. Updated
availability_test.go to reflect.
All other DefaultAppAt cases — including case-fold (Archive/MDL),
mdl override, reviewing virtual, project root returning "", random
non-canonical names returning "" — produce bit-identical output.
Two new tests in lookups_test.go cover the propagation:
- TestDefaultToolAt_PropagatesToDescendants
- TestAutoOwnAt_DescendantCanDisable
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Schema:
- default_tool: string (tool name served at this dir's no-slash URL)
- auto_own: *bool (mkdir post-hook auto-grants the creator)
- virtual: *bool (never materialise on disk; aggregator routes)
defaults.zddc.yaml: populated with the full canonical convention via
paths:. Top-level "*" matches any project; nested archive/working/
staging/reviewing declare the project-stage tools; archive's "*" /
mdl|incoming|received|issued tree declares the per-party surfaces.
All four party folders and all four project-root folders get their
default_tool; working / staging / archive/<party>/incoming get
auto_own; reviewing / archive/<party>/mdl get virtual. None of these
need on-disk dirs to exist.
Lookups (zddc/internal/zddc/lookups.go):
DefaultToolAt(root, dir) → cascade-resolved default tool name
AutoOwnAt(root, dir) → does mkdir auto-own here?
VirtualAt(root, dir) → never materialise on disk?
IsDeclaredPath(root, dir) → does the cascade say anything about this dir?
ChildrenDeclaredAt(root, dir)→ literal child names declared by Paths
Each looks up via EffectivePolicy → leaf level → Embedded fallback,
so operators' on-disk overrides win and the embedded baseline carries
the convention.
Tests cover the embedded convention, operator overrides, and
inherit:false blocking the embedded layer. No consumer migration yet
— that's Phase 3b. Behaviour is bit-identical for current callers
since none of them consult the new lookups.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Adds the recursive paths: schema and the cascade walker that threads
ancestor virtual contributions through to descendant levels.
Schema:
paths:
"*": # literal-segment or "*" segment-wildcard key
paths: # recursive — each step matches one segment
archive:
paths:
"*":
paths:
incoming:
title: "demo"
Each on-disk .zddc and the embedded defaults can declare paths:; the
walker collects every matching subtree and merges its contributions
into chain.Levels[depth] using mergeOverlay (per-field overlay with
on-disk most specific). The matched glob descends one segment at a
time; the value's own paths: becomes a new virtual source for deeper
matches.
Semantics:
- matchGlob: literal key first (case-insensitive on segment),
"*" wildcard fallback.
- mergeOverlay: top wins per-field on scalars; maps merge key-by-
key with top overriding; lists concat-dedupe; Paths replaces
(recursive walker threads it through naturally).
- inherit:false at any on-disk level drops accumulated ancestor
virtual sources AND zeroes chain.Embedded — the operator owns
every rule from that level outward.
- Behaviour is bit-identical when no .zddc declares paths:; the
walker reduces to the prior linear cascade.
Eight new tests cover the glob match table, ancestor-paths
contribution, on-disk-wins override, paths-absent bit-identical
behaviour, and inherit:false dropping ancestor paths: contributions.
All existing tests still pass.
Phase 3 next: populate defaults.zddc.yaml with the canonical
ZDDC convention via paths:, and replace apps.DefaultAppAt /
AppAvailableAt / AutoOwnCanonicalNames / VirtualOnlyCanonicalNames /
IsProjectRootFolder / IsArchivePartyFolder with cascade lookups.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
First step of the .zddc-first-configuration rollout: pure plumbing
that makes the future move-everything-out-of-Go work mechanically
possible without changing any current behaviour.
New pieces:
1. zddc/internal/zddc/defaults.zddc.yaml — a real YAML file in the
repo. Single source of truth for the baked-in baseline; intentionally
minimal in Phase 1 (just title + empty acl) so existing deployments
stay bit-identical until Phase 2 starts populating the schema.
2. //go:embed (defaults.go) bakes the bytes into the binary so
shipped deployments don't need the file. Operators who want a
starting point export with:
zddc-server show-defaults > /var/lib/zddc/root/.zddc
3. PolicyChain gains an Embedded ZddcFile field. EffectivePolicy
layers in the embedded defaults as a baseline below the on-disk
chain. Consumers that want the full effective view consult both;
existing consumers that only read chain.Levels keep working
bit-identically (the new field is additive).
4. New top-level `inherit:` key on ZddcFile. Default true. Set
`inherit: false` on any on-disk .zddc to zero out chain.Embedded
— the operator owns every rule from that level outward. Useful at
the on-disk root to fully reject the embedded defaults; useful at
deeper levels for sandbox subtrees.
5. `zddc-server show-defaults` (also accepts --show-defaults) subcommand
dumps the embedded bytes to stdout — same shape as --print-rego.
No flag plumbing needed beyond the existing args walk.
6. Tests: parse-roundtrip on the embedded file, presence in chain by
default, inherit:false drops it, explicit inherit:true is a no-op
versus the default.
Phase 2 (next): add a `paths:` recursive map + `default_tool:` /
`auto_own:` / `virtual:` keys, populate defaults.zddc.yaml with the
canonical ZDDC convention, and migrate apps.DefaultAppAt /
AutoOwnCanonicalNames / VirtualOnlyCanonicalNames to cascade lookups.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Drop files anywhere on a browse page; if the current scope is inside
a working/, staging/, or incoming/ subtree the files are PUT to the
current directory via the existing file API. Per-file ACL is enforced
server-side (authorizeAction); a 403 surfaces as a per-file error
toast and the rest of the batch proceeds.
UX:
- dragenter → semi-transparent overlay with a dashed-border panel
showing the destination path. Hides immediately on dragleave or
drop.
- drop → "Uploading N files…" toast, then per-file failure toasts
inline, then a summary toast (success / partial / all-failed).
- listing auto-refreshes after the batch so new files appear in
the tree without a manual reload.
Scope:
- upload-eligible paths are matched by /\/(working|staging|incoming)
(\/|$)/i — same convention as the new grid-mode URL token.
- 256 MiB per-file cap (UPLOAD_MAX_BYTES) since browse's single-
body PUT loads the file as a Blob in the tab; larger uploads
should use a dedicated client.
- Outside the upload-eligible set the overlay never appears; drops
are silently ignored (drag effect = none).
Sequential uploads keep progress predictable; parallel batching can
land later if needed. The module hooks document-level dragenter/leave
/over/drop so it works regardless of which pane the user drags over.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Three coupled fixes:
1. landing MDL card: Open button now navigates to /<project>/archive/
<party>/mdl (no trailing slash) so the tables tool loads. The
slash form would route to browse instead, which is not what users
want when they click "Open MDL".
2. zddc-server canonical-folder fallback extended to
archive/<party>/{mdl,incoming,received,issued}. New
zddc.IsArchivePartyFolder() recognises any of the four party
folders at depth 4. fs.ListDirectory returns [] for missing
on-disk variants (mirroring the project-root behavior added in
commit 3fc3717); the dispatcher routes slash forms to
ServeDirectory and the no-slash mdl form to ServeTable, with
non-mdl no-slash forms 302'ing to the slash form.
So /Project-N/archive/<party>/incoming/ now lands on an empty
browse listing rather than 404 when nobody has dropped files yet.
3. Fixture seeded with 3 files per party under incoming/ — naming
intentionally NOT in transmittal-envelope form, so classifier
(loaded automatically by browse's grid mode at /incoming/
per the URL-driven view convention) has something to rename.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
User feedback: the Grid toggle button was on every page and showed an
explanatory empty state when classifier wasn't available — adding UI
to explain why UI didn't work. Cleaner approach: drop the button,
make the URL the source of truth, and default grid mode automatically
inside the only context where it's meaningful.
Behavior:
- Inside any incoming/ path (case-insensitive segment match):
→ grid mode by default
- Everywhere else:
→ browse mode
- Explicit overrides via query string:
?view=grid forces grid (only honored where classifier is
available; otherwise falls back to browse)
?view=browse forces browse (always)
UI changes:
- The Browse/Grid pill toggle is gone.
- grid.js drops both empty-state messages; outside an incoming/
path it just does nothing.
- events.js owns resolveViewMode() / applyResolvedViewMode(),
called on initial mount and after every client-side rescope
(dblclick + popstate).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
