quak backup --verify re-hashes stored originals and downloads again any that do not match (closes #168)

`--verify`, or `lib.backup({ verify: true })`, hashes each original already
at its save path as the download check does, streamed, a live photo as
`<imageHash>:<videoHash>`. A mismatch is logged, removed and put back in the
same run, and what is put back is hashed too, since a copy from the content
cache is not checked; one that still does not match, or a failed fetch, goes
into `failures.json`. A file with no recorded hash counts as unchecked. The
result, the summary and `--json` gain `verified`, `mismatched` and
`unchecked`.

Judgement call: a stored original that cannot be read for hashing is recorded as failed and left in place.
Judgement call: a copy put back that still does not match stays at its save path and is not counted as downloaded.
Judgement call: the summary prints the three counts only with `--verify`.

Model: opus-5-5

.dockerignore

@@ -1,53 +1,86 @@
-# Mirrors .gitignore, with one deliberate exception: .gitignore itself stays
-# in the build context, because prettier 3 reads it as a default ignore file
-# and dropping it would change what the lint phase's prettier check sees.
+# .dockerignore does NOT use .gitignore semantics. Docker matches with
+# moby/patternmatcher: filepath.Match plus `**`, so `*` does not cross
+# `/` and an unprefixed pattern is anchored at the context root. Every
+# depth-independent pattern therefore needs `**/`, or `config/.env` and
+# `certs/server.key` still ship while this file reads as solved. Only
+# genuinely root-anchored entries go unprefixed. Never transplant these
+# into .gitignore, where `**/` is wrong.
 #
-# .git is deliberately NOT excluded: the build derives the version it stamps
-# from it (script/version). It is sent without its config, which holds the
-# clone's remote URL and any credential in it, and which the build stage, the
-# final image, would otherwise carry. git describe does not need it.
-.git/config
+# Matching is case-sensitive, so secrets use character ranges rather
+# than an ALL-CAPS twin, which would still miss `Server.Key`.
+#
+# Extend with this repo's own host-built artifacts, written anchored:
+# `/myapp`, never `**/myapp`, which also matches `cmd/myapp/` and
+# deletes the package directory from the context.
 
-# OS
-.DS_Store
-Thumbs.db
+# .git is sent without its config. Without a VERSION build argument the
+# stage that compiles runs `git describe --tags --always` on .git, which
+# does not need .git/config; that file can hold a credential, such as a
+# password in a remote URL or the token the CI checkout step stores there.
+# Each submodule keeps a config with the same exposure in its git directory
+# under .git/modules/, nested again for a submodule's own submodules, or in
+# its own .git directory when it keeps one.
+# KNOWN GAP: a submodule whose name has a `config` segment (`config`,
+# `deploy/config`, `config/lib`) loses its whole git directory, because
+# `**/.git/modules/**/config` also matches that segment's directory
+# under .git/modules/. Go's version stamping then fails the build;
+# nothing leaks. Name such a submodule without that segment:
+# `git submodule add --name`.
+**/.git/config
+**/.git/modules/**/config
 
-# Editors
-*.swp
-*.swo
-*~
-*.bak
-.idea/
-.vscode/
-*.sublime-*
+# Agent scratch: one full checkout of the repo per in-flight agent.
+# Anchored because it occurs once where agents run at the repo root.
+# KNOWN GAP: a repo running agents in subdirectories still ships
+# `services/api/.claude/` and must add its own anchored entry.
+.claude
 
-# Node
-node_modules
+# Environment files. `*.env` covers bare `.env` and the `prod.env`
+# convention. Re-include a committed template with a negation if the
+# build needs one: `!docs/example.env`.
+**/*.[eE][nN][vV]
+**/.[eE][nN][vV].*
+**/.[eE][nN][vV][rR][cC]
 
-# TypeScript / build artifacts
-dist
-build
-*.tsbuildinfo
-coverage
-.nyc_output/
+# Private keys and the bundles carrying them. Public certificates
+# (*.crt, *.cer) are deliberately absent: they are legitimate inputs.
+**/*.[pP][eE][mM]
+**/*.[kK][eE][yY]
+**/*.[pP]12
+**/*.[pP][fF][xX]
+**/[iI][dD]_[rR][sS][aA]
+**/[iI][dD]_[dD][sS][aA]
+**/[iI][dD]_[eE][cC][dD][sS][aA]
+**/[iI][dD]_[eE][cC][dD][sS][aA]_[sS][kK]
+**/[iI][dD]_[eE][dD]25519
+**/[iI][dD]_[eE][dD]25519_[sS][kK]
 
-# Vitest
-.vitest-cache/
+# Dependencies: restored inside the image, never copied in.
+**/node_modules
 
-# Environment / secrets
-.env
-.env.*
-*.pem
-*.key
+# OS metadata.
+**/.DS_Store
+**/Thumbs.db
+
+# Editor state: never a build input, and it churns COPY.
+**/*.swp
+**/*.swo
+**/*~
+**/*.bak
+**/.idea
+**/.vscode
+**/*.sublime-*
+
+# TypeScript / build artifacts: the image compiles its own.
+/dist
+/build
+/*.tsbuildinfo
+/coverage
+/.nyc_output
+/.vitest-cache
 
 # Compiled binary (built by make build-bin); around 100 MB
-bin/quak
+/bin/quak
 
 # quak runtime data (in case anyone runs the CLI from inside the repo)
-.quak/
-
-# Local per-developer tool state, including agent worktrees. Correctness,
-# not context size: a worktree copied in here has its own test/ tree, which
-# vitest globs alongside the real one, so the containerised suite runs N+1
-# times over and still reports success.
-.claude/
+/.quak

.gitignore

@@ -11,9 +11,41 @@ Thumbs.db
 .vscode/
 *.sublime-*
 
+# Agent scratch (worktrees of this repo, created and destroyed by
+# in-flight tooling). Unanchored: .gitignore patterns already match at
+# every depth, so no prefix is wanted here. This is not a .dockerignore
+# entry and must not be given a `**/` prefix on the way into one.
+.claude/
+
 # Node
 node_modules/
 
+# Secrets. Unanchored like every entry above, so each matches at every
+# depth. Matching is case-sensitive on Linux, so names use character
+# ranges rather than a lowercase form that misses `Server.Key`.
+
+# Environment files. `*.env` covers bare `.env` and the `prod.env`
+# convention. Only the templates `example.env` and `sample.env` are
+# re-included below. A repository that commits any other template adds
+# its own negation after these lines, for example `!.env.example`.
+*.[eE][nN][vV]
+.[eE][nN][vV].*
+.[eE][nN][vV][rR][cC]
+!example.env
+!sample.env
+
+# Private keys and the bundles carrying them.
+*.[pP][eE][mM]
+*.[kK][eE][yY]
+*.[pP]12
+*.[pP][fF][xX]
+[iI][dD]_[rR][sS][aA]
+[iI][dD]_[dD][sS][aA]
+[iI][dD]_[eE][cC][dD][sS][aA]
+[iI][dD]_[eE][cC][dD][sS][aA]_[sS][kK]
+[iI][dD]_[eE][dD]25519
+[iI][dD]_[eE][dD]25519_[sS][kK]
+
 # TypeScript / build artifacts
 dist/
 build/
@@ -24,18 +56,8 @@ coverage/
 # Vitest
 .vitest-cache/
 
-# Environment / secrets
-.env
-.env.*
-*.pem
-*.key
-
 # Compiled binary (built by make build-bin)
 bin/quak
 
 # quak runtime data (in case anyone runs the CLI from inside the repo)
 .quak/
-
-# Local per-developer tool settings and scratch state, including the
-# worktrees agents check out under this directory
-.claude/

.prettierignore

@@ -1,5 +1,2 @@
 node_modules/
 yarn.lock
-dist/
-build/
-coverage/

Dockerfile

@@ -58,6 +58,8 @@ COPY --from=test /app/package.json /dev/null
 COPY script/ script/
 COPY package.json yarn.lock ./
 RUN script/bootstrap
+# A tar-stream context keeps the sender's file owners, which git refuses.
+RUN git config --system --add safe.directory /app
 
 COPY . .
 

README.md

@@ -137,16 +137,15 @@ alpine. We provide:
 - `script/lint` — run eslint and a prettier check, by building the `lint` phase
   of the `Dockerfile`; requires docker (see Linting and testing below)
 - `script/fmt` — format all files with prettier (writes)
-- `script/fmt-check` — check formatting on the host (read-only); standalone, and
-  not called by `script/check` or `script/precommit`, because `script/lint`
-  already checks formatting in the container
-- `script/check` — run all checks: `test`, `lint` (our own extension)
+- `script/fmt-check` — check formatting on the host (read-only)
+- `script/check` — run all checks: `test`, `lint`, `fmt-check` (our own
+  extension)
 - `script/docker` — build the image, tagged via `script/projectname`
-- `script/cibuild` — build the image (what CI runs); its last stage depends on
-  the `lint` and `test` phases, so this one build lints, tests and compiles
+- `script/cibuild` — what CI runs: `script/bootstrap`, then `script/check`, then
+  the image build
 - `script/precommit` — run by the git pre-commit hook (our own extension); runs
-  `script/lint`, which checks both lint and formatting, but deliberately not the
-  tests, so the TDD red-phase commit can land
+  `script/lint` and `script/fmt-check` but deliberately not the tests, so the
+  TDD red-phase commit can land
 - `script/install-precommit` — installs the git pre-commit hook (our own
   extension); `make hooks` shims to it
 
@@ -162,22 +161,19 @@ script runs the tools on the host: docker is required, and that also works where
 the docker daemon is remote and bind mounts are impossible.
 
 The last stage of the `Dockerfile` compiles the package, and it copies a file
-from each phase, so it cannot be built unless lint and the tests pass. That is
-why `script/cibuild` is a single `docker build`: it runs lint and the tests once
-each and then compiles.
+from each phase, so it cannot be built unless lint and the tests pass. The image
+build in `script/cibuild` therefore runs lint and the tests a second time, after
+`script/check` has run them.
 
 Every `docker build` in `script/` passes `--no-cache`. On an unchanged tree
 Docker would otherwise serve the lint and test steps from cache, nothing would
 run, and the build would still exit 0.
 
-The formatting check is part of the `lint` phase, not a step beside it, so
-`script/check` and `script/precommit` do not call `script/fmt-check` as well;
-that would run prettier a second time over the same tree for the same verdict.
-`script/fmt-check` remains as a standalone entrypoint for asking the formatting
-question on the host. Its verdict matches the container's: prettier is pinned to
-an exact version, installed from `yarn.lock` under `--frozen-lockfile` in both
-places, and reads `.gitignore` as its default ignore file — which is why
-`.dockerignore` keeps `.gitignore` in the build context.
+`script/fmt-check` runs prettier on the host. Its verdict matches the `lint`
+phase's: prettier is pinned to an exact version, installed from `yarn.lock`
+under `--frozen-lockfile` in both places, and reads `.gitignore` as its default
+ignore file — which is why `.dockerignore` keeps `.gitignore` in the build
+context.
 
 ### Version
 
@@ -259,11 +255,10 @@ All work on quak is test-driven. No exceptions.
    history must still show tests landing before (or with) the matching
    implementation.
 8. The pre-commit hook installed by `make hooks` runs `script/precommit`, which
-   runs `script/lint` — eslint and the prettier check, in the container — but
-   not the tests, and so not the full `make check`. This is deliberate so the
-   TDD red-phase commit (failing tests, no implementation yet) can land. The
-   `test` phase is part of the image build, which is what CI executes via
-   `script/cibuild`, so a red branch still cannot reach `next`.
+   runs `script/lint` and `script/fmt-check` but not the tests, and so not the
+   full `make check`. This is deliberate so the TDD red-phase commit (failing
+   tests, no implementation yet) can land. CI executes `script/cibuild`, which
+   runs the tests, so a red branch still cannot reach `next`.
 
 ## Design
 
@@ -420,18 +415,24 @@ Backoff is exponential with full jitter: the delay before retry _n_ is
 `random() * min(maxDelayMs, baseDelayMs * 2 ** (n - 1))`. The exponential term
 is the ceiling and the wait is drawn below it, so a client that lost many
 parallel downloads to one CDN blip does not send them all again at the same
-instant. Defaults, configurable through `ApiClientOptions.retry`:
+instant. The numbers, configurable through `ApiClientOptions.retry`:
 
-| Option        | Default | Meaning                             |
-| ------------- | ------- | ----------------------------------- |
-| `attempts`    | `4`     | total calls, not retries            |
-| `baseDelayMs` | `500`   | ceiling for the first retry's delay |
-| `maxDelayMs`  | `10000` | upper bound on that ceiling         |
+| Option        | Default | `quak backup` | Meaning                             |
+| ------------- | ------- | ------------- | ----------------------------------- |
+| `attempts`    | `4`     | `10`          | total calls, not retries            |
+| `baseDelayMs` | `500`   | `1000`        | ceiling for the first retry's delay |
+| `maxDelayMs`  | `10000` | `60000`       | upper bound on that ceiling         |
 
-With those defaults a file that is going to fail gives up after at most three
-and a half seconds of waiting. `sleep` and `random` are injectable through the
-same option, which is how the test suite exercises the whole policy without
-waiting.
+With the defaults a file that is going to fail gives up after at most three and
+a half seconds of waiting. `quak backup` usually runs from cron with nobody
+watching, so every request it makes uses the `quak backup` column instead,
+exported as `UNATTENDED_RETRY_OPTIONS`: a request that keeps failing gives up
+after at most 243 seconds of waiting, and usually after about half that, since
+each wait is drawn at random below its ceiling. Every other command uses the
+defaults. A library user gets the same budget by passing
+`UNATTENDED_RETRY_OPTIONS` as `ApiClientOptions.retry`. `sleep` and `random` are
+injectable through the same option, which is how the test suite exercises the
+whole policy without waiting.
 
 Two deadlines, renewed for each attempt:
 
@@ -523,10 +524,10 @@ stack trace. All three exit with status 3, which means the user must run
 `quak login` again. `quak logout` is the exception: with no file it says there
 is no session and exits 0, and it handles a corrupt file or a failed server call
 as described below. `quak backup` meets an expired session on the refresh that
-starts every run, before it touches any file. A session that stops working
-partway through a backup instead fails each remaining file into `failures.json`,
-so that run exits 1 and the next one stops at its refresh with status 3. No
-command but `quak login` ever prompts.
+starts every run, before it touches any file but its lock (see "Backup layout").
+A session that stops working partway through a backup instead fails each
+remaining file into `failures.json`, so that run exits 1 and the next one stops
+at its refresh with status 3. No command but `quak login` ever prompts.
 
 `quak logout` ends the session on the server, so the token in `session.json`
 stops working even in a copy of the file, and then deletes the file. If the
@@ -548,7 +549,7 @@ quak collections [--json]                                 list all collections
 quak files --collection <id> [--json]                     list files in a collection
 quak get <fileID> [--out path] [--collection]             download and decrypt a file
 quak get-thumb <fileID> [--out] [--collection]            download and decrypt a thumbnail
-quak backup <dir> [--json]                                full incremental backup
+quak backup <dir> [--json] [--verify]                     full incremental backup
 quak backup-metadata <dir> [--exif]                       dump the metadata quak keeps as JSON
 quak helper list-missing-thumbnails [--json]              find files with missing thumbnails
 quak helper fix-missing-thumbnails [--file ids] [--json]  generate + upload missing thumbnails
@@ -581,6 +582,10 @@ reads no tag from is recorded, base64, as `exifRaw`, with the reason in
 `exifError`. `collections`, `files`, `backup`, `helper list-missing-thumbnails`
 and `helper fix-missing-thumbnails` take `--json` for machine-readable output.
 
+`backup --verify` also hashes the originals already in the backup and replaces
+any that do not match the content hash Ente records; one it cannot replace goes
+into `failures.json` (see "Backup layout").
+
 `backup-metadata` fetches ML data in requests of up to 200 files. When a request
 fails, the error is logged, each of its files is written with the reason in an
 `mlDataError` field instead of `mlData`, and the dump goes on. The exit code is
@@ -609,8 +614,9 @@ the smallest does not.
                                     per unique file, two for a live photo: see
                                     below)
         YYYY-MM-DD.<fileID>.json    the file's basic metadata fields quak
-                                    keeps, and its private and public magic
-                                    metadata
+                                    keeps, its update time, its private and
+                                    public magic metadata, its ML data, and
+                                    its original's EXIF, XMP and dimensions
         YYYY-MM-DD.<fileID>.livephoto.json
                                     which of a live photo's two files is which
     collections/
@@ -618,6 +624,8 @@ the smallest does not.
             <title> -> ../../YYYY/YYYY-MM/YYYY-MM-DD/YYYY-MM-DD.<fileID>.<ext>
                                     (symlink)
         <name>.json                 collection metadata + file list
+    account.json                    the account's email and user ID
+    backup.lock                     the lock a running backup holds (see below)
     failures.json                   files that failed and have not yet succeeded
 ```
 
@@ -629,12 +637,56 @@ name as uploaded, case kept, or `.bin` when it has none or it holds anything but
 letters and digits. When the date or the time zone changes, the next run saves
 the original at its new path and leaves the old copy where it is.
 
+`account.json` holds the account's `email` and `userID`, as `backup-metadata`
+writes it. A collection's JSON holds its `id`, `name`, `type`, `ownerID`,
+`isShared` and `updationTime`, its `magicMetadata`, `pubMagicMetadata` and
+`sharedMagicMetadata` when it has them, as `backup-metadata`'s
+`_collection.json` does, and `files`, each file's `id` and `metadata`. A file's
+JSON holds its `id`, `collectionID`, `ownerID`, `metadata` and `updationTime`,
+and its `magicMetadata` and `pubMagicMetadata` when it has them. Update times
+are in microseconds, as Ente records them.
+
+A file's JSON holds Ente's ML data for it (its faces and its CLIP embedding) as
+`mlData`, the same payload `backup-metadata` writes; a file Ente has no ML data
+for has no `mlData`. The backup waits for the library's ML data fetch to finish
+before it writes the JSON files. If that fetch fails, each file whose ML data is
+not in the cache gets the reason in `mlDataError` instead and counts as failed,
+and the next run fetches it again. The JSON files are rewritten on every run, so
+ML data that arrived since the last run appears.
+
+A file's JSON holds, as `imageMetadata`, what `backup-metadata --exif` records
+from its original, or from a live photo's image: `format`, `width` and `height`
+for a JPEG, `exif` (or `exifRaw` and `exifError`), and `xmp`. An original with
+none of these gets `{}`. A video gets no `imageMetadata`: reading a whole video
+to look for tags is not worth it. When the original cannot be read, the reason
+is in `imageMetadataError` instead; neither the file nor the run fails. The
+original is read when the run stores it, or when the JSON beside it has neither
+field, as one an earlier version wrote. Otherwise the field is taken from that
+JSON when it is rewritten, so a run does not read every stored original again.
+
 `failures.json` records each failed file with the kind of failure, how many
 times it has been tried and when it was last tried. A file leaves it once it
 succeeds, or once it is no longer in the library or in the backup's scope. The
 library's `lib.backup({ includeThumbnails: true })` also writes
 `thumbnails/<fileID>.jpg` beside `collections/`; `quak backup` does not.
 
+`backup.lock` keeps two backups of the same directory from running at once, such
+as a cron run that starts while the previous one is still going. A backup
+creates `<dir>` if it is missing and takes the lock before its refresh, and
+removes the lock when it ends, whether it succeeds or fails; `quak backup` takes
+it before it opens its library, so a refused run sends no request. The lock is a
+directory that
+[proper-lockfile](https://github.com/moxystudio/node-proper-lockfile) creates
+and keeps touching while the backup runs. A second backup of the directory, from
+another process or the same one, fails at once: `quak backup` prints
+`quak: another backup of <dir> is running` and exits with status 2. A run
+stopped with Ctrl-C or `kill` removes the lock as it exits. Only a run that
+cannot, such as one killed with SIGKILL or cut off by a crash or power loss,
+leaves it behind; once it has gone 10 seconds untouched, the next run takes it
+over, so nobody has to remove it. The lock is outside the date folders and
+`collections/`, so it is never taken for an original, and the removal of old
+album directories never touches it.
+
 A collection's directory and JSON are named after the collection, and a symlink
 after the file's title, both with unsafe characters replaced. When two
 collections would get the same name, or two symlinks in one collection the same
@@ -665,8 +717,26 @@ subsequent runs, existing originals are skipped. If a download fails, the error
 is logged and the backup continues with the next file. The exit code is non-zero
 if any files failed. `quak backup` opens its library with the thumbnail and
 originals precache off, so the only file content it fetches is the originals the
-backup stores. The library's ML data fetch still runs and fills the cache's
-`mldata/`.
+backup stores.
+
+With `--verify`, or `lib.backup({ verify: true })`, a run also hashes each
+original already at its save path the way a download is checked (see "On-disk
+cache layout" below): its bytes, read in chunks, or a live photo's image and
+video, joined as `<imageHash>:<videoHash>`. An original that matches the content
+hash its metadata records is left as it is. One that does not is logged on one
+line naming the file, deleted (a live photo's image and video both), and put
+back in the same run like a missing one: downloaded, or copied from the cache if
+the cache holds it. What is put back is hashed too, because a copy from the
+cache is not checked as a download is. If it still does not match, it stays at
+its save path and the file goes into `failures.json`, as it does when the
+download fails. A file whose metadata records no hash is left as it is and
+counted as unchecked. A stored original that cannot be read is left as it is and
+counts as failed. The summary and `--json` add the counts `verified`,
+`mismatched` and `unchecked`, all of originals that were already stored; one
+first downloaded in this run is in none of them. A mismatch that was put back
+with matching bytes does not make the exit code non-zero. Without `--verify`
+nothing is hashed, the summary is unchanged, and the three counts are 0 in
+`--json`.
 
 Each original is written to a temporary file in the same directory, synced to
 disk, and renamed into place, so an original is either complete or absent, even
@@ -883,16 +953,23 @@ photos newest first). `lib.subscribe({ onChange })` delivers a `LibraryChange`
   `SimilarResult[]` (`{ fileID, score }`, cosine similarity, most similar first,
   default limit 20). quak bundles no text encoder, so `searchByEmbedding` takes
   a query vector the caller produced elsewhere.
-- `await lib.backup(opts?)` → `BackupResult`. It waits for a refresh as
+- `await lib.backup(opts?)` → `BackupResult`. It takes the lock in the download
+  directory, and fails at once with an error whose `code` is `ELOCKED` while
+  another backup of it runs. A caller can instead take the lock itself, before
+  it opens its library, as `quak backup` does: `await lockBackupDirectory(dir)`
+  takes it, failing the same way, and returns the function that releases it, and
+  the caller passes `lockHeld: true` to the backup. It waits for a refresh as
   `fresh()` does, puts every in-scope original not already at its save path
   there as `photo.download()` does (and, with `includeThumbnails`, fetches
-  thumbnails) through the content cache, and rebuilds the on-disk backup tree
-  with a durable failure ledger. A fetched original is written straight to its
-  save path and not into the cache, which then counts it as present; one the
-  cache already held is copied from there. `BackupOptions`: `downloadDirectory`
-  (falls back to the library's), `includeOriginals` (default `true`),
-  `includeThumbnails` (default `false`), `onlyAlbumNames`, and `onProgress`. See
-  Backup layout above for the tree it writes.
+  thumbnails) through the content cache, waits for an ML data fetch, and
+  rebuilds the on-disk backup tree, each file's JSON with its ML data and its
+  original's EXIF, XMP and dimensions, with a durable failure ledger. A fetched
+  original is written straight to its save path and not into the cache, which
+  then counts it as present; one the cache already held is copied from there.
+  `BackupOptions`: `downloadDirectory` (falls back to the library's),
+  `includeOriginals` (default `true`), `includeThumbnails` (default `false`),
+  `onlyAlbumNames`, `verify` (default `false`), `onProgress`, and `lockHeld`
+  (default `false`). See Backup layout above for the tree it writes.
 
 ### Request pools
 
@@ -987,14 +1064,12 @@ documents:
   before the implementation. Tests are the canonical API documentation and must
   be commented thoroughly. `main` and `next` are always green.
 
-- **Required checks before every commit:** `make lint` must pass — that is
-  eslint plus the prettier check, and it builds the `lint` phase of the
-  `Dockerfile`, so it needs docker. The pre-commit hook enforces exactly that.
-  `make check` (which also runs the tests) must pass before merging into `next`.
-  `make fmt-check` is available for a host-side formatting check on its own, but
-  it is not a separate requirement: `make lint` already covers it, and running
-  both would check formatting twice. Never invoke eslint or prettier directly;
-  linting runs in the container only.
+- **Required checks before every commit:** `make lint` and `make fmt-check` must
+  pass. `make lint` is eslint plus the prettier check, and it builds the `lint`
+  phase of the `Dockerfile`, so it needs docker. The pre-commit hook enforces
+  exactly that. `make check` (which also runs the tests) must pass before
+  merging into `next`. Never invoke eslint or prettier directly; linting runs in
+  the container only.
 
 - **Formatting:** prettier with 4-space indents and `proseWrap: always` for
   markdown. Use `make fmt` to format. Use `yarn` not `npm`.

REPO_POLICIES.md

@@ -1,6 +1,6 @@
 ---
 title: Repository Policies
-last_modified: 2026-09-08
+last_modified: 2026-10-04
 ---
 
 This document covers repository structure, tooling, and workflow standards. Code
@@ -104,10 +104,14 @@ style conventions are in separate documents:
   `lint` phase and a `test` phase, with the final stage depending on both so the
   image cannot be built unless they pass. For non-server repos the final stage
   brings up a development environment; for server repos it is the runtime image.
-  Dockerfiles install development prerequisites by running `script/bootstrap`
-  rather than duplicating installs inline; COPY `script/` and the dependency
-  manifests (`package.json` + `yarn.lock`, `go.mod` + `go.sum`, etc.) before
-  running it.
+  The gate phases and the build stage start from their pinned base images and
+  install what those images lack either inline, as the canonical Go `Dockerfile`
+  below does for `git`, or by running `script/bootstrap`, as the `prompts`
+  repo's own `Dockerfile` does for its yarn packages. The development
+  environment stage installs development prerequisites by running
+  `script/bootstrap` rather than duplicating its installs inline. A stage that
+  runs `script/bootstrap` COPYs `script/` and the dependency manifests
+  (`package.json` + `yarn.lock`, `go.mod` + `go.sum`, etc.) before running it.
 
 - **Linting and testing run in Docker, as phases of the `Dockerfile`.** There is
   no separate lint file. `script/lint` and `script/test` each build one phase
@@ -156,11 +160,14 @@ style conventions are in separate documents:
   not evidence that anything ran: a sub-second build reporting success is a
   cache hit, not a result. Never invalidate by pruning — `docker builder prune`
   and friends destroy a build cache shared with every other build on the host.
+  When a check is added or changed, prove it works by planting a defect it must
+  catch and watching the run fail on it, then revert the defect. A green run
+  alone shows neither that the check ran nor that it covers what it should.
 
 - **The gate phases are separate stages, and the build stage depends on both.**
   The lint phase is based on the `golangci/golangci-lint` image (pinned by
   hash), so lint failures surface in seconds rather than after a full compile,
-  and the test phase is based on the Go image. The canonical Go repo
+  and the test phase is based on the Debian Go image. The canonical Go repo
   `Dockerfile`:
 
     ```dockerfile
@@ -173,8 +180,9 @@ style conventions are in separate documents:
     COPY . .
     RUN golangci-lint run --config .golangci.yml ./...
 
-    # Test phase
-    # golang:1.x-alpine, YYYY-MM-DD
+    # Test phase. -race needs cgo and so a C compiler, which the Debian Go
+    # image ships and the alpine one does not.
+    # golang:1.x, YYYY-MM-DD
     FROM golang@sha256:... AS test
     WORKDIR /src
     COPY go.mod go.sum ./
@@ -191,13 +199,27 @@ style conventions are in separate documents:
     FROM golang@sha256:... AS builder
     COPY --from=lint /src/go.sum /dev/null
     COPY --from=test /src/go.sum /dev/null
+    RUN apk add --no-cache git
+    # A tar-stream context keeps the sender's file owners, which git refuses.
+    RUN git config --system --add safe.directory /src
     WORKDIR /src
     COPY go.mod go.sum ./
     RUN go mod download
     COPY . .
 
-    ARG VERSION=dev
-    RUN CGO_ENABLED=0 go build -trimpath \
+    # The VERSION build arg when one is given, otherwise
+    # `git describe --tags --always` on the .git in the build context. With
+    # .git present, a version that is still empty, dev or unknown fails the
+    # build: git is missing or could not read the checkout.
+    ARG VERSION
+    RUN VERSION="${VERSION:-$(git describe --tags --always)}"; \
+        if [ -e .git ]; then \
+            case "$VERSION" in ""|dev|unknown) \
+                echo "version is '$VERSION' although .git is present" >&2; \
+                exit 1 ;; \
+            esac; \
+        fi; \
+        CGO_ENABLED=0 go build -trimpath \
             -ldflags="-s -w -X main.Version=${VERSION}" \
             -o /app ./cmd/app/
 
@@ -221,10 +243,41 @@ style conventions are in separate documents:
       (e.g. a web frontend compiled in a separate stage), the lint phase must
       create placeholder files so the embed directives resolve. Example:
       `RUN mkdir -p web/dist && touch web/dist/index.html web/dist/style.css`.
-    - If the project requires CGO or system libraries for linting (e.g.
-      `vips-dev`), install them in the lint phase with `apk add`.
-    - `ARG VERSION=dev` is declared in the stage that compiles and supplied by
-      `script/docker` and `script/cibuild`; no stage may call `git describe`.
+    - If the project requires CGO or system libraries for linting, install them
+      in the lint phase. The `golangci/golangci-lint` image is Debian-based and
+      has no `apk`, so install with `apt-get` under the Debian package name
+      (`libvips-dev`, where alpine says `vips-dev`), and delete the package
+      lists in the same `RUN`, so the layer does not keep them:
+
+        ```dockerfile
+        RUN apt-get update \
+            && apt-get install -y --no-install-recommends libvips-dev \
+            && rm -rf /var/lib/apt/lists/*
+        ```
+
+    - `.dockerignore` lets `.git` into the build context. It keeps out every git
+      `config` at any depth (`**/.git/config`, `**/.git/modules/**/config`): the
+      repository's own, each submodule's under `.git/modules/`, and that of a
+      submodule keeping its own `.git` directory. `git describe` does not need
+      them, and each can hold a credential: a password in a remote URL, or the
+      token the CI checkout step stores there. A submodule whose name has a
+      `config` segment (`config`, `deploy/config`, `config/lib`) loses its whole
+      git directory to `**/.git/modules/**/config`, and Go's version stamping
+      then fails the build: give it a name without that segment
+      (`git submodule add --name`). The stage that compiles has `git` (the
+      Debian Go image has it; an alpine one needs `apk add --no-cache git`) and
+      takes the version from the `VERSION` build argument when one is given,
+      otherwise from `git describe --tags --always`. That gives the tag on a
+      tagged commit; on a later commit, the tag, the number of commits since it
+      and the short commit (`v1.2.3-4-gabc1234`); and the short commit when no
+      tag is reachable. The stage that compiles also marks its working directory
+      safe for git (`git config --system --add safe.directory /src`): a context
+      sent as a tar stream keeps the sender's file owners, and git refuses a
+      checkout owned by another user, so the version would come out empty.
+      `ARG VERSION` has no default, and the build fails if the context carries
+      `.git` and the version still comes out empty, `dev` or `unknown`. A plain
+      `docker build .` with no build arguments must succeed; a Dockerfile that
+      refuses an empty build argument drops that refusal and keeps the argument.
 
 - Every repo should have a Gitea Actions workflow (`.gitea/workflows/`) that
   runs `script/cibuild` on push, and checks out the repo as its only other step.
@@ -233,7 +286,12 @@ style conventions are in separate documents:
   carry the same guarantee, because its gate phases may come from the cache. The
   image build is uncached and so runs the gate phases a second time. That is the
   price of the rule above, and it is worth paying: the image that ships is built
-  from a run of its own gates rather than from a cache entry.
+  from a run of its own gates rather than from a cache entry. A separate
+  workflow limited to `main` by a `branches` list under `on: push` cannot be
+  checked by review: to try a change to it, add the feature branch to that list
+  and push, then remove the branch from the list again before merging. Keep any
+  job in it that publishes behind `if: github.ref_name == 'main'`, so the run
+  from the feature branch publishes nothing.
 
 - Use platform-standard formatters: `black` for Python, `prettier` for
   JS/CSS/Markdown/HTML, `go fmt` for Go. Always use default configuration with
@@ -286,17 +344,19 @@ style conventions are in separate documents:
     ```
 
     `-count=1` is required on both invocations: it defeats Go's test _result_
-    cache, so the target cannot report a pass it did not earn, and the rerun
-    reproduces a failure instead of replaying it. It leaves the build cache
-    alone, so it costs the runtime of the suite and no recompilation.
+    cache, so neither run can report a stored pass in place of running the
+    tests. It leaves the build cache alone, so it costs the runtime of the suite
+    and no recompilation.
 
-    Note that this is a second, independent cache, stacked below the Docker
-    layer cache that [issue #26](https://git.eeqj.de/sneak/prompts/issues/26)
-    addresses. `CHECK_EPOCH` guarantees the `RUN make test` _step_ re-executes;
-    it does not guarantee `go test` inside that step does any work, because the
-    `GOCACHE` baked into earlier image layers survives into the re-executed
-    step. They are two separate defects requiring two separate fixes, and a fix
-    for one must not be recorded as covering the other.
+    That cache is Go's own, separate from Docker's layer cache. Go stores a
+    passing result in its cache directory (`GOCACHE`), and when the same tests
+    run again on unchanged code it prints that result, marked `(cached)`,
+    without running them. That matters on a developer's machine, where this
+    target runs and the directory lasts from one run to the next. The `test`
+    phase of the `Dockerfile` needs no `-count=1`: its base image holds no
+    result for this repo's tests and nothing before its `go test` step runs a
+    test, so there is nothing to replay. `--no-cache` (above) is what makes that
+    step run on an unchanged tree.
 
     Python example:
 
@@ -340,7 +400,7 @@ style conventions are in separate documents:
   — which is more dangerous than a short file with no secret patterns at all,
   because it reads as solved and stops anyone looking. Give every
   depth-independent pattern the `**/` prefix and leave only genuinely
-  root-anchored entries unprefixed: `.git`, and the repo's own host-built
+  root-anchored entries unprefixed: `.claude`, and the repo's own host-built
   binary, written `/myapp` and never `**/myapp`, which would also match
   `cmd/myapp/` and delete the package directory from the context. Matching is
   case-sensitive, and an ALL-CAPS twin per pattern still misses `Server.Key`, so
@@ -365,12 +425,13 @@ style conventions are in separate documents:
   directory, so a repo running agents in subdirectories still ships
   `services/api/.claude/` and must add its own anchored entry there.
 
-- **Excluding `.git` means `git describe` cannot run inside any build stage, and
-  it fails quietly there.** In a build stage there is no repository, so
-  `git describe` writes nothing to stdout, `-X main.Version=` comes out empty,
-  the binary reports no version at all, and the build still exits 0. Compute the
-  version on the host and thread it in as a build arg. `script/docker` and
-  `script/cibuild` do this, byte-identically across repos:
+- **A plain `docker build .` of a clone stamps the version that
+  `git describe --tags --always` gives**, derived from the `.git` in the build
+  context as the canonical `Dockerfile` above shows. Without its failure check,
+  a missing `git` or an unreadable checkout would leave `-X main.Version=` empty
+  and the build would still exit 0. `script/docker` and `script/cibuild` pass
+  the version they compute on the host; it takes precedence. They do this
+  byte-identically across repos:
 
     ```sh
     # Own line: a failing command substitution inside an argument does not
@@ -387,7 +448,7 @@ style conventions are in separate documents:
     fallback is applied — a live check that fires on a build from an export with
     no `.git` and on a repository with no commits yet. Do not fold it into the
     substitution as `|| echo unknown`, which makes the guard unreachable. The
-    Dockerfile's side is `ARG VERSION=dev` in the stage that compiles, declared
+    Dockerfile's side is `ARG VERSION` in the stage that compiles, declared
     there because `ARG` is stage-scoped; passing `VERSION` to a repo whose
     Dockerfile declares no such `ARG` is ignored and costs nothing, which is why
     the scripts stay byte-identical. One consequence for CI: the standard
@@ -426,12 +487,18 @@ style conventions are in separate documents:
   `test-support` depguard rule, where a repo names its own test-support packages
   by full import path. A repo adds entries there and changes nothing else, and a
   re-vendor carries its entries forward. The canonical golangci-lint version is
-  v2.12.2 (released 2026-05-06), pinned as the digest of the lint phase's base
+  v2.14.0 (released 2026-09-24), pinned as the digest of the lint phase's base
   image
-  (`golangci/golangci-lint@sha256:5cceeef04e53efe1470638d4b4b4f5ceefd574955ab3941b2d9a68a8c9ad5240`,
-  which reports `2.12.2 built with go1.26.2 from c0d3ddc9`). That digest is the
-  only pin, since no repo installs golangci-lint on the host: bumping the
-  version means changing it and nothing else.
+  (`golangci/golangci-lint@sha256:ad862ba6b3798cbe0fd9fd7408d498fd74fbd2623a92406b2fd3898faf0bf98f`,
+  which reports `2.14.0 built with go1.27.0 from 114493f9`). A module's `go`
+  directive must not name a newer Go minor version than the one golangci-lint
+  was built with, or golangci-lint refuses to lint it: this release lints
+  `go 1.27.1` but not `go 1.28`. That digest is the only pin, since no repo
+  installs golangci-lint on the host. A repo sets the lint phase digest to the
+  one named here and re-vendors `.golangci.yml` in the same commit, whichever of
+  the two prompted the change: the canonical copy can name linters that an older
+  golangci-lint rejects, and a newer golangci-lint can add linters that
+  `default: all` switches on until the canonical copy disables them.
 
 - **`script/bootstrap` installs a pinned tool by comparing versions, never by
   testing presence.** An `if ! command -v <tool>; then install; fi` guard tests
@@ -455,6 +522,11 @@ style conventions are in separate documents:
 
     Keep it POSIX sh: no arrays, no `[[`, no `grep -P`.
 
+    A Go tool a repo needs on the host is installed with `go install` pinned to
+    a commit hash (`go install <package>@<commit hash>`). It is never tracked as
+    a `go.mod` tool dependency or through a `tools.go` file, either of which
+    pulls the tool's own dependencies into the repo's `go.mod` and `go.sum`.
+
 - When pinning images or packages by hash, add a comment above the reference
   with the version and date (YYYY-MM-DD).
 
@@ -567,10 +639,10 @@ style conventions are in separate documents:
   settings.
 
 - Avoid putting files in the repo root unless necessary. Root should contain
-  only project-level config files (`README.md`, `Makefile`, `Dockerfile`,
-  `LICENSE`, `.gitignore`, `.editorconfig`, `REPO_POLICIES.md`, and
-  language-specific config). Everything else goes in a subdirectory. Canonical
-  subdirectory names:
+  only project-level config files (`README.md`, `AGENTS.md`, `Makefile`,
+  `Dockerfile`, `LICENSE`, `.gitignore`, `.editorconfig`, `REPO_POLICIES.md`,
+  and language-specific config). Everything else goes in a subdirectory.
+  Canonical subdirectory names:
     - `bin/` — executable scripts and tools
     - `cmd/` — Go command entrypoints; thin only: one `main.go` per binary whose
       body is a single call into `internal/` or `pkg/`, no project logic in
@@ -601,3 +673,7 @@ style conventions are in separate documents:
     - Go: `go.mod`, `go.sum`, `.golangci.yml`
     - JS: `package.json`, `yarn.lock`, `.prettierrc`, `.prettierignore`
     - Python: `pyproject.toml`
+
+- Guidance for coding agents lives in one `AGENTS.md` at the repository root. It
+  is never committed under a file or directory named after one agent tool, such
+  as `CLAUDE.md` or `.claude/`, and never split into separate memory files.

TODO.md

@@ -25,14 +25,73 @@ declares one.
 
 # Completed Steps
 
-- 2026-10-05: A command whose saved session the server no longer accepts
-  (HTTP 401) prints one line,
+- 2026-10-06: `quak backup --verify` and `lib.backup({ verify: true })` hash
+  each original already at its save path as the download check does, streamed, a
+  live photo as `<imageHash>:<videoHash>` (issue 168). One that does not match
+  the content hash its metadata records is logged, removed (both files of a live
+  photo) and put back in the same run, downloaded or copied from the cache, and
+  what is put back is hashed too. One that still does not match, or whose
+  download fails, goes into `failures.json`. One with no recorded hash is left
+  alone. The result, `--json` and the summary gain `verified`, `mismatched` and
+  `unchecked`. Without `--verify` nothing is hashed.
+
+- 2026-10-06: Two backups of the same directory never run at once (issue 169).
+  `lib.backup()` takes a lock, `backup.lock` in its download directory, made
+  with `proper-lockfile`, before its refresh, and removes it when it ends,
+  whether it succeeds or fails. A second backup of the directory, from another
+  process or the same one, fails at once with an error naming the directory.
+  `quak backup` takes the lock before it opens its library, so a refused run
+  sends no request; it prints the error as one line and exits 2. A lock that has
+  gone 10 seconds untouched, left by a run that could not remove it, is taken
+  over by the next run.
+
+- 2026-10-06: `quak backup` writes each original's EXIF, XMP and dimensions into
+  the file's JSON as `imageMetadata`, what `backup-metadata --exif` records
+  (issue 167): for a live photo from its image, for a video nothing, and `{}`
+  for an original with none of them. A failed read puts the reason in
+  `imageMetadataError` and fails neither the file nor the run. An original is
+  read when the run stores it or when its JSON has neither field; otherwise the
+  field is taken from that JSON. The hand-built JPEGs moved to
+  `test/exif-jpeg.ts`, beside the HEIC.
+
+- 2026-10-06: When the server answers HTTP 401 and that ends a command that
+  loads the saved session, because the server no longer accepts its token, the
+  command prints one line,
   `quak: the saved session is no longer valid; run "quak login"`, and exits 3
-  (issue 164). A missing or corrupt session file keeps its message and now also
-  exits 3, so a cron job can tell that the user must log in again. `run` in
-  `src/cli-run.ts` recognises the 401, which reaches it unchanged from the
-  refresh, so `quak backup` stops there before it touches any file. Every other
-  error still exits 1.
+  (issue 164). `quak backup` meets that 401 on the refresh it starts with,
+  before it touches any file. For the same commands, a missing or corrupt
+  session file keeps its message and also exits 3, so a cron job can tell that
+  the user must log in again. `quak logout` is unchanged. `run` in
+  `src/cli-run.ts` recognises the 401, which reaches it unchanged.
+
+- 2026-10-06: `quak backup` retries a failed request for longer than the other
+  commands do (issue 165). `src/retry.ts` exports `UNATTENDED_RETRY_OPTIONS`
+  beside the unchanged default: 10 attempts, a 1 s base delay and a 60 s cap, so
+  a request that keeps failing waits at most 243 s before it gives up.
+  `bin/quak.ts` loads the backup's session with them, so its refresh, ML data
+  and downloads all use them. What is retried and the backoff formula are
+  unchanged.
+
+- 2026-10-06: The files this repository copies from `sneak/prompts` are copied
+  again from its commit `dd4027b` (issue 171). `.gitignore` and `.dockerignore`
+  keep out more secret files, and this repository's build artifacts follow the
+  copied content. `script/check` runs `script/fmt-check` again, `script/cibuild`
+  runs `script/bootstrap` and `script/check` before the image build,
+  `script/fmt` and `script/fmt-check` find the pinned yarn under nvm, and the
+  image's last stage marks `/app` safe for git.
+
+- 2026-10-06: `quak backup` writes the account and album records
+  `backup-metadata` writes (issue 166): `account.json` with the account's
+  `email` and `userID`, and in each album's JSON its `ownerID`, `isShared`,
+  `updationTime` and, when present, its three layers of magic metadata. Each
+  file's JSON gains its `updationTime`.
+
+- 2026-10-05: `quak backup` writes each file's ML data (its faces and CLIP
+  embedding) into the file's JSON as `mlData`, the payload
+  `lib.mldata.forFile()` returns (issue 163). `lib.backup()` waits for an ML
+  data fetch before it writes the JSON files. When that fetch fails, each file
+  whose ML data is not cached gets the reason in `mlDataError` and counts as
+  failed, and the next run fetches it again.
 
 - 2026-10-03: The download-albums example test no longer fails on a disk with
   under 50 GiB free (issue 160). It opens its libraries with

bin/quak.ts

@@ -23,6 +23,7 @@ import { run as runCommand } from "../src/cli-run.js";
 import { loadSession } from "../src/cli-session.js";
 import { Client } from "../src/client.js";
 import { VERSION } from "../src/index.js";
+import { UNATTENDED_RETRY_OPTIONS } from "../src/retry.js";
 
 const paths = envPaths("quak", { suffix: "" });
 
@@ -129,8 +130,24 @@ program
     )
     .argument("<dir>", "Output directory")
     .option("--json", "Print result as JSON instead of human-readable summary")
-    .action((dir: string, opts: { json?: boolean }) =>
-        run(backupCommand(context(), dir, opts)),
+    .option(
+        "--verify",
+        "Re-hash stored originals and download again any that do not match",
+    )
+    // A backup usually runs from cron with nobody watching, so every request
+    // it makes retries for longer than the other commands' requests do.
+    .action((dir: string, opts: { json?: boolean; verify?: boolean }) =>
+        run(
+            backupCommand(
+                {
+                    ...context(),
+                    loadSession: (path) =>
+                        loadSession(path, { retry: UNATTENDED_RETRY_OPTIONS }),
+                },
+                dir,
+                opts,
+            ),
+        ),
     );
 
 const helper = program

package.json

@@ -36,6 +36,7 @@
     "devDependencies": {
         "@eslint/js": "9.38.0",
         "@types/node": "22.18.13",
+        "@types/proper-lockfile": "4.1.4",
         "eslint": "9.38.0",
         "prettier": "3.8.1",
         "typescript": "5.9.3",
@@ -50,6 +51,7 @@
         "fast-srp-hap": "2.0.4",
         "fflate": "0.8.3",
         "jpeg-js": "0.4.4",
-        "libsodium-wrappers-sumo": "0.8.4"
+        "libsodium-wrappers-sumo": "0.8.4",
+        "proper-lockfile": "4.1.2"
     }
 }

script/check

@@ -1,11 +1,8 @@
 #!/bin/sh
-# script/check: run all checks (test, lint). Our own extension to
-# scripts-to-rule-them-all. Both are Docker phases. Must not modify any
-# files.
-#
-# script/fmt-check is not called here, unlike the template: the lint
-# phase already runs `prettier --check .`, so calling it would run
-# prettier a second time over the same tree for the same verdict.
+# script/check: run all checks (test, lint, fmt-check). Our own
+# extension to scripts-to-rule-them-all. test and lint are Docker
+# phases; fmt-check is native, because a formatter writes the working
+# tree. Must not modify any files.
 set -eu
 
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd -P)"
@@ -13,6 +10,7 @@ SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd -P)"
 main() {
     "$SCRIPT_DIR/test"
     "$SCRIPT_DIR/lint"
+    "$SCRIPT_DIR/fmt-check"
 }
 
 main "$@"

script/cibuild

@@ -1,8 +1,7 @@
 #!/bin/sh
-# script/cibuild: run the CI build. The image's last stage depends on the
-# lint and test phases, so this one build runs eslint, prettier and the
-# suite once each and then compiles. Unlike the template it does not run
-# script/check first, which would run lint and the tests a second time.
+# script/cibuild: run the CI build. It bootstraps first: a CI runner
+# checks out and runs this and nothing else, and script/fmt-check runs
+# the formatter on the host, which a pristine checkout cannot do.
 # --no-cache for the same reason as script/docker: the gate phases the
 # final stage depends on are RUN steps, and a cached one is a check that
 # did not run.
@@ -13,11 +12,12 @@ ROOT="$(cd "$SCRIPT_DIR/.." && pwd -P)"
 
 main() {
     cd "$ROOT"
+    "$SCRIPT_DIR/bootstrap"
+    "$SCRIPT_DIR/check"
     # Own line: a failing command substitution inside an argument does
     # not trip `set -e`, so the inline form degrades silently to an
-    # empty constant. The version resolved here goes in as the VERSION
-    # build arg, which takes precedence over what the build would derive
-    # from the .git in its context.
+    # empty constant. The VERSION build argument takes precedence over
+    # the version a build stage derives from the .git in the context.
     version="$(git describe --tags --always --dirty 2>/dev/null || true)"
     [ -n "$version" ] || version="unknown"
     docker build --no-cache \

script/docker

@@ -12,9 +12,8 @@ main() {
     cd "$ROOT"
     # Own line: a failing command substitution inside an argument does
     # not trip `set -e`, so the inline form degrades silently to an
-    # empty constant. The version resolved here goes in as the VERSION
-    # build arg, which takes precedence over what the build would derive
-    # from the .git in its context.
+    # empty constant. The VERSION build argument takes precedence over
+    # the version a build stage derives from the .git in the context.
     version="$(git describe --tags --always --dirty 2>/dev/null || true)"
     [ -n "$version" ] || version="unknown"
     docker build --no-cache \

script/fmt

@@ -4,9 +4,28 @@ set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
+# Must match the pin in script/bootstrap.
+NODE_VERSION="22.17.0"
+
+# script/bootstrap installs node and yarn under nvm and leaves neither
+# on the PATH of the shell that called it, so resolve the pinned
+# toolchain here the way bootstrap's own install step does. nvm is a
+# bash script, hence the subshell.
+run_yarn() {
+    if command -v yarn >/dev/null 2>&1; then
+        exec yarn "$@"
+    fi
+    if [ ! -s "$HOME/.nvm/nvm.sh" ]; then
+        echo "fmt: no yarn; run script/bootstrap first" >&2
+        exit 1
+    fi
+    exec bash -c '. "$HOME/.nvm/nvm.sh" && nvm use "$1" >/dev/null &&
+        shift && exec yarn "$@"' bash "$NODE_VERSION" "$@"
+}
+
 main() {
     cd "$ROOT"
-    yarn run prettier --write .
+    run_yarn run prettier --write .
 }
 
 main "$@"

script/fmt-check

@@ -4,9 +4,28 @@ set -eu
 
 ROOT="$(cd "$(dirname "$0")/.." && pwd -P)"
 
+# Must match the pin in script/bootstrap.
+NODE_VERSION="22.17.0"
+
+# script/bootstrap installs node and yarn under nvm and leaves neither
+# on the PATH of the shell that called it, so resolve the pinned
+# toolchain here the way bootstrap's own install step does. nvm is a
+# bash script, hence the subshell.
+run_yarn() {
+    if command -v yarn >/dev/null 2>&1; then
+        exec yarn "$@"
+    fi
+    if [ ! -s "$HOME/.nvm/nvm.sh" ]; then
+        echo "fmt-check: no yarn; run script/bootstrap first" >&2
+        exit 1
+    fi
+    exec bash -c '. "$HOME/.nvm/nvm.sh" && nvm use "$1" >/dev/null &&
+        shift && exec yarn "$@"' bash "$NODE_VERSION" "$@"
+}
+
 main() {
     cd "$ROOT"
-    yarn run prettier --check .
+    run_yarn run prettier --check .
 }
 
 main "$@"

script/precommit

@@ -2,17 +2,17 @@
 # script/precommit: run by the git pre-commit hook; fails the commit if
 # checks fail. Our own extension to scripts-to-rule-them-all.
 #
-# Runs lint but deliberately NOT the tests, so the TDD red-phase commit
-# (failing tests, no implementation yet) can land. CI runs
-# script/cibuild, whose image build includes the test phase, and so
-# catches any branch that ships red. The lint phase includes the
-# prettier check, so a badly formatted tree still fails the commit.
+# Runs lint and fmt-check but deliberately NOT the tests, so the TDD
+# red-phase commit (failing tests, no implementation yet) can land. CI
+# runs script/cibuild, which runs the tests, and so catches any branch
+# that ships red.
 set -eu
 
 SCRIPT_DIR="$(cd "$(dirname "$0")" && pwd -P)"
 
 main() {
     "$SCRIPT_DIR/lint"
+    "$SCRIPT_DIR/fmt-check"
 }
 
 main "$@"

src/backup.ts

@@ -1,18 +1,25 @@
 // The backup command, rebuilt on the library API (issue #51).
 //
-// `lib.backup()` waits for a completed refresh of the library (a failed one
-// fails the backup before any file is touched), then, for every file in scope,
-// puts its original at its save path under `downloadDirectory`, as
-// `Photo.download()` does, and rebuilds the derived views (per-file sidecars,
-// per-collection symlink trees, per-collection JSON) from the model. The
-// on-disk layout:
+// `lib.backup()` takes the lock in `downloadDirectory` (unless its caller
+// already holds it), and fails at once when another backup of it holds the
+// lock. It waits for a completed refresh of the library (a failed one fails the
+// backup before any file is touched), then, for every file in scope, puts its
+// original at its save path under `downloadDirectory`, as `Photo.download()`
+// does, waits for an ML data fetch, and rebuilds the derived views (per-file
+// sidecars, per-collection symlink trees, per-collection JSON) from the model.
+// The on-disk layout:
 //
 //     <downloadDirectory>/
 //         YYYY/YYYY-MM/YYYY-MM-DD/
 //             YYYY-MM-DD.<fileID>.<ext>    the decrypted bytes (the save path)
-//             YYYY-MM-DD.<fileID>.json     per-file metadata sidecar
+//             YYYY-MM-DD.<fileID>.json     per-file metadata sidecar, with
+//                                          the file's ML data and its
+//                                          original's EXIF, XMP and
+//                                          dimensions
 //         collections/<name>/<title>       symlink to the original
 //         collections/<name>.json          per-collection metadata
+//         account.json                     the account's email and user ID
+//         backup.lock                      the lock, while a backup runs
 //         failures.json                    durable ledger of unresolved failures
 //
 // A live photo's original is its image and its video, each with its own
@@ -26,18 +33,29 @@
 // no unique state, so they are rebuilt every run; that repairs stale sidecars
 // and missing or broken symlinks left by an earlier crash. A rebuild also
 // removes the symlinks to originals that no longer belong to an album, and the
-// directories of albums that no longer exist.
+// directories of albums that no longer exist. The one thing a sidecar takes
+// from the sidecar it replaces is its original's EXIF, XMP and dimensions (or
+// why they could not be read), so that a run does not read every stored
+// original again; a sidecar without them gets them read from the original.
+//
+// With `verify`, each original already at its save path is hashed as the
+// download check hashes it, and one that does not match the content hash its
+// metadata records is removed and fetched again in the same run. What is put
+// back is hashed too, and recorded as failed if it still does not match.
 //
 // Resilience (issue #8): no per-file condition aborts the run. A failed
-// download or a failed symlink is caught, recorded in `failures.json` with a
-// classification, a running attempt count, and the last-tried time, and the run
-// continues. `result.failed` — and thus the CLI's exit code — stays non-zero
-// while any failure remains unresolved and clears once every one succeeds. Each
-// run reconciles the ledger against the files it attempted, so an entry for a
-// file that has since left the library (deleted) or this run's scope is dropped
-// rather than counted forever, which would poison a scheduled backup's exit code.
+// download, a failed symlink, or ML data missing because the ML data fetch
+// failed is caught, recorded in `failures.json` with a classification, a
+// running attempt count, and the last-tried time, and the run continues.
+// `result.failed` — and thus the CLI's exit code — stays non-zero while any
+// failure remains unresolved and clears once every one succeeds. Each run
+// reconciles the ledger against the files it attempted, so an entry for a file
+// that has since left the library (deleted) or this run's scope is dropped
+// rather than counted forever, which would poison a scheduled backup's exit
+// code.
 
 import {
+    createReadStream,
     lstatSync,
     mkdirSync,
     readdirSync,
@@ -49,8 +67,16 @@ import {
     symlinkSync,
     writeFileSync,
 } from "node:fs";
+import { readFile } from "node:fs/promises";
 import { dirname, extname, join, relative, resolve } from "node:path";
+import lockfile from "proper-lockfile";
 
+import {
+    chunkHashFinal,
+    chunkHashInit,
+    chunkHashUpdate,
+    init,
+} from "./crypto/index.js";
 import { removeLeftoverTempFiles } from "./download/index.js";
 import { sanitizeFileName, withExtension } from "./filename.js";
 import {
@@ -60,7 +86,9 @@ import {
     storedAtSavePath,
 } from "./library/content.js";
 import { representative } from "./library/records.js";
-import type { Collection, EnteFile } from "./model/types.js";
+import { extractImageMetadata } from "./metadata-backup.js";
+import type { MLData } from "./mldata-fetch.js";
+import type { Collection, EnteFile, FileMetadata } from "./model/types.js";
 
 export type ProgressCallback = (message: string) => void;
 
@@ -76,7 +104,14 @@ export interface BackupOptions {
     includeThumbnails?: boolean;
     // Restrict the backup to albums with these names; others are left untouched.
     onlyAlbumNames?: string[];
+    // Hash each original already at its save path, and fetch again any whose
+    // bytes do not match the content hash its metadata records. Default false.
+    verify?: boolean;
     onProgress?: ProgressCallback;
+    // The caller already holds the lock in `downloadDirectory`, taken with
+    // `lockBackupDirectory`, and releases it itself, so the backup does not
+    // take it. `quak backup` takes it before it opens its library.
+    lockHeld?: boolean;
 }
 
 export interface BackupError {
@@ -93,6 +128,13 @@ export interface BackupResult {
     downloaded: number;
     // Originals already at their save path and left untouched.
     skipped: number;
+    // With `verify`, the originals already at their save path whose hash
+    // matched, those whose hash did not (each removed and fetched again), and
+    // those whose metadata records no hash (left as they are). All three are
+    // zero without `verify`.
+    verified: number;
+    mismatched: number;
+    unchecked: number;
     // Files with an unresolved failure after this run (the ledger size); the
     // CLI exits non-zero while this is above zero. A file can be both
     // downloaded and failed if its bytes landed but its symlink did not.
@@ -104,6 +146,8 @@ export interface BackupResult {
 // The slice of the library that backup drives. `Library` implements it; a test
 // can drive backup with a stand-in.
 export interface BackupLibrary {
+    // The account the library belongs to.
+    whoami(): { email: string; userID: number };
     refresh(): Promise<void>;
     listCollections(): Collection[];
     listFiles(collectionID: number): EnteFile[];
@@ -117,6 +161,13 @@ export interface BackupLibrary {
         destination: string,
     ): Promise<{ path: string; videoPath?: string }>;
     thumbnail(fileID: number): Promise<{ path: string }>;
+    // Wait for an ML data fetch to complete, joining one already running or
+    // starting one. Rejects with the reason when it fails; resolves at once
+    // when the library cannot fetch ML data.
+    fetchMLData(): Promise<void>;
+    // A file's cached ML data, as `lib.mldata.forFile()` returns it, or
+    // undefined when none is cached.
+    mlData(fileID: number): Promise<MLData | undefined>;
 }
 
 type FailureClass = "transient" | "permanent" | "unknown";
@@ -340,33 +391,126 @@ const saveLedger = (path: string, ledger: Map<number, FailureEntry>): void => {
     );
 };
 
-const writeSidecar = (path: string, file: EnteFile): void => {
+// The content hash of the original stored at `stored`, computed as the download
+// check computes it: over each file's bytes, read in chunks, and for a live
+// photo `<imageHash>:<videoHash>`.
+const storedHash = async (stored: {
+    path: string;
+    videoPath?: string;
+}): Promise<string> => {
+    await init();
+    const hashFile = async (path: string): Promise<string> => {
+        const state = chunkHashInit();
+        for await (const chunk of createReadStream(path)) {
+            chunkHashUpdate(state, chunk as Buffer);
+        }
+        return chunkHashFinal(state);
+    };
+    const hash = await hashFile(stored.path);
+    if (stored.videoPath === undefined) return hash;
+    return `${hash}:${await hashFile(stored.videoPath)}`;
+};
+
+// A file's EXIF, XMP and dimensions as its JSON holds them: what
+// `extractImageMetadata` found in its original, or why the original could not
+// be read.
+interface ImageMetadata {
+    imageMetadata?: Record<string, unknown>;
+    imageMetadataError?: string;
+}
+
+// The image metadata for the file whose original is at `originalPath` (for a
+// live photo, its image) and whose JSON is at `jsonPath`. A video gets none,
+// as `photo.exif()` reads none. An original stored before this run is not read
+// again when its JSON already holds image metadata: that is kept. A failed
+// read gives the reason, and fails neither the file nor the run. An original
+// with no EXIF, XMP or JPEG dimensions gets `{}`, so it is not read again.
+const imageMetadataFor = async (
+    file: EnteFile,
+    originalPath: string,
+    jsonPath: string,
+    storedThisRun: boolean,
+): Promise<ImageMetadata> => {
+    if (file.metadata.fileType === "video") return {};
+    if (!storedThisRun) {
+        try {
+            const { imageMetadata, imageMetadataError } = JSON.parse(
+                readFileSync(jsonPath, "utf-8"),
+            ) as ImageMetadata;
+            if (imageMetadata !== undefined || imageMetadataError !== undefined)
+                return { imageMetadata, imageMetadataError };
+        } catch {
+            // No JSON yet, or one that cannot be parsed: read the original.
+        }
+    }
+    try {
+        const bytes = await readFile(originalPath);
+        return { imageMetadata: extractImageMetadata(bytes) ?? {} };
+    } catch (err) {
+        return { imageMetadataError: errorMessage(err) };
+    }
+};
+
+// The file's JSON: its basic fields, its magic metadata, its ML data or the
+// reason the ML data is missing, and its image metadata.
+const writeSidecar = (
+    path: string,
+    file: EnteFile,
+    ml: { mlData?: MLData; mlDataError?: string },
+    image: ImageMetadata,
+): void => {
     const meta: Record<string, unknown> = {
         id: file.id,
         collectionID: file.collectionID,
         ownerID: file.ownerID,
         metadata: file.metadata,
+        updationTime: file.updationTime,
     };
     if (file.magicMetadata) meta.magicMetadata = file.magicMetadata;
     if (file.pubMagicMetadata) meta.pubMagicMetadata = file.pubMagicMetadata;
+    if (ml.mlData) meta.mlData = ml.mlData;
+    if (ml.mlDataError) meta.mlDataError = ml.mlDataError;
+    if (image.imageMetadata) meta.imageMetadata = image.imageMetadata;
+    if (image.imageMetadataError) {
+        meta.imageMetadataError = image.imageMetadataError;
+    }
     writeFileSync(path, JSON.stringify(meta, null, 2));
 };
 
-export const runBackup = async (
+// The album's JSON: its basic fields, its magic metadata, and its files.
+const writeAlbumJSON = (
+    path: string,
+    c: Collection,
+    files: { id: number; metadata: FileMetadata }[],
+): void => {
+    const album: Record<string, unknown> = {
+        id: c.id,
+        name: c.name,
+        type: c.type,
+        ownerID: c.ownerID,
+        isShared: c.isShared,
+        updationTime: c.updationTime,
+    };
+    if (c.magicMetadata) album.magicMetadata = c.magicMetadata;
+    if (c.pubMagicMetadata) album.pubMagicMetadata = c.pubMagicMetadata;
+    if (c.sharedMagicMetadata) {
+        album.sharedMagicMetadata = c.sharedMagicMetadata;
+    }
+    album.files = files;
+    writeFileSync(path, JSON.stringify(album, null, 2));
+};
+
+// The backup itself, which `runBackup` below runs while the lock is held.
+const runLockedBackup = async (
     lib: BackupLibrary,
     opts: BackupOptions,
+    downloadDirectory: string,
 ): Promise<BackupResult> => {
-    const downloadDirectory = opts.downloadDirectory;
-    if (!downloadDirectory) {
-        throw new Error(
-            "backup requires a downloadDirectory (pass one to backup() or " +
-                "open the library with one)",
-        );
-    }
     const includeOriginals = opts.includeOriginals ?? true;
     const includeThumbnails = opts.includeThumbnails ?? false;
     const log = opts.onProgress ?? (() => {});
     const only = opts.onlyAlbumNames ? new Set(opts.onlyAlbumNames) : undefined;
+    const verify = opts.verify ?? false;
 
     log("Refreshing library...");
     await lib.refresh();
@@ -374,6 +518,11 @@ export const runBackup = async (
     const collectionsDir = join(downloadDirectory, "collections");
     const thumbnailsDir = join(downloadDirectory, "thumbnails");
     mkdirSync(collectionsDir, { recursive: true });
+    const { email, userID } = lib.whoami();
+    writeFileSync(
+        join(downloadDirectory, "account.json"),
+        JSON.stringify({ email, userID }, null, 2),
+    );
     if (includeThumbnails) mkdirSync(thumbnailsDir, { recursive: true });
     removeLeftoverTempFiles(thumbnailsDir);
     for (const dir of dateFolders(downloadDirectory)) {
@@ -417,8 +566,12 @@ export const runBackup = async (
 
     const errors: BackupError[] = [];
     const failedThisRun = new Set<number>();
+    const storedThisRun = new Set<number>();
     let downloaded = 0;
     let skipped = 0;
+    let verified = 0;
+    let mismatched = 0;
+    let unchecked = 0;
 
     const recordFailure = (
         file: EnteFile,
@@ -450,10 +603,47 @@ export const runBackup = async (
 
     // Phase 1: get the bytes. Put each pending original at its save path
     // through the content cache/pools, as `Photo.download()` does, and fetch
-    // the optional thumbnails; a present file is left as is.
+    // the optional thumbnails; a present file is left as is. With `verify`, a
+    // present original is hashed first, and one that does not match the hash
+    // its metadata records is removed and fetched like a missing one. One
+    // that cannot be read is recorded as failed and left where it is.
     if (includeOriginals) {
         for (const [fileID, file] of distinct) {
-            if (storedAtSavePath(downloadDirectory, file) !== undefined) {
+            let stored = storedAtSavePath(downloadDirectory, file);
+            let mismatch = false;
+            if (stored !== undefined && verify) {
+                try {
+                    if (file.metadata.hash === undefined) {
+                        unchecked++;
+                    } else if (
+                        (await storedHash(stored)) === file.metadata.hash
+                    ) {
+                        verified++;
+                    } else {
+                        log(
+                            `MISMATCH original ${file.metadata.title} (${fileID}): its bytes do not match its content hash`,
+                        );
+                        mismatched++;
+                        mismatch = true;
+                        rmSync(stored.path);
+                        if (stored.videoPath !== undefined) {
+                            rmSync(stored.videoPath);
+                        }
+                        stored = undefined;
+                    }
+                } catch (err) {
+                    log(
+                        `FAILED verifying original ${file.metadata.title}: ${errorMessage(err)}`,
+                    );
+                    recordFailure(
+                        file,
+                        collectionName.get(file.collectionID) ?? "",
+                        err,
+                    );
+                    continue;
+                }
+            }
+            if (stored !== undefined) {
                 skipped++;
                 continue;
             }
@@ -462,9 +652,24 @@ export const runBackup = async (
                 // A fetched original is written straight to its save path (a
                 // live photo beside it); only one that was already cached
                 // elsewhere is copied.
-                await placeOriginal(downloadDirectory, file, (dest) =>
-                    lib.original(fileID, dest),
+                const placed = await placeOriginal(
+                    downloadDirectory,
+                    file,
+                    (dest) => lib.original(fileID, dest),
                 );
+                storedThisRun.add(fileID);
+                // A copy from the cache is not checked as a download is and can
+                // hold the same bad bytes, so what is put back after a
+                // mismatch is hashed too. A bad copy stays where it is and the
+                // file fails.
+                if (
+                    mismatch &&
+                    (await storedHash(placed)) !== file.metadata.hash
+                ) {
+                    throw new Error(
+                        "the original put back does not match its content hash either",
+                    );
+                }
                 downloaded++;
             } catch (err) {
                 log(
@@ -497,12 +702,43 @@ export const runBackup = async (
     }
 
     // Phase 2: rebuild the derived views from the model. Sidecars first, for
-    // every present original (this repairs stale ones).
+    // every present original (this repairs stale ones), each with the file's
+    // ML data once an ML data fetch has completed, and its image metadata.
+    // When the fetch fails, a file with no cached ML data gets the reason
+    // instead and is recorded as failed. The next run fetches its ML data
+    // again because none is cached.
     if (includeOriginals) {
+        let mlDataError: string | undefined;
+        try {
+            log("Fetching ML data...");
+            await lib.fetchMLData();
+        } catch (err) {
+            mlDataError = errorMessage(err);
+            log(`FAILED ML data: ${mlDataError}`);
+        }
         for (const file of distinct.values()) {
-            if (storedAtSavePath(downloadDirectory, file) !== undefined) {
-                const path = savePath(downloadDirectory, file);
-                writeSidecar(withExtension(path, ".json"), file);
+            const stored = storedAtSavePath(downloadDirectory, file);
+            if (stored === undefined) continue;
+            const path = withExtension(
+                savePath(downloadDirectory, file),
+                ".json",
+            );
+            const image = await imageMetadataFor(
+                file,
+                stored.path,
+                path,
+                storedThisRun.has(file.id),
+            );
+            const mlData = await lib.mlData(file.id);
+            if (mlData === undefined && mlDataError !== undefined) {
+                writeSidecar(path, file, { mlDataError }, image);
+                recordFailure(
+                    file,
+                    collectionName.get(file.collectionID) ?? "",
+                    new Error(`ML data: ${mlDataError}`),
+                );
+            } else {
+                writeSidecar(path, file, { mlData }, image);
             }
         }
     }
@@ -574,13 +810,10 @@ export const runBackup = async (
             }
         }
 
-        writeFileSync(
+        writeAlbumJSON(
             join(collectionsDir, `${colDirName}.json`),
-            JSON.stringify(
-                { id: c.id, name: c.name, type: c.type, files: metaFiles },
-                null,
-                2,
-            ),
+            c,
+            metaFiles,
         );
     }
 
@@ -599,7 +832,58 @@ export const runBackup = async (
         totalFiles: distinct.size,
         downloaded,
         skipped,
+        verified,
+        mismatched,
+        unchecked,
         failed: ledger.size,
         errors,
     };
 };
+
+// Only one backup of a directory runs at a time, in this process or another.
+// Creates `downloadDirectory` if it is missing, takes the lock in it and
+// returns the function that releases it; while another backup holds the lock,
+// fails at once with an error whose `code` is `ELOCKED`. The lock is the
+// directory `backup.lock`, whose modification time proper-lockfile keeps
+// current while it is held. One it has not touched for 10 seconds was left by
+// a run that could not remove it, and is taken over.
+export const lockBackupDirectory = async (
+    downloadDirectory: string,
+): Promise<() => Promise<void>> => {
+    mkdirSync(downloadDirectory, { recursive: true });
+    try {
+        return await lockfile.lock(downloadDirectory, {
+            lockfilePath: join(downloadDirectory, "backup.lock"),
+        });
+    } catch (err) {
+        if ((err as NodeJS.ErrnoException).code !== "ELOCKED") throw err;
+        throw Object.assign(
+            new Error(`another backup of ${downloadDirectory} is running`),
+            { code: "ELOCKED" },
+        );
+    }
+};
+
+// Runs the backup holding the lock, which it takes and releases itself unless
+// the caller already holds it (`opts.lockHeld`).
+export const runBackup = async (
+    lib: BackupLibrary,
+    opts: BackupOptions,
+): Promise<BackupResult> => {
+    const downloadDirectory = opts.downloadDirectory;
+    if (!downloadDirectory) {
+        throw new Error(
+            "backup requires a downloadDirectory (pass one to backup() or " +
+                "open the library with one)",
+        );
+    }
+    if (opts.lockHeld) {
+        return runLockedBackup(lib, opts, downloadDirectory);
+    }
+    const release = await lockBackupDirectory(downloadDirectory);
+    try {
+        return await runLockedBackup(lib, opts, downloadDirectory);
+    } finally {
+        await release();
+    }
+};

src/cli-commands.ts

@@ -20,6 +20,7 @@ import {
     type ClientSnapshot,
     type LoginOptions,
 } from "./client.js";
+import { lockBackupDirectory } from "./backup.js";
 import { init } from "./crypto/index.js";
 import {
     defaultCacheDirectory,
@@ -401,13 +402,24 @@ export const backupMetadataCommand = async (
 export const backupCommand = async (
     ctx: CliContext,
     dir: string,
-    opts: { json?: boolean },
+    opts: { json?: boolean; verify?: boolean },
 ): Promise<number> => {
     await init();
     const client = requireSession(ctx);
     if (!client) return 3;
 
     ctx.stderr.write("Starting backup...\n");
+    // The lock is taken before the library opens and starts its refresh, so a
+    // run refused while another backup of `dir` runs sends no request.
+    let release: () => Promise<void>;
+    try {
+        release = await lockBackupDirectory(dir);
+    } catch (err) {
+        if ((err as NodeJS.ErrnoException).code !== "ELOCKED") throw err;
+        ctx.stderr.write(`quak: ${(err as Error).message}\n`);
+        return 2;
+    }
+    try {
         // The precache is off: the backup fetches what it needs, and must not
         // also fill the cache with every thumbnail and the recent originals.
         const lib = await Library.open({
@@ -420,6 +432,8 @@ export const backupCommand = async (
         try {
             const result = await lib.backup({
                 downloadDirectory: dir,
+                lockHeld: true,
+                verify: opts.verify,
                 onProgress: (msg) => {
                     if (!opts.json) ctx.stderr.write(msg + "\n");
                 },
@@ -432,6 +446,11 @@ export const backupCommand = async (
                 ctx.stderr.write(`  Total files: ${result.totalFiles}\n`);
                 ctx.stderr.write(`  Downloaded:  ${result.downloaded}\n`);
                 ctx.stderr.write(`  Skipped:     ${result.skipped}\n`);
+                if (opts.verify) {
+                    ctx.stderr.write(`  Verified:    ${result.verified}\n`);
+                    ctx.stderr.write(`  Mismatched:  ${result.mismatched}\n`);
+                    ctx.stderr.write(`  Unchecked:   ${result.unchecked}\n`);
+                }
                 ctx.stderr.write(`  Failed:      ${result.failed}\n`);
                 if (result.errors.length > 0) {
                     ctx.stderr.write("\nFailed files:\n");
@@ -447,6 +466,9 @@ export const backupCommand = async (
         } finally {
             await lib.close();
         }
+    } finally {
+        await release();
+    }
 };
 
 export const listMissingThumbnailsCommand = async (

src/index.ts

@@ -27,6 +27,7 @@ export {
     isRetryable,
     isSafeToReplay,
     resolveRetryOptions,
+    UNATTENDED_RETRY_OPTIONS,
     withRetry,
     type ResolvedRetryOptions,
     type RetryOptions,
@@ -66,6 +67,7 @@ export {
     type EnsureOptions,
     type EnsureResult,
     type EnsureEvent,
+    lockBackupDirectory,
     runBackup,
     type BackupOptions,
     type BackupResult,

src/library/index.ts

@@ -94,6 +94,7 @@ import type { Collection, EnteFile } from "../model/types.js";
 import { runBackup, type BackupOptions, type BackupResult } from "../backup.js";
 
 export {
+    lockBackupDirectory,
     runBackup,
     type BackupOptions,
     type BackupResult,
@@ -279,7 +280,8 @@ export class Library {
     private cycle?: Promise<void>;
     // Guards the ML fetch pass so a slow backfill never runs twice at once; a
     // refresh whose pass is still running kicks nothing new. Holds the running
-    // pass, so `close()` can wait for it.
+    // pass, so `close()` and `backup()` can wait for it. It rejects when the
+    // pass fails.
     private mlFetch?: Promise<void>;
     private closed = false;
     private lastRefreshAt?: number;
@@ -565,12 +567,17 @@ export class Library {
 
     // Back up every in-scope file to `opts.downloadDirectory`, or else the
     // library's, each original at its save path, with a durable failure
-    // ledger (issue #51). Waits for a completed refresh first, as `fresh()`
-    // does, joining one already running, and rejects before touching any file
-    // when it fails. Then puts pending originals at their save paths as
-    // `Photo.download()` does (and optional thumbnails) through the content
-    // cache and pools, and rebuilds the derived symlink/JSON views from the
-    // model. Throws before any network work when no content cache backs the
+    // ledger (issue #51). Takes the lock in that directory first, unless
+    // `opts.lockHeld` says the caller holds it, failing at once while another
+    // backup of it runs (see `runBackup`). Waits for a
+    // completed refresh, as `fresh()` does, joining one already running, and
+    // rejects before touching any file but the lock when it fails. Then puts
+    // pending originals at their save paths as `Photo.download()` does (and
+    // optional thumbnails) through the content cache and pools, waits for an
+    // ML data fetch, and rebuilds the derived symlink/JSON views from the
+    // model, each file's JSON with its ML data, beside an `account.json` with
+    // the account's email and user ID.
+    // Throws before any network work when no content cache backs the
     // originals it must fetch.
     backup(opts?: BackupOptions): Promise<BackupResult> {
         const downloadDirectory =
@@ -587,12 +594,15 @@ export class Library {
         const cache = this.cache;
         return runBackup(
             {
+                whoami: () => this.client.whoami(),
                 refresh: () => this.refreshNow(),
                 listCollections: () => this.store.listCollections(),
                 listFiles: (id) => this.store.listFiles(id),
                 original: (fileID, destination) =>
                     cache!.backupOriginal(fileID, destination),
                 thumbnail: (fileID) => cache!.thumbnail(fileID),
+                fetchMLData: () => this.fetchMLDataNow(),
+                mlData: (fileID) => this.mldata.forFile({ fileID }),
             },
             { ...opts, downloadDirectory },
         );
@@ -612,7 +622,7 @@ export class Library {
             this.timer = undefined;
         }
         await this.cycle?.catch(() => {});
-        await this.mlFetch;
+        await this.mlFetch?.catch(() => {});
         await precacheClosed;
     }
 
@@ -676,10 +686,9 @@ export class Library {
             // Backfill ML data for the files this refresh knows about. It runs
             // outside the refresh's success/failure so a fetch or disk problem
             // there never marks the metadata refresh failed, and it is not
-            // awaited so it never stalls the refresh interval.
-            this.mlFetch ??= this.runMLFetch().finally(() => {
-                this.mlFetch = undefined;
-            });
+            // awaited so it never stalls the refresh interval. Its failure is
+            // reported through `status()` and `onProgress`.
+            void this.fetchMLDataNow().catch(() => {});
         } catch (err) {
             const error = err instanceof Error ? err.message : String(err);
             this.lastError = error;
@@ -784,10 +793,19 @@ export class Library {
         }
     }
 
+    // Join the running ML fetch pass, or start one when none runs. Resolves at
+    // once when the client cannot fetch ML data; rejects when the pass fails.
+    private fetchMLDataNow(): Promise<void> {
+        this.mlFetch ??= this.runMLFetch().finally(() => {
+            this.mlFetch = undefined;
+        });
+        return this.mlFetch;
+    }
+
     // One ML fetch pass: fetch, decrypt and store the ML data for every file
     // the store knows about that is not cached (or whose `updationTime` has
-    // advanced), through the metadata pool, and update the CLIP index. Guarded
-    // so passes never overlap; a failure is reported, not thrown.
+    // advanced), through the metadata pool, and update the CLIP index. A
+    // failure is reported through `status()` and `onProgress`, then thrown.
     private async runMLFetch(): Promise<void> {
         const mldata = this.mlStore;
         // Bind so the call keeps the client as its receiver when invoked
@@ -831,6 +849,7 @@ export class Library {
             const error = err instanceof Error ? err.message : String(err);
             this.lastMLError = error;
             this.emit({ operation: "fetchMLData", status: "failed", error });
+            throw err;
         }
     }
 

src/retry.ts

@@ -37,6 +37,16 @@ export const DEFAULT_RETRY_OPTIONS: ResolvedRetryOptions = {
     random: Math.random,
 };
 
+// For a run nobody is watching, such as `quak backup` from cron. A request that
+// keeps failing waits at most 243 s in all before it gives up, usually about
+// half that, since each wait is drawn at random below its ceiling.
+export const UNATTENDED_RETRY_OPTIONS: ResolvedRetryOptions = {
+    ...DEFAULT_RETRY_OPTIONS,
+    attempts: 10,
+    baseDelayMs: 1_000,
+    maxDelayMs: 60_000,
+};
+
 export const resolveRetryOptions = (
     opts?: RetryOptions,
 ): ResolvedRetryOptions => ({

test/cli/backup.test.ts

@@ -12,6 +12,7 @@
  *         collections/
  *             <name>/<title>          symlink to the original (rebuilt each run)
  *             <name>.json             per-collection metadata (rebuilt each run)
+ *         account.json                the account's email and user ID
  *         failures.json               durable ledger of unresolved failures
  *
  * The properties that distinguish backup from a naive download loop, and that
@@ -32,6 +33,7 @@
  */
 
 import {
+    chmodSync,
     existsSync,
     lstatSync,
     mkdirSync,
@@ -41,6 +43,7 @@ import {
     readlinkSync,
     rmSync,
     symlinkSync,
+    utimesSync,
     writeFileSync,
 } from "node:fs";
 import { spawnSync } from "node:child_process";
@@ -52,11 +55,17 @@ import { runBackup, type BackupLibrary } from "../../src/backup.js";
 import { Library } from "../../src/library/index.js";
 import type { ContentSource } from "../../src/library/content.js";
 import type { CollectionsPage, FilesPage } from "../../src/client.js";
+import { extractImageMetadata } from "../../src/metadata-backup.js";
+import type { MLData } from "../../src/mldata-fetch.js";
 import type { Collection, EnteFile } from "../../src/model/types.js";
+import { HEIC_WITH_EXIF } from "../exif-heic.js";
+import { JPEG_WITH_EXIF } from "../exif-jpeg.js";
 import {
     asLivePhoto,
+    blake2b,
     cdnSource,
     IMAGE,
+    livePhotoHash,
     livePhotoZip,
     VIDEO,
 } from "../live-photo.js";
@@ -823,7 +832,617 @@ describe("the refresh before a backup", () => {
         );
 
         expect(source.originalCalls).toBe(0);
-        expect(existsSync(outDir)).toBe(false);
+        // The backup made the directory for its lock, and removed the lock.
+        expect(readdirSync(outDir)).toEqual([]);
+        await lib.close();
+    });
+});
+
+describe("the backup lock", () => {
+    const lockPath = (outDir: string): string => join(outDir, "backup.lock");
+
+    it("refuses a second backup of the directory while one runs", async () => {
+        await fillCache();
+        const client = new HeldClient();
+        const lib = await openLibrary(stubSource(), client);
+        const outDir = join(root, "backup");
+        // The first backup holds the lock once it starts its refresh, which
+        // `HeldClient` keeps from finishing.
+        let refreshing!: () => void;
+        const started = new Promise<void>((resolve) => {
+            refreshing = resolve;
+        });
+        const first = lib.backup({
+            downloadDirectory: outDir,
+            onProgress: (msg) => {
+                if (msg === "Refreshing library...") refreshing();
+            },
+        });
+        await started;
+
+        await expect(lib.backup({ downloadDirectory: outDir })).rejects.toThrow(
+            `another backup of ${outDir} is running`,
+        );
+
+        client.release();
+        expect((await first).failed).toBe(0);
+        await lib.close();
+    });
+
+    it("releases the lock after a backup succeeds", async () => {
+        const lib = await openLibrary(stubSource());
+        const outDir = join(root, "backup");
+
+        await lib.backup({ downloadDirectory: outDir });
+
+        expect(existsSync(lockPath(outDir))).toBe(false);
+        // So the next backup of the directory runs.
+        expect((await lib.backup({ downloadDirectory: outDir })).skipped).toBe(
+            3,
+        );
+        await lib.close();
+    });
+
+    it("releases the lock after a backup fails", async () => {
+        const lib = await openLibrary(stubSource(), new FailingClient());
+        const outDir = join(root, "backup");
+
+        await expect(lib.backup({ downloadDirectory: outDir })).rejects.toThrow(
+            "HTTP 401 from server",
+        );
+
+        expect(existsSync(lockPath(outDir))).toBe(false);
+        await lib.close();
+    });
+
+    it("takes over a lock left by a run that was killed", async () => {
+        const outDir = join(root, "backup");
+        // A killed run's lock, last touched a minute ago.
+        mkdirSync(lockPath(outDir), { recursive: true });
+        const minuteAgo = new Date(Date.now() - 60_000);
+        utimesSync(lockPath(outDir), minuteAgo, minuteAgo);
+        const lib = await openLibrary(stubSource());
+
+        const result = await lib.backup({ downloadDirectory: outDir });
+
+        expect(result.downloaded).toBe(3);
+        expect(existsSync(lockPath(outDir))).toBe(false);
+        await lib.close();
+    });
+});
+
+// Also serves ML data: file 100 has `ML_PAYLOAD`, the other two have none.
+// While `mlError` is set, every ML data request fails with it.
+const ML_PAYLOAD = { face: { faces: [] }, clip: { embedding: [0.5, 0.25] } };
+
+class MLClient extends MockClient {
+    mlError?: string;
+    async fetchMLData(args: {
+        fileIDs: number[];
+    }): Promise<Map<number, MLData>> {
+        if (this.mlError) throw new Error(this.mlError);
+        const result = new Map<number, MLData>();
+        if (args.fileIDs.includes(100)) result.set(100, ML_PAYLOAD);
+        return result;
+    }
+}
+
+// The JSON the backup in `outDir` wrote beside the original of `fileID`.
+const fileJSON = (outDir: string, fileID: number): Record<string, unknown> =>
+    JSON.parse(readFileSync(saved(outDir, `${fileID}.json`), "utf-8"));
+
+describe("ML data in each file's JSON", () => {
+    it("writes a file's ML data, and no ML field for a file that has none", async () => {
+        const lib = await openLibrary(stubSource(), new MLClient());
+        const outDir = join(root, "backup");
+
+        const result = await lib.backup({ downloadDirectory: outDir });
+
+        expect(result.failed).toBe(0);
+        expect(fileJSON(outDir, 100).mlData).toEqual(ML_PAYLOAD);
+        expect(fileJSON(outDir, 101)).not.toHaveProperty("mlData");
+        expect(fileJSON(outDir, 101)).not.toHaveProperty("mlDataError");
+        await lib.close();
+    });
+
+    it("gives each file with no cached ML data the reason when the fetch fails, and fails the run", async () => {
+        const client = new MLClient();
+        const outDir = join(root, "backup");
+        const first = await openLibrary(stubSource(), client);
+        await first.backup({ downloadDirectory: outDir });
+        await first.close();
+
+        client.mlError = "HTTP 503 from server";
+        const lib = await openLibrary(stubSource(), client);
+        const result = await lib.backup({ downloadDirectory: outDir });
+
+        // File 100's ML data was cached by the first run and is kept.
+        expect(fileJSON(outDir, 100).mlData).toEqual(ML_PAYLOAD);
+        expect(fileJSON(outDir, 100)).not.toHaveProperty("mlDataError");
+        for (const fileID of [101, 200]) {
+            expect(fileJSON(outDir, fileID).mlDataError).toBe(
+                "HTTP 503 from server",
+            );
+        }
+        expect(result.failed).toBe(2);
+        expect(result.errors.map((e) => [e.fileID, e.error])).toEqual([
+            [101, "ML data: HTTP 503 from server"],
+            [200, "ML data: HTTP 503 from server"],
+        ]);
+        expect(Object.keys(readLedger(outDir).files)).toEqual(["101", "200"]);
+        await lib.close();
+    });
+
+    it("writes the ML data on the run after a failed fetch", async () => {
+        const client = new MLClient();
+        client.mlError = "HTTP 503 from server";
+        const outDir = join(root, "backup");
+        const first = await openLibrary(stubSource(), client);
+        expect((await first.backup({ downloadDirectory: outDir })).failed).toBe(
+            3,
+        );
+        expect(fileJSON(outDir, 100)).not.toHaveProperty("mlData");
+        await first.close();
+
+        client.mlError = undefined;
+        const lib = await openLibrary(stubSource(), client);
+        const result = await lib.backup({ downloadDirectory: outDir });
+
+        expect(result.failed).toBe(0);
+        expect(result.skipped).toBe(3);
+        expect(fileJSON(outDir, 100).mlData).toEqual(ML_PAYLOAD);
+        for (const fileID of [100, 101, 200]) {
+            expect(fileJSON(outDir, fileID)).not.toHaveProperty("mlDataError");
+        }
+        expect(existsSync(join(outDir, "failures.json"))).toBe(false);
+        await lib.close();
+    });
+});
+
+// The fields `RecordsClient` gives the Vacation album: shared by another
+// account, with all three layers of magic metadata.
+const SHARED_ALBUM = {
+    ownerID: 7,
+    isShared: true,
+    updationTime: 1700,
+    magicMetadata: { visibility: 0 },
+    pubMagicMetadata: { coverID: 100 },
+    sharedMagicMetadata: { visibility: 2 },
+};
+
+// Serves the Vacation album with `SHARED_ALBUM`'s fields, and each file with
+// the update time 5000 + its ID.
+class RecordsClient extends MockClient {
+    override async collectionsSince(): Promise<CollectionsPage> {
+        const page = await super.collectionsSince();
+        return {
+            ...page,
+            collections: page.collections.map((c) =>
+                c.id === 1 ? { ...c, ...SHARED_ALBUM } : c,
+            ),
+        };
+    }
+    override async filesSince(args: {
+        collectionID: number;
+    }): Promise<FilesPage> {
+        const page = await super.filesSince(args);
+        return {
+            ...page,
+            files: page.files.map((f) => ({
+                ...f,
+                updationTime: 5000 + f.id,
+            })),
+        };
+    }
+}
+
+// The JSON the backup in `outDir` wrote for the album named `name`.
+const albumJSON = (outDir: string, name: string): Record<string, unknown> =>
+    JSON.parse(
+        readFileSync(join(outDir, "collections", `${name}.json`), "utf-8"),
+    );
+
+describe("account and album records", () => {
+    it("writes account.json with the account's email and user ID", async () => {
+        const lib = await openLibrary(stubSource());
+        const outDir = join(root, "backup");
+
+        await lib.backup({ downloadDirectory: outDir });
+
+        expect(
+            JSON.parse(readFileSync(join(outDir, "account.json"), "utf-8")),
+        ).toEqual({ email: "backup@example.com", userID: USER_ID });
+        await lib.close();
+    });
+
+    it("writes each album's owner, sharing, update time and magic metadata into its JSON", async () => {
+        const lib = await openLibrary(stubSource(), new RecordsClient());
+        const outDir = join(root, "backup");
+
+        await lib.backup({ downloadDirectory: outDir });
+
+        expect(albumJSON(outDir, "Vacation")).toEqual({
+            id: 1,
+            name: "Vacation",
+            type: "album",
+            ...SHARED_ALBUM,
+            files: [
+                { id: 100, metadata: file(100, 1, "beach.jpg").metadata },
+                { id: 101, metadata: file(101, 1, "sunset.jpg").metadata },
+            ],
+        });
+        // An album with no magic metadata gets no magic metadata fields.
+        expect(albumJSON(outDir, "Work")).toEqual({
+            id: 2,
+            name: "Work",
+            type: "album",
+            ownerID: USER_ID,
+            isShared: false,
+            updationTime: 1,
+            files: [
+                { id: 200, metadata: file(200, 2, "diagram.png").metadata },
+            ],
+        });
+        await lib.close();
+    });
+
+    it("writes each file's update time into its JSON", async () => {
+        const lib = await openLibrary(stubSource(), new RecordsClient());
+        const outDir = join(root, "backup");
+
+        await lib.backup({ downloadDirectory: outDir });
+
+        for (const fileID of [100, 101, 200]) {
+            expect(fileJSON(outDir, fileID).updationTime).toBe(5000 + fileID);
+        }
+        await lib.close();
+    });
+});
+
+// What a file's JSON holds as `imageMetadata` for an original of `bytes`.
+const imageMetadataOf = (bytes: Uint8Array): unknown =>
+    JSON.parse(JSON.stringify(extractImageMetadata(bytes)));
+
+describe("image metadata in each file's JSON", () => {
+    // Serves `files` in the Vacation album and nothing in Work.
+    class FilesClient extends MockClient {
+        constructor(private readonly files: EnteFile[]) {
+            super();
+        }
+        override async filesSince(args: {
+            collectionID: number;
+        }): Promise<FilesPage> {
+            return {
+                files: args.collectionID === 1 ? this.files : [],
+                deleted: [],
+                cursor: 1,
+            };
+        }
+    }
+
+    // Writes `originals.get(fileID)` as each file's original, looked up at
+    // each fetch, so a test can change it between runs.
+    const bytesSource = (
+        originals: Map<number, Uint8Array>,
+    ): ContentSource => ({
+        original: async ({ file: f, destination }) => {
+            const bytes = originals.get(f.id)!;
+            writeFileSync(destination, bytes);
+            return { bytesWritten: bytes.length };
+        },
+        thumbnail: async () => {
+            throw new Error("no thumbnails in this source");
+        },
+    });
+
+    // A JPEG, a HEIC, a video, and a file holding no image metadata. The
+    // video's bytes are the JPEG's, so reading it would find EXIF.
+    const clip = file(602, 1, "clip.mov");
+    const files = [
+        file(600, 1, "photo.jpg"),
+        file(601, 1, "photo.heic"),
+        { ...clip, metadata: { ...clip.metadata, fileType: "video" as const } },
+        file(603, 1, "notes.png"),
+    ];
+    const originals = (): Map<number, Uint8Array> =>
+        new Map([
+            [600, JPEG_WITH_EXIF],
+            [601, HEIC_WITH_EXIF],
+            [602, JPEG_WITH_EXIF],
+            [603, new TextEncoder().encode("not an image")],
+        ]);
+    const open = (bytes = originals()): Promise<Library> =>
+        openLibrary(bytesSource(bytes), new FilesClient(files));
+
+    // Rewrite the JSON of `fileID` without its image metadata, as an earlier
+    // version of quak wrote it.
+    const dropImageMetadata = (outDir: string, fileID: number): void => {
+        const json = fileJSON(outDir, fileID);
+        delete json.imageMetadata;
+        writeFileSync(saved(outDir, `${fileID}.json`), JSON.stringify(json));
+    };
+
+    it("writes each new original's image metadata, and none for a video", async () => {
+        const lib = await open();
+        const outDir = join(root, "backup");
+
+        const result = await lib.backup({ downloadDirectory: outDir });
+
+        expect(result).toMatchObject({ downloaded: 4, failed: 0 });
+        expect(fileJSON(outDir, 600).imageMetadata).toEqual(
+            imageMetadataOf(JPEG_WITH_EXIF),
+        );
+        expect(fileJSON(outDir, 601).imageMetadata).toEqual(
+            imageMetadataOf(HEIC_WITH_EXIF),
+        );
+        expect(fileJSON(outDir, 602)).not.toHaveProperty("imageMetadata");
+        expect(fileJSON(outDir, 602)).not.toHaveProperty("imageMetadataError");
+        // Empty, so that the next run does not read it again.
+        expect(fileJSON(outDir, 603).imageMetadata).toEqual({});
+        await lib.close();
+    });
+
+    it("reads a live photo's image", async () => {
+        const { file: live, body } = await asLivePhoto(
+            file(500, 1, "IMG_0500.HEIC"),
+            livePhotoZip({ "image.heic": HEIC_WITH_EXIF, "video.mov": VIDEO }),
+            livePhotoHash(HEIC_WITH_EXIF, VIDEO),
+        );
+        const lib = await openLibrary(
+            cdnSource(new Map([[500, body]])),
+            new FilesClient([live]),
+        );
+        const outDir = join(root, "backup");
+
+        await lib.backup({ downloadDirectory: outDir });
+
+        expect(fileJSON(outDir, 500).imageMetadata).toEqual(
+            imageMetadataOf(HEIC_WITH_EXIF),
+        );
+        await lib.close();
+    });
+
+    it("keeps the image metadata of an original stored by an earlier run without reading it again", async () => {
+        const lib = await open();
+        const outDir = join(root, "backup");
+        await lib.backup({ downloadDirectory: outDir });
+        // A read of the original now would give the HEIC's.
+        writeFileSync(saved(outDir, "600.jpg"), HEIC_WITH_EXIF);
+
+        const second = await lib.backup({ downloadDirectory: outDir });
+
+        expect(second).toMatchObject({ downloaded: 0, failed: 0 });
+        expect(fileJSON(outDir, 600).imageMetadata).toEqual(
+            imageMetadataOf(JPEG_WITH_EXIF),
+        );
+        await lib.close();
+    });
+
+    it("reads an original stored by an earlier version, whose JSON has no image metadata", async () => {
+        const lib = await open();
+        const outDir = join(root, "backup");
+        await lib.backup({ downloadDirectory: outDir });
+        dropImageMetadata(outDir, 600);
+
+        const second = await lib.backup({ downloadDirectory: outDir });
+
+        expect(second).toMatchObject({ downloaded: 0, failed: 0 });
+        expect(fileJSON(outDir, 600).imageMetadata).toEqual(
+            imageMetadataOf(JPEG_WITH_EXIF),
+        );
+        await lib.close();
+    });
+
+    it("reads an original the run stores again, though its JSON has image metadata", async () => {
+        const bytes = originals();
+        const lib = await open(bytes);
+        const outDir = join(root, "backup");
+        await lib.backup({ downloadDirectory: outDir });
+        rmSync(saved(outDir, "600.jpg"));
+        bytes.set(600, HEIC_WITH_EXIF);
+
+        const second = await lib.backup({ downloadDirectory: outDir });
+
+        expect(second).toMatchObject({ downloaded: 1, failed: 0 });
+        expect(fileJSON(outDir, 600).imageMetadata).toEqual(
+            imageMetadataOf(HEIC_WITH_EXIF),
+        );
+        await lib.close();
+    });
+
+    // Root ignores file permissions, so this fails when run as root. The
+    // `test` phase of the `Dockerfile` runs as the `node` user.
+    it("gives the reason an original could not be read, failing neither the file nor the run", async () => {
+        const lib = await open();
+        const outDir = join(root, "backup");
+        await lib.backup({ downloadDirectory: outDir });
+        dropImageMetadata(outDir, 600);
+        const original = saved(outDir, "600.jpg");
+        chmodSync(original, 0o000);
+
+        const second = await lib
+            .backup({ downloadDirectory: outDir })
+            .finally(() => chmodSync(original, 0o600));
+
+        expect(second).toMatchObject({ failed: 0, errors: [] });
+        expect(existsSync(join(outDir, "failures.json"))).toBe(false);
+        expect(fileJSON(outDir, 600)).not.toHaveProperty("imageMetadata");
+        expect(fileJSON(outDir, 600).imageMetadataError).toMatch(/EACCES/);
+        await lib.close();
+    });
+});
+
+// With `verify`, each original already stored is hashed, and one that does not
+// match the content hash its metadata records is downloaded again.
+describe("backup with verify", () => {
+    // MockClient's files, each recording the hash of what `stubSource` writes
+    // for it, except diagram.png (200), which records none.
+    class HashedClient extends MockClient {
+        override async filesSince(args: {
+            collectionID: number;
+        }): Promise<FilesPage> {
+            const page = await super.filesSince(args);
+            const files = page.files.map((f) =>
+                f.id === 200
+                    ? f
+                    : {
+                          ...f,
+                          metadata: {
+                              ...f.metadata,
+                              hash: blake2b(Buffer.alloc(SIZE_BY_ID[f.id]!)),
+                          },
+                      },
+            );
+            return { ...page, files };
+        }
+    }
+
+    // A backup of the account in `root/backup`, the library that made it, and
+    // the source it fetched from.
+    const backedUp = async (): Promise<{
+        lib: Library;
+        source: StubSource;
+        outDir: string;
+    }> => {
+        const source = stubSource();
+        const lib = await openLibrary(source, new HashedClient());
+        const outDir = join(root, "backup");
+        await lib.backup({ downloadDirectory: outDir });
+        return { lib, source, outDir };
+    };
+
+    // What `stubSource` writes for beach.jpg (100), and other bytes of the
+    // same length.
+    const good = Buffer.alloc(SIZE_BY_ID[100]!);
+    const corrupt = Buffer.alloc(SIZE_BY_ID[100]!, 1);
+
+    it("leaves an original that matches its hash, and one with no hash, as they are", async () => {
+        const { lib, source, outDir } = await backedUp();
+        const calls = source.originalCalls;
+
+        const result = await lib.backup({
+            downloadDirectory: outDir,
+            verify: true,
+        });
+
+        expect(result).toMatchObject({
+            downloaded: 0,
+            skipped: 3,
+            verified: 2,
+            mismatched: 0,
+            unchecked: 1,
+            failed: 0,
+        });
+        expect(source.originalCalls).toBe(calls);
+        await lib.close();
+    });
+
+    it("downloads again an original that does not match its hash", async () => {
+        const { lib, outDir } = await backedUp();
+        writeFileSync(saved(outDir, "100.jpg"), corrupt);
+        const log: string[] = [];
+
+        const result = await lib.backup({
+            downloadDirectory: outDir,
+            verify: true,
+            onProgress: (msg) => log.push(msg),
+        });
+
+        expect(result).toMatchObject({
+            downloaded: 1,
+            skipped: 2,
+            verified: 1,
+            mismatched: 1,
+            unchecked: 1,
+            failed: 0,
+        });
+        expect(readFileSync(saved(outDir, "100.jpg"))).toEqual(good);
+        expect(log.filter((msg) => msg.startsWith("MISMATCH"))).toEqual([
+            "MISMATCH original beach.jpg (100): its bytes do not match its content hash",
+        ]);
+        await lib.close();
+    });
+
+    it("records a failed download in failures.json, with the original removed", async () => {
+        const { lib, source, outDir } = await backedUp();
+        writeFileSync(saved(outDir, "100.jpg"), corrupt);
+        source.failID = 100;
+
+        const result = await lib.backup({
+            downloadDirectory: outDir,
+            verify: true,
+        });
+
+        expect(result).toMatchObject({
+            downloaded: 0,
+            mismatched: 1,
+            failed: 1,
+        });
+        expect(Object.keys(readLedger(outDir).files)).toEqual(["100"]);
+        expect(existsSync(saved(outDir, "100.jpg"))).toBe(false);
+        await lib.close();
+    });
+
+    it("records in failures.json an original the cache puts back with the same bad bytes", async () => {
+        const lib = await openLibrary(stubSource(), new HashedClient());
+        const outDir = join(root, "backup");
+        // The cache holds a bad copy, and a backup copies an original the
+        // cache holds to its save path.
+        const cached = await lib.photos.byID({ fileID: 100 })!.original();
+        writeFileSync(cached.path, corrupt);
+        await lib.backup({ downloadDirectory: outDir });
+
+        const result = await lib.backup({
+            downloadDirectory: outDir,
+            verify: true,
+        });
+
+        expect(result).toMatchObject({
+            downloaded: 0,
+            mismatched: 1,
+            failed: 1,
+        });
+        expect(result.errors.map((e) => e.error)).toEqual([
+            "the original put back does not match its content hash either",
+        ]);
+        expect(Object.keys(readLedger(outDir).files)).toEqual(["100"]);
+        expect(readFileSync(saved(outDir, "100.jpg"))).toEqual(corrupt);
+        await lib.close();
+    });
+
+    it("hashes nothing without verify", async () => {
+        const { lib, outDir } = await backedUp();
+        writeFileSync(saved(outDir, "100.jpg"), corrupt);
+
+        const result = await lib.backup({ downloadDirectory: outDir });
+
+        expect(result).toMatchObject({
+            downloaded: 0,
+            skipped: 3,
+            verified: 0,
+            mismatched: 0,
+            unchecked: 0,
+            failed: 0,
+        });
+        expect(readFileSync(saved(outDir, "100.jpg"))).toEqual(corrupt);
+        await lib.close();
+    });
+
+    // Root ignores file permissions, so this fails when run as root. The
+    // `test` phase of the `Dockerfile` runs as the `node` user.
+    it("records an original it cannot read as failed, and leaves it", async () => {
+        const { lib, outDir } = await backedUp();
+        const original = saved(outDir, "100.jpg");
+        chmodSync(original, 0o000);
+
+        const result = await lib
+            .backup({ downloadDirectory: outDir, verify: true })
+            .finally(() => chmodSync(original, 0o600));
+
+        expect(result).toMatchObject({ downloaded: 0, verified: 1, failed: 1 });
+        expect(result.errors.map((e) => e.fileID)).toEqual([100]);
+        expect(result.errors[0]!.error).toMatch(/EACCES/);
+        expect(readFileSync(original)).toEqual(good);
         await lib.close();
     });
 });
@@ -871,6 +1490,9 @@ describe("backup album folders", () => {
         thumbnail: async () => {
             throw new Error("no thumbnails in this stand-in");
         },
+        fetchMLData: async () => {},
+        mlData: async () => undefined,
+        whoami: () => ({ email: "backup@example.com", userID: USER_ID }),
     });
 
     const albumID = (outDir: string, jsonName: string): number =>
@@ -1320,6 +1942,64 @@ describe("backup of live photos", () => {
         },
     );
 
+    it("verifies a live photo's image and video together, and downloads it again when one does not match", async () => {
+        const { file: live, body } = await asLivePhoto(
+            file(500, 10, "IMG_0500.HEIC"),
+        );
+        const lib = await open([live], new Map([[500, body]]));
+        const outDir = join(root, "backup");
+        await lib.backup({ downloadDirectory: outDir });
+
+        const good = await lib.backup({
+            downloadDirectory: outDir,
+            verify: true,
+        });
+        expect(good).toMatchObject({ skipped: 1, verified: 1, mismatched: 0 });
+
+        const video = saved(outDir, "500.mov");
+        writeFileSync(video, "another few seconds of video");
+        const bad = await lib.backup({
+            downloadDirectory: outDir,
+            verify: true,
+        });
+
+        expect(bad).toMatchObject({
+            downloaded: 1,
+            verified: 0,
+            mismatched: 1,
+            failed: 0,
+        });
+        expect(readFileSync(saved(outDir, "500.heic"))).toEqual(
+            Buffer.from(IMAGE),
+        );
+        expect(readFileSync(video)).toEqual(Buffer.from(VIDEO));
+        expect(tree(outDir)).toEqual(linked);
+        await lib.close();
+    });
+
+    it("removes both files of a live photo that does not match its hash when downloading it again fails", async () => {
+        const { file: live, body } = await asLivePhoto(
+            file(500, 10, "IMG_0500.HEIC"),
+        );
+        const bodies = new Map([[500, body]]);
+        const lib = await open([live], bodies);
+        const outDir = join(root, "backup");
+        await lib.backup({ downloadDirectory: outDir });
+        writeFileSync(saved(outDir, "500.mov"), "another few seconds of video");
+        bodies.delete(500);
+
+        const result = await lib.backup({
+            downloadDirectory: outDir,
+            verify: true,
+        });
+
+        expect(result).toMatchObject({ mismatched: 1, failed: 1 });
+        expect(Object.keys(readLedger(outDir).files)).toEqual(["500"]);
+        expect(existsSync(saved(outDir, "500.heic"))).toBe(false);
+        expect(existsSync(saved(outDir, "500.mov"))).toBe(false);
+        await lib.close();
+    });
+
     it("serves a live photo the backup stored to a library reading the backup", async () => {
         const { file: live, body } = await asLivePhoto(
             file(500, 10, "IMG_0500.HEIC"),

test/cli/bin.test.ts

@@ -0,0 +1,71 @@
+/**
+ * Tests for the retry options `bin/quak.ts` loads the saved session with:
+ * `quak backup` gets the unattended ones, every other command the default.
+ *
+ * Each test runs `bin/quak.ts`, as the smoke test does, with its session loader
+ * replaced by one that records the options it is given and reports no session,
+ * so the command stops before it makes any request.
+ */
+
+import { mkdtempSync, rmSync } from "node:fs";
+import { tmpdir } from "node:os";
+import { join } from "node:path";
+import { afterAll, afterEach, describe, expect, it, vi } from "vitest";
+import type { ApiClientOptions } from "../../src/api/client.js";
+import { UNATTENDED_RETRY_OPTIONS } from "../../src/retry.js";
+
+const loaded = vi.hoisted(() => [] as (ApiClientOptions | undefined)[]);
+
+vi.mock("../../src/cli-session.js", () => ({
+    loadSession: (_path: string, apiOptions?: ApiClientOptions) => {
+        loaded.push(apiOptions);
+        return null;
+    },
+}));
+
+const argv = process.argv;
+const dir = mkdtempSync(join(tmpdir(), "quak-bin-test-"));
+
+// Run `quak <args>` to completion. The "Not logged in" message and the exit
+// are swallowed.
+const quak = async (...args: string[]): Promise<void> => {
+    vi.spyOn(process.stderr, "write").mockImplementation(() => true);
+    vi.spyOn(process, "exit").mockImplementation(() => undefined as never);
+    process.argv = ["node", "quak", ...args];
+    vi.resetModules();
+    await import("../../bin/quak.js");
+};
+
+afterEach(() => {
+    process.argv = argv;
+    loaded.length = 0;
+    vi.restoreAllMocks();
+});
+
+afterAll(() => {
+    rmSync(dir, { recursive: true, force: true });
+});
+
+describe("bin/quak.ts session loading", () => {
+    it("loads the session for backup with the unattended retry options", async () => {
+        await quak("backup", dir);
+        // `vi.resetModules()` gave `bin/quak.ts` its own copy of
+        // `src/retry.ts`, whose `sleep` is a different function, so the
+        // options are compared by their numbers.
+        const { attempts, baseDelayMs, maxDelayMs } = UNATTENDED_RETRY_OPTIONS;
+        expect(loaded).toEqual([
+            {
+                retry: expect.objectContaining({
+                    attempts,
+                    baseDelayMs,
+                    maxDelayMs,
+                }),
+            },
+        ]);
+    });
+
+    it("loads the session for another command with the default options", async () => {
+        await quak("collections");
+        expect(loaded).toEqual([undefined]);
+    });
+});

test/cli/commands.test.ts

@@ -18,6 +18,7 @@ import {
     readFileSync,
     rmSync,
     statSync,
+    utimesSync,
     writeFileSync,
 } from "node:fs";
 import { join } from "node:path";
@@ -52,13 +53,14 @@ import {
 import { run } from "../../src/cli-run.js";
 import { loadSession } from "../../src/cli-session.js";
 import type { Client, ClientSnapshot, LoginOptions } from "../../src/client.js";
-import type { ContentSource } from "../../src/library/content.js";
+import { savePath, type ContentSource } from "../../src/library/content.js";
 import type { Collection, EnteFile } from "../../src/model/types.js";
 import { init, toBase64 } from "../../src/crypto/index.js";
 import { defaultCacheDirectory } from "../../src/library/index.js";
 import { HEIC_WITH_EXIF } from "../exif-heic.js";
 import {
     asLivePhoto,
+    blake2b,
     cdnSource,
     IMAGE,
     livePhotoHash,
@@ -703,6 +705,7 @@ describe("backup", () => {
                 "  Failed:      0\n",
         );
         expect(stdout.text).toBe("");
+        expect(existsSync(join(dir, "backup.lock"))).toBe(false);
     });
 
     // The backup opens its library with the precache off: it fetches the
@@ -742,6 +745,77 @@ describe("backup", () => {
         expect(stderr.text).toBe("Starting backup...\n");
     });
 
+    it("--verify downloads again an original that does not match its hash, prints the counts, and exits 0", async () => {
+        // Each file records the hash of the original the fake writes for it.
+        const client = {
+            ...fakeClient(),
+            filesSince: async (args: { collectionID: number }) => ({
+                files: (FILES[args.collectionID] ?? []).map((f) => ({
+                    ...f,
+                    metadata: {
+                        ...f.metadata,
+                        hash: blake2b(Buffer.alloc(7, f.id & 0xff)),
+                    },
+                })),
+                deleted: [],
+                cursor: 1,
+            }),
+        } as unknown as Client;
+        const ctx = context(client);
+        const dir = join(root, "backup");
+        expect(await backupCommand(ctx, dir, {})).toBe(0);
+        writeFileSync(savePath(dir, FILES[1]![0]!), "corrupt");
+
+        expect(await backupCommand(ctx, dir, { verify: true })).toBe(0);
+
+        expect(stderr.text).toContain(
+            "MISMATCH original beach.jpg (100): its bytes do not match its content hash\n",
+        );
+        expect(stderr.text).toContain(
+            "  Downloaded:  1\n" +
+                "  Skipped:     2\n" +
+                "  Verified:    2\n" +
+                "  Mismatched:  1\n" +
+                "  Unchecked:   0\n" +
+                "  Failed:      0\n",
+        );
+    });
+
+    it("--verify --json adds the verified, mismatched and unchecked counts", async () => {
+        const dir = join(root, "backup");
+        expect(await backupCommand(context(), dir, {})).toBe(0);
+
+        const code = await backupCommand(context(), dir, {
+            verify: true,
+            json: true,
+        });
+
+        expect(code).toBe(0);
+        expect(JSON.parse(stdout.text)).toMatchObject({
+            skipped: 3,
+            verified: 0,
+            mismatched: 0,
+            unchecked: 3,
+            failed: 0,
+        });
+    });
+
+    it("exits 1 and lists each file when the ML data fetch fails", async () => {
+        const client = {
+            ...fakeClient(),
+            fetchMLData: async () => {
+                throw new Error("HTTP 503 from server");
+            },
+        } as unknown as Client;
+        expect(
+            await backupCommand(context(client), join(root, "backup"), {}),
+        ).toBe(1);
+        expect(stderr.text).toContain("  Failed:      3\n");
+        expect(stderr.text).toContain(
+            "  [Vacation] beach.jpg (id 100): ML data: HTTP 503 from server\n",
+        );
+    });
+
     // Runs `backup` through `run`, as `bin/quak.ts` does, which prints a thrown
     // error; returns the exit code and what `run` printed.
     const backupThroughRun = async (
@@ -776,7 +850,7 @@ describe("backup", () => {
         expect(code).toBe(1);
         expect(runText).toBe("quak: HTTP 503 from server\n");
         expect(stderr.text).toBe("Starting backup...\nRefreshing library...\n");
-        expect(existsSync(dir)).toBe(false);
+        expect(readdirSync(dir)).toEqual([]);
     });
 
     // A real saved session, read back by `loadSession`, whose server answers
@@ -807,7 +881,38 @@ describe("backup", () => {
             `quak: the saved session is no longer valid; run "quak login"\n`,
         );
         expect(stderr.text).toBe("Starting backup...\nRefreshing library...\n");
-        expect(existsSync(dir)).toBe(false);
+        expect(readdirSync(dir)).toEqual([]);
+    });
+
+    it("exits 2 with one line naming the directory, sending no request, while another backup of it runs", async () => {
+        const dir = join(root, "backup");
+        // The lock another backup holds. Its modification time is set an hour
+        // ahead, so it stays current however long this test takes.
+        const lock = join(dir, "backup.lock");
+        mkdirSync(lock, { recursive: true });
+        const hourAhead = new Date(Date.now() + 3_600_000);
+        utimesSync(lock, hourAhead, hourAhead);
+        // A client whose refresh never finishes, so a run that started one
+        // before it exits would never return.
+        let requests = 0;
+        const never = (): Promise<never> => {
+            requests++;
+            return new Promise(() => {});
+        };
+        const client = {
+            ...fakeClient(),
+            collectionsSince: never,
+            filesSince: never,
+        } as unknown as Client;
+
+        expect(await backupCommand(context(client), dir, {})).toBe(2);
+        expect(requests).toBe(0);
+        expect(stderr.text).toBe(
+            `Starting backup...\nquak: another backup of ${dir} is running\n`,
+        );
+        expect(readdirSync(dir)).toEqual(["backup.lock"]);
+        // Opening the library would have made its cache directory.
+        expect(existsSync(join(root, "cache"))).toBe(false);
     });
 });
 

test/client/session.test.ts

@@ -12,6 +12,10 @@ import { afterAll, beforeAll, describe, expect, it } from "vitest";
 import { init, toBase64 } from "../../src/crypto/index.js";
 import { Client, type ClientSnapshot } from "../../src/client.js";
 import { loadSession } from "../../src/cli-session.js";
+import {
+    DEFAULT_RETRY_OPTIONS,
+    UNATTENDED_RETRY_OPTIONS,
+} from "../../src/retry.js";
 
 const validSnapshot = (): ClientSnapshot => {
     const kp = sodium.crypto_box_keypair();
@@ -176,6 +180,20 @@ describe("loadSession", () => {
         });
     });
 
+    it("gives the restored client the retry options it is passed", () => {
+        const path = join(dir, "retry.json");
+        writeFileSync(path, JSON.stringify(validSnapshot()));
+        const retryOptions = (client: Client | null) =>
+            client!.getApiClient().getRetryOptions();
+
+        expect(
+            retryOptions(
+                loadSession(path, { retry: UNATTENDED_RETRY_OPTIONS }),
+            ),
+        ).toEqual(UNATTENDED_RETRY_OPTIONS);
+        expect(retryOptions(loadSession(path))).toEqual(DEFAULT_RETRY_OPTIONS);
+    });
+
     it("says the file is corrupt when it is not JSON", () => {
         const path = join(dir, "truncated.json");
         writeFileSync(path, '{"email": "user@exa');

test/exif-heic.ts

@@ -1,7 +1,7 @@
 /**
  * `exif.heic`, beside this file: a real 64x64 HEIC whose EXIF holds the same
- * values as the hand-built JPEG in `library/content-library.test.ts`, for the
- * tests of `exif()` and `backup-metadata --exif`.
+ * values as the hand-built JPEG in `exif-jpeg.ts`, for the tests of `exif()`,
+ * `backup-metadata --exif` and the image metadata `quak backup` records.
  *
  * It was made once, in a throwaway node:22-alpine container (Alpine 3.23.3),
  * with libheif 1.23.0 and exiftool 13.55:

test/exif-jpeg.ts

@@ -0,0 +1,89 @@
+/**
+ * Hand-built JPEGs for the tests of `exif()` and of the image metadata
+ * `quak backup` records: one whose EXIF holds the same values as `exif.heic`
+ * (see `exif-heic.ts`), and one whose EXIF cannot be parsed.
+ */
+
+// Big-endian bytes for the hand-built JPEG below.
+const u16 = (n: number): number[] => [n >> 8, n & 0xff];
+const u32 = (n: number): number[] => [...u16(n >>> 16), ...u16(n & 0xffff)];
+const ascii = (s: string): number[] => [...new TextEncoder().encode(s), 0];
+const rational = (num: number, den: number): number[] => [
+    ...u32(num),
+    ...u32(den),
+];
+// One IFD entry: tag, type (1 BYTE, 2 ASCII, 3 SHORT, 4 LONG, 5 RATIONAL),
+// count, then the value when it fits in 4 bytes, else its offset.
+const entry = (
+    tag: number,
+    type: number,
+    count: number,
+    value: number[],
+): number[] => [...u16(tag), ...u16(type), ...u32(count), ...value];
+
+// The TIFF block of a JPEG's EXIF segment, holding every field `Photo`'s typed
+// methods return: the camera in the first IFD, the exposure in the Exif IFD,
+// and a GPS position of 40°26'46" N, 79°58'56" W, 12.5 m below sea level.
+// Offsets count from the start of this block.
+const TIFF = [
+    ...[0x4d, 0x4d, 0x00, 0x2a], // big-endian TIFF
+    ...u32(8), // the first IFD's offset
+    // The first IFD, at 8: five entries, then no next IFD.
+    ...u16(5),
+    ...entry(0x010f, 2, 6, u32(74)), // Make
+    ...entry(0x0110, 2, 7, u32(80)), // Model
+    ...entry(0x0112, 3, 1, [...u16(6), 0, 0]), // Orientation
+    ...entry(0x8769, 4, 1, u32(88)), // the Exif IFD's offset
+    ...entry(0x8825, 4, 1, u32(246)), // the GPS IFD's offset
+    ...u32(0),
+    ...ascii("Canon"), // at 74
+    ...ascii("EOS R5"), // at 80
+    0, // a pad byte
+    // The Exif IFD, at 88: seven entries, then no next IFD.
+    ...u16(7),
+    ...entry(0x829a, 5, 1, u32(178)), // ExposureTime
+    ...entry(0x829d, 5, 1, u32(186)), // FNumber
+    ...entry(0x8827, 3, 1, [...u16(400), 0, 0]), // ISOSpeedRatings
+    ...entry(0x9003, 2, 20, u32(194)), // DateTimeOriginal
+    ...entry(0x9011, 2, 7, u32(214)), // OffsetTimeOriginal
+    ...entry(0x920a, 5, 1, u32(222)), // FocalLength
+    ...entry(0xa434, 2, 16, u32(230)), // LensModel
+    ...u32(0),
+    ...rational(1, 250), // at 178
+    ...rational(28, 10), // at 186
+    ...ascii("2021:07:15 14:30:00"), // at 194
+    ...ascii("+02:00"), // at 214
+    0, // a pad byte
+    ...rational(50, 1), // at 222
+    ...ascii("RF50mm F1.8 STM"), // at 230
+    // The GPS IFD, at 246: six entries, then no next IFD.
+    ...u16(6),
+    ...entry(0x0001, 2, 2, [...ascii("N"), 0, 0]), // GPSLatitudeRef
+    ...entry(0x0002, 5, 3, u32(324)), // GPSLatitude
+    ...entry(0x0003, 2, 2, [...ascii("W"), 0, 0]), // GPSLongitudeRef
+    ...entry(0x0004, 5, 3, u32(348)), // GPSLongitude
+    ...entry(0x0005, 1, 1, [1, 0, 0, 0]), // GPSAltitudeRef: below sea level
+    ...entry(0x0006, 5, 1, u32(372)), // GPSAltitude
+    ...u32(0),
+    ...[...rational(40, 1), ...rational(26, 1), ...rational(46, 1)], // at 324
+    ...[...rational(79, 1), ...rational(58, 1), ...rational(56, 1)], // at 348
+    ...rational(25, 2), // at 372
+];
+
+export const JPEG_WITH_EXIF = new Uint8Array([
+    ...[0xff, 0xd8], // start of image
+    ...[0xff, 0xe1, ...u16(2 + 6 + TIFF.length)], // APP1 and its length
+    ...[...ascii("Exif"), 0], // "Exif\0\0"
+    ...TIFF,
+    ...[0xff, 0xda, 0x00, 0x02], // start of scan
+]);
+
+// A JPEG whose EXIF segment is laid out correctly but holds "XX" where the TIFF
+// byte order belongs, so exifreader cannot parse it.
+export const JPEG_WITH_BAD_EXIF = new Uint8Array([
+    ...[0xff, 0xd8], // start of image
+    ...[0xff, 0xe1, ...u16(2 + 6 + 2)], // APP1 and its length
+    ...[...ascii("Exif"), 0], // "Exif\0\0"
+    ...[0x58, 0x58], // "XX"
+    ...[0xff, 0xda, 0x00, 0x02], // start of scan
+]);

test/library/content-library.test.ts

@@ -29,6 +29,7 @@ import type { CollectionsPage, FilesPage } from "../../src/client.js";
 import type { Collection, EnteFile } from "../../src/model/types.js";
 import { readPhotoExif, type PhotoExif } from "../../src/exif.js";
 import { HEIC_WITH_EXIF } from "../exif-heic.js";
+import { JPEG_WITH_BAD_EXIF, JPEG_WITH_EXIF } from "../exif-jpeg.js";
 import {
     asLivePhoto,
     cdnSource,
@@ -256,90 +257,6 @@ describe("Library content wiring", () => {
     });
 });
 
-// Big-endian bytes for the hand-built JPEG below.
-const u16 = (n: number): number[] => [n >> 8, n & 0xff];
-const u32 = (n: number): number[] => [...u16(n >>> 16), ...u16(n & 0xffff)];
-const ascii = (s: string): number[] => [...new TextEncoder().encode(s), 0];
-const rational = (num: number, den: number): number[] => [
-    ...u32(num),
-    ...u32(den),
-];
-// One IFD entry: tag, type (1 BYTE, 2 ASCII, 3 SHORT, 4 LONG, 5 RATIONAL),
-// count, then the value when it fits in 4 bytes, else its offset.
-const entry = (
-    tag: number,
-    type: number,
-    count: number,
-    value: number[],
-): number[] => [...u16(tag), ...u16(type), ...u32(count), ...value];
-
-// The TIFF block of a JPEG's EXIF segment, holding every field `Photo`'s typed
-// methods return: the camera in the first IFD, the exposure in the Exif IFD,
-// and a GPS position of 40°26'46" N, 79°58'56" W, 12.5 m below sea level.
-// Offsets count from the start of this block.
-const TIFF = [
-    ...[0x4d, 0x4d, 0x00, 0x2a], // big-endian TIFF
-    ...u32(8), // the first IFD's offset
-    // The first IFD, at 8: five entries, then no next IFD.
-    ...u16(5),
-    ...entry(0x010f, 2, 6, u32(74)), // Make
-    ...entry(0x0110, 2, 7, u32(80)), // Model
-    ...entry(0x0112, 3, 1, [...u16(6), 0, 0]), // Orientation
-    ...entry(0x8769, 4, 1, u32(88)), // the Exif IFD's offset
-    ...entry(0x8825, 4, 1, u32(246)), // the GPS IFD's offset
-    ...u32(0),
-    ...ascii("Canon"), // at 74
-    ...ascii("EOS R5"), // at 80
-    0, // a pad byte
-    // The Exif IFD, at 88: seven entries, then no next IFD.
-    ...u16(7),
-    ...entry(0x829a, 5, 1, u32(178)), // ExposureTime
-    ...entry(0x829d, 5, 1, u32(186)), // FNumber
-    ...entry(0x8827, 3, 1, [...u16(400), 0, 0]), // ISOSpeedRatings
-    ...entry(0x9003, 2, 20, u32(194)), // DateTimeOriginal
-    ...entry(0x9011, 2, 7, u32(214)), // OffsetTimeOriginal
-    ...entry(0x920a, 5, 1, u32(222)), // FocalLength
-    ...entry(0xa434, 2, 16, u32(230)), // LensModel
-    ...u32(0),
-    ...rational(1, 250), // at 178
-    ...rational(28, 10), // at 186
-    ...ascii("2021:07:15 14:30:00"), // at 194
-    ...ascii("+02:00"), // at 214
-    0, // a pad byte
-    ...rational(50, 1), // at 222
-    ...ascii("RF50mm F1.8 STM"), // at 230
-    // The GPS IFD, at 246: six entries, then no next IFD.
-    ...u16(6),
-    ...entry(0x0001, 2, 2, [...ascii("N"), 0, 0]), // GPSLatitudeRef
-    ...entry(0x0002, 5, 3, u32(324)), // GPSLatitude
-    ...entry(0x0003, 2, 2, [...ascii("W"), 0, 0]), // GPSLongitudeRef
-    ...entry(0x0004, 5, 3, u32(348)), // GPSLongitude
-    ...entry(0x0005, 1, 1, [1, 0, 0, 0]), // GPSAltitudeRef: below sea level
-    ...entry(0x0006, 5, 1, u32(372)), // GPSAltitude
-    ...u32(0),
-    ...[...rational(40, 1), ...rational(26, 1), ...rational(46, 1)], // at 324
-    ...[...rational(79, 1), ...rational(58, 1), ...rational(56, 1)], // at 348
-    ...rational(25, 2), // at 372
-];
-
-const JPEG_WITH_EXIF = new Uint8Array([
-    ...[0xff, 0xd8], // start of image
-    ...[0xff, 0xe1, ...u16(2 + 6 + TIFF.length)], // APP1 and its length
-    ...[...ascii("Exif"), 0], // "Exif\0\0"
-    ...TIFF,
-    ...[0xff, 0xda, 0x00, 0x02], // start of scan
-]);
-
-// A JPEG whose EXIF segment is laid out correctly but holds "XX" where the TIFF
-// byte order belongs, so exifreader cannot parse it.
-const JPEG_WITH_BAD_EXIF = new Uint8Array([
-    ...[0xff, 0xd8], // start of image
-    ...[0xff, 0xe1, ...u16(2 + 6 + 2)], // APP1 and its length
-    ...[...ascii("Exif"), 0], // "Exif\0\0"
-    ...[0x58, 0x58], // "XX"
-    ...[0xff, 0xda, 0x00, 0x02], // start of scan
-]);
-
 describe("Photo save path, local copy, content and EXIF", () => {
     // The same account, with `files` in its album instead.
     class FilesClient extends MockClient {

test/live-photo.ts

@@ -30,7 +30,8 @@ export const livePhotoZip = (
     },
 ): Uint8Array => zipSync(entries);
 
-const blake2b = (bytes: Uint8Array): string =>
+// The content hash Ente's clients record for an original's bytes.
+export const blake2b = (bytes: Uint8Array): string =>
     createHash("blake2b512").update(bytes).digest("base64");
 
 // The hash Ente's clients record for a live photo: the unkeyed BLAKE2b-512 of

test/packaging/build-context.test.ts

@@ -29,18 +29,19 @@ const patterns = (name: string): string[] =>
 const dockerignore = patterns(".dockerignore");
 
 describe(".dockerignore", () => {
-    // Everything here is either generated, enormous, or secret. `.claude/` is
-    // the correctness one: see the header comment and issue #25.
+    // Everything here is either generated, enormous, or secret. `.claude` is
+    // the correctness one: see the header comment and issue #25. The leading
+    // `/` anchors an entry at the root of the context.
     it.each([
-        ".claude/",
-        ".quak/",
-        "bin/quak",
-        "node_modules",
-        "coverage",
-        "dist",
-        ".vitest-cache/",
-        ".nyc_output/",
-        "*.tsbuildinfo",
+        ".claude",
+        "/.quak",
+        "/bin/quak",
+        "**/node_modules",
+        "/coverage",
+        "/dist",
+        "/.vitest-cache",
+        "/.nyc_output",
+        "/*.tsbuildinfo",
     ])("keeps %s out of the build context", (pattern) => {
         expect(dockerignore).toContain(pattern);
     });
@@ -57,7 +58,7 @@ describe(".dockerignore", () => {
     // The build stage is the final image, so a .git/config sent in would
     // ship the clone's remote URL and any credential in it.
     it("sends .git without its config", () => {
-        expect(dockerignore).toContain(".git/config");
+        expect(dockerignore).toContain("**/.git/config");
     });
 
     // BuildKit lets a `Dockerfile.dockerignore` shadow the root one; such a

test/retry/retry.test.ts

@@ -43,6 +43,7 @@ import {
     isRetryable,
     isSafeToReplay,
     resolveRetryOptions,
+    UNATTENDED_RETRY_OPTIONS,
     withRetry,
 } from "../../src/retry.js";
 import { ApiError, TruncatedStreamError } from "../../src/errors.js";
@@ -639,3 +640,33 @@ describe("retry defaults", () => {
         }
     });
 });
+
+describe("unattended retry options", () => {
+    it("allow ten attempts, a 1 s base delay and a 60 s cap", () => {
+        // The numbers the README documents for `quak backup`.
+        expect(UNATTENDED_RETRY_OPTIONS.attempts).toBe(10);
+        expect(UNATTENDED_RETRY_OPTIONS.baseDelayMs).toBe(1_000);
+        expect(UNATTENDED_RETRY_OPTIONS.maxDelayMs).toBe(60_000);
+    });
+
+    it("give up on a request that keeps failing after at most 243 s of waiting", async () => {
+        // `random: () => 1` makes every wait its ceiling, the worst case.
+        const { sleep, delays } = recordingSleep();
+        let calls = 0;
+        await expect(
+            withRetry(
+                () => {
+                    calls++;
+                    return Promise.reject(new ApiError("HTTP 503", 503));
+                },
+                { ...UNATTENDED_RETRY_OPTIONS, sleep, random: () => 1 },
+            ),
+        ).rejects.toThrow("HTTP 503");
+
+        expect(calls).toBe(10);
+        expect(delays).toEqual([
+            1_000, 2_000, 4_000, 8_000, 16_000, 32_000, 60_000, 60_000, 60_000,
+        ]);
+        expect(delays.reduce((sum, ms) => sum + ms, 0)).toBe(243_000);
+    });
+});

yarn.lock

@@ -535,6 +535,18 @@
   dependencies:
     undici-types "~6.21.0"
 
+"@types/proper-lockfile@4.1.4":
+  version "4.1.4"
+  resolved "https://registry.yarnpkg.com/@types/proper-lockfile/-/proper-lockfile-4.1.4.tgz#cd9fab92bdb04730c1ada542c356f03620f84008"
+  integrity sha512-uo2ABllncSqg9F1D4nugVl9v93RmjxF6LJzQLMLDdPaXCUIDPeOJ21Gbqi43xNKzBi/WQ0Q0dICqufzQbMjipQ==
+  dependencies:
+    "@types/retry" "*"
+
+"@types/retry@*":
+  version "0.12.5"
+  resolved "https://registry.yarnpkg.com/@types/retry/-/retry-0.12.5.tgz#f090ff4bd8d2e5b940ff270ab39fd5ca1834a07e"
+  integrity sha512-3xSjTp3v03X/lSQLkczaN9UIEwJMoMCA1+Nb5HfbJEQWogdeQIyVtTvxPXDQjZ5zws8rFQfVfRdz03ARihPJgw==
+
 "@typescript-eslint/eslint-plugin@8.46.2":
   version "8.46.2"
   resolved "https://registry.yarnpkg.com/@typescript-eslint/eslint-plugin/-/eslint-plugin-8.46.2.tgz#dc4ab93ee3d7e6c8e38820a0d6c7c93c7183e2dc"
@@ -1140,6 +1152,11 @@ globals@^14.0.0:
   resolved "https://registry.yarnpkg.com/globals/-/globals-14.0.0.tgz#898d7413c29babcf6bafe56fcadded858ada724e"
   integrity sha512-oahGvuMGQlPw/ivIYBjVSrWAfWLBeku5tpPE2fOPLi+WHffIWbuh2tCjhyQhTBPMf5E9jDEH4FOmTYgYwbKwtQ==
 
+graceful-fs@^4.2.4:
+  version "4.2.11"
+  resolved "https://registry.yarnpkg.com/graceful-fs/-/graceful-fs-4.2.11.tgz#4183e4e8bf08bb6e05bbb2f7d2e0c8f712ca40e3"
+  integrity sha512-RbJ5/jmFcNNCcDV5o9eTnBLJ/HszWV0P73bc+Ff4nS/rJj+YaS6IGyiOL0VoBYX+l1Wrl3k63h/KrH+nhJ0XvQ==
+
 graphemer@^1.4.0:
   version "1.4.0"
   resolved "https://registry.yarnpkg.com/graphemer/-/graphemer-1.4.0.tgz#fb2f1d55e0e3a1849aeffc90c4fa0dd53a0e66c6"
@@ -1414,6 +1431,15 @@ prettier@3.8.1:
   resolved "https://registry.yarnpkg.com/prettier/-/prettier-3.8.1.tgz#edf48977cf991558f4fcbd8a3ba6015ba2a3a173"
   integrity sha512-UOnG6LftzbdaHZcKoPFtOcCKztrQ57WkHDeRD9t/PTQtmT0NHSeWWepj6pS0z/N7+08BHFDQVUrfmfMRcZwbMg==
 
+proper-lockfile@4.1.2:
+  version "4.1.2"
+  resolved "https://registry.yarnpkg.com/proper-lockfile/-/proper-lockfile-4.1.2.tgz#c8b9de2af6b2f1601067f98e01ac66baa223141f"
+  integrity sha512-TjNPblN4BwAWMXU8s9AEz4JmQxnD1NNL7bNOY/AKUzyamc379FWASUhc/K1pL2noVb+XmZKLL68cjzLsiOAMaA==
+  dependencies:
+    graceful-fs "^4.2.4"
+    retry "^0.12.0"
+    signal-exit "^3.0.2"
+
 punycode@^2.1.0:
   version "2.3.1"
   resolved "https://registry.yarnpkg.com/punycode/-/punycode-2.3.1.tgz#027422e2faec0b25e1549c3e1bd8309b9133b6e5"
@@ -1429,6 +1455,11 @@ resolve-from@^4.0.0:
   resolved "https://registry.yarnpkg.com/resolve-from/-/resolve-from-4.0.0.tgz#4abcd852ad32dd7baabfe9b40e00a36db5f392e6"
   integrity sha512-pb/MYmXstAkysRFx8piNI1tGFNQIFA3vkE3Gq4EuA1dF6gHp/+vgZqsCGJapvy8N3Q+4o7FwvquPJcnZ7RYy4g==
 
+retry@^0.12.0:
+  version "0.12.0"
+  resolved "https://registry.yarnpkg.com/retry/-/retry-0.12.0.tgz#1b42a6266a21f07421d1b0b54b7dc167b01c013b"
+  integrity sha512-9LkiTwjUh6rT555DtE9rTX+BKByPfrMzEAtnlEtdEwr3Nkffwiihqe2bWADg+OQRjt9gl6ICdmB/ZFDCGAtSow==
+
 reusify@^1.0.4:
   version "1.1.0"
   resolved "https://registry.yarnpkg.com/reusify/-/reusify-1.1.0.tgz#0fe13b9522e1473f51b558ee796e08f11f9b489f"
@@ -1502,6 +1533,11 @@ siginfo@^2.0.0:
   resolved "https://registry.yarnpkg.com/siginfo/-/siginfo-2.0.0.tgz#32e76c70b79724e3bb567cb9d543eb858ccfaf30"
   integrity sha512-ybx0WO1/8bSBLEWXZvEd7gMW3Sn3JFlW3TvX1nREbDLRNQNaeNN8WK0meBwPdAaOI7TtRRRJn/Es1zhrrCHu7g==
 
+signal-exit@^3.0.2:
+  version "3.0.7"
+  resolved "https://registry.yarnpkg.com/signal-exit/-/signal-exit-3.0.7.tgz#a9a1767f8af84155114eaabd73f99273c8f59ad9"
+  integrity sha512-wnD2ZE+l+SPC/uoS0vXeE9L1+0wuaMqKlfz9AMUo38JsyLSBWSFcHR1Rri62LZc12vLr1gb3jl7iwQhgwpAbGQ==
+
 signal-exit@^4.1.0:
   version "4.1.0"
   resolved "https://registry.yarnpkg.com/signal-exit/-/signal-exit-4.1.0.tgz#952188c1cbd546070e2dd20d0f41c0ae0530cb04"