Files
multica/server/internal/handler/file_test.go
beast 60048172a7 fix(lark): ingest inbound images and videos as chat attachments (MUL-4934) (#5580)
* fix: ingest feishu media as chat attachments

* fix: ingest feishu post embedded media

* fix(lark): make inbound media retries safe

* fix lark media resource limit

* fix(lark): move inbound media off ack path

* fix(channel): make inbound media runs durable

* fix(channel): close enqueue-vs-append race on media deferral

EnqueueChatTask read the session-wide media deadline in one statement and
sealed the input batch in a later one. Under READ COMMITTED a media message
committing between the two got sealed into a task the deadline read had
already decided was 'queued', so the daemon could claim it before its
attachment bound — the agent received the bare placeholder, and the later
media-ready promotion was a no-op against a non-deferred task.

After the seal, re-derive the deferral from the sealed batch itself in the
same transaction (DeferChatTaskForSealedPendingMedia): if any sealed message
still carries an unexpired media marker, flip the task to deferred with
fire_at aligned to the latest marker. The existing post-commit promote fence
already covers the opposite direction (marker cleared mid-transaction).

Adds a deterministic regression test that injects the media append between
the deadline read and the seal via a wrapped pgx.Tx.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): keep committed chat task out of enqueue error path

The post-commit media-ready fence returned its error from EnqueueChatTask
even though the deferred task was already durably committed. The router
flush treats any enqueue error as "no task exists": it clears the typing
indicator and logs an enqueue failure while the run still happens at its
fire_at deadline. Log the fence failure instead — the claim-path deferred
promoter re-queues the task regardless.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): cap global media resolution concurrency

Media jobs were serialized per session but unbounded across sessions: a
burst could open arbitrarily many concurrent 45s Lark downloads, and each
unknown-length upload may buffer up to the 100 MiB resource cap in memory.
Gate resolveAndBindMedia behind a global slot semaphore (default 8,
RouterConfig.MediaConcurrency). Per-session ordering is unchanged; on
shutdown a job cancelled while waiting for a slot proceeds straight to the
bounded DB finalize so marker clearing stays prompt. Also document that the
per-message media budget spans queue/slot waits (it must match the
persisted fire_at) and why timed-out uploads cannot leak unbounded orphans.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(chat): keep channel-sealed user messages on task cancel

Sealing the channel input batch stamps task_id onto channel user
messages, which exposed them to the cancel draft-restore path: an
empty-transcript cancel would DeleteUserChatMessageByTask the sealed
Feishu/Slack messages and detach their attachments. Those messages are
the durable record of what the platform sender wrote — the sender has
no Multica composer to restore a draft into.

Gate the restore-delete on ChatSessionHasChannelBinding in both the
synchronous finalize and the deferred finalize (the latter covers
markers left by an older replica during a rolling deploy); a bound
session now settles as "Stopped." instead.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): skip the media pipeline for messages without media

Every inbound message on a Media-enabled platform persisted a 45s
media deadline and queued a resolution job, so a plain text message
could wait behind the global media semaphore (its task deferred while
other sessions download 100 MiB videos) and a crash between append and
clear delayed a pure-text run to the full 45s fallback.

Add MediaResolver.HasMedia — a pure in-memory probe the Router calls
on the ACK path — and only persist the deadline / enqueue the job when
the message actually references platform media. The Feishu resolver
decodes the already-received payload and reports standalone image or
video keys and post-embedded img/media spans.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(chat): gate cancel restore on immutable channel provenance

The previous guard keyed the cancel restore-delete off
ChatSessionHasChannelBinding, but a binding only proves routing exists
right now: archiving a session and rebinding an installation both
delete the binding while preserving chat history, so a still-cancellable
sealed task could again restore-delete the original inbound messages.

Persist provenance on the message instead: migration 203 adds
chat_message.channel_ingested, stamped inside the channel append
transaction and never mutated, and both cancel finalize paths now gate
on TaskHasChannelIngestedMessages over the task's sealed batch. The
binding-existence query is removed. Regression tests cover ingest ->
archive/unbind -> cancel for a queued and a started task.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): reclaim media uploads that never gain an attachment row

Deadline expiry dropped already-resolved refs and a BindMedia failure
was log-only, leaving uploaded objects with no attachment row and no
reclaim path — the dedup mark commits with the message before media
runs, so a redelivery is dropped as a duplicate and never re-resolves
(and thus never overwrites) those keys, and workspace/session deletion
only enumerates the attachment table.

Add MediaResolver.DiscardMedia — a best-effort delete by StorageKey —
and call it from both failure paths in resolveAndBindMedia. The Feishu
resolver forwards to the storage backend's Delete. Tests cover a
partial upload discarded at the deadline, discard on bind failure, and
key-level deletion in the resolver.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs(server): refresh comments stale after detached media ingestion

Channel tasks now seal a self-owned input batch, media ingestion is no
longer out of scope for the flattener, and MediaRefs are filled by the
detached resolver after append rather than by feishuChannel pre-engine.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(chat): stop keying channel empty-completion silence off chat_input_task_id

Sealing gave channel tasks a self input-owner, which broke
writeChatCompletionOutcome's discriminator: it treated any owned task
as direct, so an empty channel completion wrote the no_response
fallback row and the outbound patcher — which forwards any non-empty
chat:done content verbatim — pushed the English fallback body to
Feishu/Slack, violating the MUL-4351 contract.

Silence is now decided by the immutable channel_ingested provenance of
the task's input batch, looked up by the batch OWNER id
(chat_input_task_id): auto-retry clones inherit the owner while their
sealed messages stay tagged with the parent's id, so keying off the
task's own id would misread a channel retry as direct. The cancel-path
provenance gates switch to the same owner key via chatInputOwnerID.
chat_input_task_id is back to meaning only "input batch owner".

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* chore(migrations): renumber to 203/204 after upstream took 202

Upstream main merged 202_runtime_profile_add_qwen while this branch
held 202/203, tripping TestMigrationNumericPrefixesStayUniqueAfterLegacySet
on the CI merge tree. channel_media_pending becomes 203 and
channel_ingested becomes 204; no content changes.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channels): gate outbound delivery on channel provenance, not owner

Merging main brought #5645 (keep direct chat replies in Multica),
whose outbound gate assumed channel tasks leave chat_input_task_id
NULL. Sealed channel tasks own an input batch too, so on the merge
tree every channel reply and failure notice was classified as direct
and silently dropped — agents stopped replying in Feishu/Slack.

Both outbound gates now call engine.TaskInputIsChannelIngested: a NULL
owner keeps #5645's deliver-by-default for pre-sealing tasks, an owned
batch delivers only when it carries the immutable channel_ingested
stamp (keyed by the owner id, so auto-retry clones inherit the
verdict). Direct replies stay in Multica; sealed channel replies reach
the platform. Tests cover both directions on both platforms.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): discard media orphans on a fresh context, after finalize

DiscardMedia shared finalizeCtx with BindMedia, so a bind that failed
because the finalize deadline expired handed the storage deletes an
already-dead context — the compensation silently no-opped and the
orphans leaked anyway. The deadline path also ran S3 deletes before
the marker clear, eating the same 5s budget the user-facing
bind/promotion needed.

Collect the refs from both failure paths, run bind + promotion on the
finalize budget first, then delete on a fresh discard context. The
bind-failure test now pins that discard receives a live context.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): compensate result-uncertain media uploads and commits

The compensation protocol treated "the call returned an error" as "the
side effect did not happen", which is wrong in both directions across
the result-uncertain windows:

- An upload error can follow a server-side write (lost response,
  deadline mid-write). The attempted key never reached the router, so
  nothing could reclaim it and dedup guarantees no re-resolve. The
  resolver now idempotently deletes the deterministic key on a fresh
  budget right at the failure site.

- A commit error is not a rollback guarantee: a lost ack can report
  failure after Postgres durably committed the attachment rows, and
  the router's discard would then delete objects those rows reference.
  BindMediaRefs now converges the ambiguity on a fresh budget — any of
  the batch's URLs present proves the atomic commit landed (bind
  reports success); none proves the rollback (discard stays safe); a
  failed verification returns ErrMediaBindResultUnknown and the router
  keeps the uploads, preferring a rare orphan over a broken attachment.

Fault-injection coverage: an upload error deletes the attempted key; a
lost-ack commit keeps the bound attachment and reports success; a
verified rollback stays a discardable error; the router keeps uploads
on the unknown-outcome sentinel.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* chore(migrations): renumber to 207/208 after upstream took 203-206

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* docs(channel): note DiscardMedia self-invocation and the unknown-outcome skip

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* chore(migrations): renumber to 212/213 after upstream took 207-211

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* feat(channel): replace inline media compensation with an intent ledger and reconciler

Inline best-effort compensation cannot answer "did my side effect
happen?" at the moment it needs the answer — the DELETE/PUT reordering
and the empty-read-vs-in-flight-COMMIT gaps were both instances of the
same two-system atomicity problem. Persist the intent instead and let
an asynchronous reconciler settle it:

- channel_media_pending_object (migration 214; claim index 215 as its
  own single-statement CONCURRENTLY migration): a state machine row
  ('pending' -> 'deleting') with lease, attempt, and backoff columns.
- The resolver upserts the row BEFORE each PUT, state-guarded so a key
  the reconciler owns is never resurrected (the resource is skipped).
  ObjectURL is a pure function of configuration, so the row carries the
  attachment URL pre-upload.
- BindMediaRefs deletes the batch's rows INSIDE the attachment-insert
  transaction: commit landed <=> intents gone, atomically, so an
  ambiguous COMMIT never needs adjudication. A key already claimed to
  'deleting' is skipped (placeholder stays).
- Nothing is ever deleted inline. The reconciler — an independent
  worker so storage latency cannot starve other sweepers — claims due
  rows ('pending' past the settle delay, or expired leases) under a
  fresh lease, checks for a durable attachment reference only AFTER the
  claim (race-free: bind can no longer succeed on the key), deletes
  unreferenced objects outside any transaction, and backs off failed
  deletes with attempt-based retry. Crash windows converge for free.
- The settle delay is a fixed constant carrying NO correctness weight;
  invariant tests pin it at >=10x every pipeline budget. Metrics cover
  deletes, referenced clears, delete failures, and ledger backlog.

Removed: MediaResolver.DiscardMedia, ErrMediaBindResultUnknown, the
post-commit verification, and both router discard branches.

Tests: intent-before-upload ordering; upload error leaves the row and
deletes nothing; bind-wins vs reconciler-wins on the same key; lost-ack
and rolled-back commit injections (intent cleared iff the attachment
landed); reconciler three-state settle; expired-lease reclaim; delete
failure backoff and retry; settle invariants.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): never build or sweep the media reconciler without storage

store is nil when S3 is unconfigured AND the local upload dir fails to
initialize, but the reconciler was constructed unconditionally and
main only gates the goroutine on the reconciler pointer — the first
unreferenced ledger row (rows can pre-exist from a boot where storage
worked) would nil-pointer panic a bare goroutine and take down the
process.

Construct the reconciler only when a storage backend exists, and guard
RunOnce defensively: with no deleter it skips the sweep without
claiming, so rows are not stranded in 'deleting' until lease expiry.
Test covers the pre-existing-row + missing-storage boot.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* chore(migrations): renumber to 213-216 after upstream took 212

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* refactor(channel): remove the dead pre-resolved MediaRefs ingress path

lark.InboundMessage.MediaRefs and the resolver's early-returns for
pre-populated refs were vestiges of the pre-detached synchronous design
— no producer fills them before the router anymore. Worse, the intent
ledger made the path actively misleading: refs arriving without ledger
rows would be silently skipped at bind (with a log blaming the
reconciler), contradicting the field's "already persisted" contract.

Delete the field, its channelMessageFromLark mapping, and both
early-returns; channel.InboundMessage.MediaRefs is now documented as
what it actually is — ResolveMedia's output channel, always empty on
ingress, attachable only through a claimed ledger intent.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): enforce workspace tenancy on every ledger query

The intent-ledger upsert's conflict branch guarded only on state, so a
cross-workspace storage_key collision could rewrite the row's
workspace/message/url ownership; release and delete keyed on
(storage_key, lease_token) alone. The derived key embeds the workspace
UUID so none of this is reachable today — but tenancy must be enforced
by the workspace column in every query, never derived from the key
string (MUL-3515 rule, restated in this PR's review).

The upsert now updates only within the same workspace (a cross-tenant
conflict updates nothing, returns no row, and the resolver skips the
upload — the fail-safe direction), and release/delete take
(workspace_id, storage_key, lease_token). Tests pin that a foreign
workspace can neither steal, release, nor delete a row.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(migrations): build the ledger primary key via a concurrent index

storage_key TEXT PRIMARY KEY created its unique index implicitly at
CREATE TABLE, against the repo convention that every migration index —
including a new table's unique index — is built CONCURRENTLY in its
own single-statement migration (the exact three-step pattern
client_usage_daily shipped in 207-209). The table now declares
storage_key NOT NULL, 216 builds the unique index concurrently, and
217 attaches the primary key USING INDEX; the claim index moves to
218. ON CONFLICT (storage_key) still resolves against the constraint,
and the full down/up round-trip is verified.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): bound each reconciler object delete with its own timeout

DeleteObject ran on the worker-lifetime context and the SDK's default
HTTP client has no overall request timeout, so one black-holed
connection would wedge the sequential sweep loop — and with it every
later batch and the backlog gauge — forever; a single-replica
deployment has no other worker to reclaim the lease. Each delete now
gets a 30s timeout (well under the 2min lease), and a timed-out delete
takes the existing release/backoff path. Covered by a blocking-deleter
test with an injectable timeout.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): anchor the media deadline to the DB clock and bound queue waits by it

Two deadline gaps from review:

- The persisted marker was an application-clock timestamp compared
  against SQL now() everywhere it is read, so a skewed app node could
  shrink the fallback window and hand the agent a placeholder before
  the resolver's local budget ended. The append transaction now anchors
  a relative budget (MediaPendingSeconds) with now() + make_interval,
  writer and readers sharing one clock; the local resolve budget stays
  monotonic app-side. A DB test pins that the remaining budget measured
  by the DB clock equals the requested one.

- enqueueMedia's waits (per-session order, global slot) only watched
  shutdown, so in a burst an already-expired job kept its goroutine and
  payload until it reached the front. Both waits now also watch the
  message's deadline; on expiry the job skips the resolver entirely and
  runs only the empty finalize (marker clear + promotion), which also
  unblocks the session's later messages. Covered by a queued-expiry
  test that finalizes while the only slot is deterministically held.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* chore(migrations): renumber to 216-221 after upstream took 213-215

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): start the local media budget before the append transaction

The DB anchors the durable fallback at insert-time now(), but the
local monotonic budget started only after AppendMessage returned — so
the resolver outlived the fallback by the append/commit latency, a
window where the deferred task is already claimable while the resolver
still runs and the agent reads a placeholder that binds moments later.
Capture the local deadline before calling AppendMessage, restoring the
ordering local-gives-up <= durable-fallback-fires. A slow-append test
pins that the resolver's context deadline is measured from the
pre-append instant.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* chore(migrations): renumber to 224-229 after upstream took 216-223

Verified against the merged tree: the numeric-prefix uniqueness test
passes and the full migration set applies cleanly from scratch.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): heartbeat the reconciler lease per row

One claim covers up to 50 rows under a single 2-minute lease, but the
batch is processed sequentially and each delete may run its full 30s
timeout — a few stalled deletes could outlive the lease mid-batch,
letting another replica reclaim the tail: duplicate concurrent
deletes, inflated attempt/backoff on rows whose owner was alive, and
skewed metrics.

The lease is now renewed before EACH row's settle work, so it only
ever needs to cover one row's worst case (invariant-tested: lease >=
2x the per-delete timeout). A renewal that matches no row means
another worker reclaimed it after a genuine expiry — the row is
skipped, leaving the new owner's state untouched. Test simulates a
mid-batch reclaim and pins the skip.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): dedup post media resources and make local writes atomic

A rich post may reference the same image_key/file_key in several spans.
The object key derives from (message, type, key), so duplicates uploaded
to the SAME key twice: LocalStorage.UploadStream truncated the
destination up front and removed it outright on a copy error, so a
second failing attempt destroyed the object the first success had
produced — leaving an attachment row pointing at nothing. A second
succeeding attempt instead produced two attachment rows for one object.

Collapse duplicate spans by (fetch type, platform key) before the
upload loop, and write local uploads through a temp file renamed into
place so a failed write can only discard its own temp file. Tests cover
a duplicated span uploading once and a failed re-upload leaving the
previous object intact.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): fence late-materializing PUTs with a tombstone schedule

A DELETE cannot be ordered against a PUT the client already abandoned:
the store may materialize the object after the delete completes. The
reconciler cleared the ledger row right after deleting, so such an
object had no row and nothing to reclaim it — which made the settle
delay the de-facto correctness barrier for the PUT/DELETE race, exactly
what the design says it must not be.

The row is now kept as a tombstone ('tombstoned' state, migration 226's
CHECK) and re-deleted on a widening schedule (15m, 1h, 6h, 24h, the
pass index carried in last_error), so a late materialization is
reclaimed by a later pass; only after the schedule is exhausted is the
row dropped. Claim, heartbeat, lease, and tenancy predicates are
unchanged — a tombstone is claimed exactly like any other due row. A
separate gauge reports tombstones so they cannot be mistaken for a
backlog of objects awaiting reclaim, and the header comment now states
precisely what state fences (bind/commit) versus what the schedule
fences (late PUTs).

Tests: the reviewer's interleaving — DELETE completes, the abandoned PUT
materializes right after, and the object is gone by the end of the
schedule — plus a full schedule walk asserting the object is counted
once and the row clears at the end.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): tombstones must re-delete, not re-ask the reference question

A tombstone revisit ran the same reference check as a first settle, so
an attachment carrying the same URL — a re-ingested copy of the object —
sent the row down the "referenced, keep it" branch: the object was kept
and the row cleared, abandoning the re-delete schedule that fences the
ORIGINAL object against an abandoned PUT. A tombstone has already been
judged unreferenced and deleted; it exists only to re-delete whatever
materializes later, so it now goes straight to the delete + schedule
tail (extracted as settleDeletedObject, shared with the first settle).

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

* fix(channel): keep the tombstone schedule position in its own column

The re-delete pass index was encoded into last_error, which the failure
path also writes: one failed re-delete erased the position and restarted
the walk. A store failing intermittently could therefore keep a tombstone
alive indefinitely — every recovery would resume at pass 1 and the row
would never reach the end of the schedule to be dropped.

tombstone_pass is now its own column (the table is introduced in this PR,
so migration 226 carries it), advanced only by a successful delete, and
the tombstone write clears the now-stale last_error. Test walks the
schedule across a failed re-delete and asserts it resumes rather than
restarts, and that the row still terminates.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(lark): derive media object keys per chat message

The object key was derived from the platform message alone, so a second
ingest of the same Feishu message reused the first ingest's ledger row.
That row can be a tombstone (up to ~31h while the re-delete schedule
runs), and the intent upsert refuses anything that has left 'pending', so
the second ingest skipped the upload and silently produced a placeholder
with no attachment. A re-ingest is reachable: the inbound dedup claim is
reclaimable once 60s stale and the dedup row is only vacuumed after 24h.

Keying on the chat message the object will attach to keeps the two
ingests independent, and nothing leaks: each one's objects are covered by
its own ledger row.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* refactor(storage): route both local upload paths through one atomic write

UploadStream wrote through a temp file and renamed into place, but the
buffered Upload path still truncated the destination up front — the
destructive shape the stream path exists to avoid, one caller away from
coming back. Both now share writeAtomic, which also restores the 0644 the
direct write used (CreateTemp makes files 0600).

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* chore(channel): gofmt the media-pending append fields

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(storage): keep the local upload chmod best-effort

The rename-into-place rewrite made a failed chmod fail the whole upload.
CreateTemp's 0600 has to be widened to the 0644 the direct write used, but
an upload dir on a mount that ignores chmod (SMB/NFS/FUSE) accepted the
old direct write fine — turning those deployments' uploads into hard
errors would be a regression for a cosmetic property. Log and continue.

Tests pin 0644 on both upload paths, and that a failed buffered upload
leaves no temp litter and no damage to a previous object.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(channel): never re-delete an object an attachment references

The tombstone pass skipped the reference check and deleted unconditionally,
so a durable attachment carrying that URL lost the only object it can read
— the dangling attachment the intent ledger exists to prevent, and the
opposite of the posture every other path here takes ("a reclaimable orphan
beats a broken attachment").

The check now runs on every pass. A positive result on a tombstone is
unreachable by design — keys are per (chat message, resource) and a bind
cannot attach a key that has left 'pending' — so reaching it means an
invariant broke: keep the object, clear the row, log it, and count it on
a dedicated reconciler_tombstone_referenced_total counter. The test's
contract is flipped to assert the referenced object survives.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(storage): make the local staging file reclaimable after a crash

os.CreateTemp's random suffix meant a crash between the staging write and
the rename left a file nothing could name: the ledger records only the
final storage key, and DeleteObject removed only the object and its
sidecar. Each leftover can approach the 100 MiB resource cap and they
accumulate without bound.

The staging path is now derived from the object key, so DeleteObject
removes it alongside the object — which makes the media reconciler reclaim
it too, since the intent row is written before the upload. Opening it 0644
directly also drops the chmod the previous commit had to make best-effort.
Both read paths refuse the staging name (keys come from the request URL,
and a half-written body should not be readable); a user-supplied ".tmp"
extension is unaffected, since object keys are generated and never
dot-prefixed.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* chore(channel): renumber the media migrations after merging main

main took 224 (agent_task_session_rollout_missing), so the ledger group
moves to 225-230 and the cross-references inside the table migration follow.
main's CompleteTask also grew a sessionRolloutMissing parameter; the three
call sites this PR added to chat_input_ownership_test.go pass false.

Verified the way the numbering is meant to be verified: full migration set
applied from scratch on the merged tree, and the whole server suite run
against that database.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* fix(channel): let Postgres compute every reconciler deadline

The reconciler built settle cutoffs, lease expiry, backoff and re-delete
times from the process clock and compared them against the database's
now(). A replica whose clock had drifted would therefore settle rows whose
upload was still in flight (the object is deleted and the bind then refuses
to attach — media silently lost), hand out leases that are born expired
(rows churn between workers, attempt/backoff inflate), or compress the
tombstone schedule that fences a late-materializing PUT.

The four settle queries now take durations and derive their timestamps from
now(), so every replica reads one clock. The parameter types are the guard:
an app-side timestamp can no longer be passed. Test asserts the persisted
lease, backoff and re-delete deadlines all track the database's now().

The generated code also picks up main's new agent_task_queue column in the
two RETURNING task.* queries this PR adds.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>

* chore(lark): drop unrelated gofmt-only churn from this PR

Six files carried whitespace/comment-reformatting with no functional
change, unrelated to the inbound media pipeline. Reverting them to the
base revision keeps the diff focused on the feature (75 -> 69 files):

  server/internal/service/empty_claim_cache.go
  server/internal/integrations/lark/markdown_detect.go
  server/internal/integrations/lark/ws_chunk_assembler.go
  server/internal/integrations/lark/ws_chunk_assembler_test.go
  server/internal/integrations/lark/ws_frame_test.go
  server/internal/integrations/lark/registration_test.go

Verified: `git diff -w` against these files was already empty, so no
behavior is affected. go vet clean; tests covering these files pass.

Co-authored-by: multica-agent <github@multica.ai>

---------

Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
Co-authored-by: Bohan-J <bohan@devv.ai>
Co-authored-by: multica-agent <github@multica.ai>
2026-07-27 15:44:39 +08:00

1838 lines
70 KiB
Go

package handler
import (
"bytes"
"context"
"crypto/rand"
"crypto/rsa"
"crypto/x509"
"encoding/base64"
"encoding/json"
"encoding/pem"
"fmt"
"io"
"mime/multipart"
"net/http"
"net/http/httptest"
"net/url"
"os"
"path/filepath"
"strings"
"sync"
"testing"
"time"
"github.com/go-chi/chi/v5"
"github.com/multica-ai/multica/server/internal/auth"
"github.com/multica-ai/multica/server/internal/storage"
db "github.com/multica-ai/multica/server/pkg/db/generated"
)
// createHandlerTestChatSession seeds a chat_session row owned by testUserID
// targeting the given agent and returns the session UUID. Cleanup runs after
// the test. Used by attachment / chat tests that need an existing session.
func createHandlerTestChatSession(t *testing.T, agentID string) string {
t.Helper()
var sessionID string
if err := testPool.QueryRow(context.Background(), `
INSERT INTO chat_session (workspace_id, agent_id, creator_id, title, status)
VALUES ($1, $2, $3, $4, 'active')
RETURNING id
`, testWorkspaceID, agentID, testUserID, "Handler Test Chat Session").Scan(&sessionID); err != nil {
t.Fatalf("failed to create handler test chat session: %v", err)
}
t.Cleanup(func() {
testPool.Exec(context.Background(), `DELETE FROM chat_session WHERE id = $1`, sessionID)
})
return sessionID
}
// mockStorage is a tiny in-memory Storage stand-in. Upload records the bytes
// keyed by the storage key so GetReader can round-trip them in tests; KeyFromURL
// strips the synthetic CDN host so consumers can pass either the URL or the
// raw key.
type mockStorage struct {
mu sync.Mutex
files map[string][]byte
presignCalls []string
presignDispositions []string
}
func (m *mockStorage) Upload(_ context.Context, key string, data []byte, _ string, _ string) (string, error) {
m.mu.Lock()
defer m.mu.Unlock()
if m.files == nil {
m.files = map[string][]byte{}
}
m.files[key] = append([]byte(nil), data...)
return fmt.Sprintf("https://cdn.example.com/%s", key), nil
}
func (m *mockStorage) Delete(_ context.Context, key string) {
m.mu.Lock()
defer m.mu.Unlock()
delete(m.files, key)
}
func (m *mockStorage) DeleteObject(ctx context.Context, key string) error {
m.Delete(ctx, key)
return nil
}
func (m *mockStorage) DeleteKeys(_ context.Context, _ []string) {}
func (m *mockStorage) ObjectURL(key string) string {
return "https://cdn.example.com/" + key
}
func (m *mockStorage) KeyFromURL(rawURL string) string {
for _, prefix := range []string{
"https://cdn.example.com/",
"http://rustfs:9000/test-bucket/",
"https://s3.example.com/test-bucket/",
} {
if strings.HasPrefix(rawURL, prefix) {
return strings.TrimPrefix(rawURL, prefix)
}
}
return rawURL
}
func (m *mockStorage) CdnDomain() string { return "cdn.example.com" }
// mockStorageNoCdn is a mockStorage variant that returns an empty CdnDomain
// to simulate a private S3 / R2 / MinIO deployment where the operator has
// NOT configured a public-facing CDN domain. buildMarkdownURL must not
// persist `a.Url` for this shape — it would write a private bucket URL
// into markdown that no client can load.
type mockStorageNoCdn struct{ mockStorage }
func (m *mockStorageNoCdn) CdnDomain() string { return "" }
func (m *mockStorage) GetReader(_ context.Context, key string) (io.ReadCloser, error) {
m.mu.Lock()
defer m.mu.Unlock()
if data, ok := m.files[key]; ok {
return io.NopCloser(bytes.NewReader(data)), nil
}
return nil, fmt.Errorf("mockStorage GetReader: key not found: %q", key)
}
func (m *mockStorage) PresignGet(_ context.Context, key string, _ time.Duration) (string, error) {
m.mu.Lock()
defer m.mu.Unlock()
m.presignCalls = append(m.presignCalls, key)
return "https://signed.example.com/" + key + "?X-Amz-Signature=mock", nil
}
func (m *mockStorage) PresignGetWithContentDisposition(_ context.Context, key string, _ time.Duration, contentDisposition string) (string, error) {
m.mu.Lock()
defer m.mu.Unlock()
m.presignCalls = append(m.presignCalls, key)
m.presignDispositions = append(m.presignDispositions, contentDisposition)
u := url.URL{
Scheme: "https",
Host: "signed.example.com",
Path: "/" + key,
}
q := u.Query()
q.Set("X-Amz-Signature", "mock")
if contentDisposition != "" {
q.Set("response-content-disposition", contentDisposition)
}
u.RawQuery = q.Encode()
return u.String(), nil
}
func (m *mockStorage) put(key string, data []byte) {
m.mu.Lock()
defer m.mu.Unlock()
if m.files == nil {
m.files = map[string][]byte{}
}
m.files[key] = append([]byte(nil), data...)
}
// seekableReadCloser adapts a *bytes.Reader (which is an io.ReadSeeker) into an
// io.ReadCloser, mirroring what LocalStorage.GetReader returns (an *os.File is
// seekable). Used to exercise proxyAttachmentDownload's http.ServeContent path.
type seekableReadCloser struct{ *bytes.Reader }
func (seekableReadCloser) Close() error { return nil }
// seekableMockStorage is a mockStorage whose GetReader returns a seekable body,
// so proxyAttachmentDownload takes the http.ServeContent branch (the local-disk
// shape) instead of the manual single-range fallback.
type seekableMockStorage struct{ mockStorage }
func (m *seekableMockStorage) GetReader(_ context.Context, key string) (io.ReadCloser, error) {
m.mu.Lock()
defer m.mu.Unlock()
if data, ok := m.files[key]; ok {
return seekableReadCloser{bytes.NewReader(data)}, nil
}
return nil, fmt.Errorf("seekableMockStorage GetReader: key not found: %q", key)
}
// failingReader serves up to failAfter bytes and then errors on the next Read.
// It is forward-only (no Seek), so proxyAttachmentDownload routes it through the
// manual serveProxyRange path — letting us simulate a storage backend that dies
// mid-stream while the handler is discarding bytes to reach a Range start.
type failingReader struct {
data []byte
pos int
failAfter int
}
func (f *failingReader) Read(p []byte) (int, error) {
if f.pos >= f.failAfter {
return 0, fmt.Errorf("simulated storage read failure at offset %d", f.pos)
}
if remaining := f.failAfter - f.pos; len(p) > remaining {
p = p[:remaining]
}
n := copy(p, f.data[f.pos:])
f.pos += n
return n, nil
}
func (f *failingReader) Close() error { return nil }
// failingMockStorage returns a failingReader so the non-seekable Range path hits
// a read error partway through the skip-to-start CopyN.
type failingMockStorage struct {
mockStorage
failAfter int
}
func (m *failingMockStorage) GetReader(_ context.Context, key string) (io.ReadCloser, error) {
m.mu.Lock()
defer m.mu.Unlock()
if data, ok := m.files[key]; ok {
return &failingReader{data: data, failAfter: m.failAfter}, nil
}
return nil, fmt.Errorf("failingMockStorage GetReader: key not found: %q", key)
}
func TestUploadFileForeignWorkspace(t *testing.T) {
origStorage := testHandler.Storage
testHandler.Storage = &mockStorage{}
defer func() { testHandler.Storage = origStorage }()
var body bytes.Buffer
writer := multipart.NewWriter(&body)
part, err := writer.CreateFormFile("file", "test.txt")
if err != nil {
t.Fatal(err)
}
part.Write([]byte("hello world"))
writer.Close()
foreignWorkspaceID := "00000000-0000-0000-0000-000000000099"
req := httptest.NewRequest("POST", "/api/upload-file", &body)
req.Header.Set("Content-Type", writer.FormDataContentType())
req.Header.Set("X-User-ID", testUserID)
req.Header.Set("X-Workspace-ID", foreignWorkspaceID)
w := httptest.NewRecorder()
testHandler.UploadFile(w, req)
if w.Code != http.StatusForbidden {
t.Fatalf("UploadFile with foreign workspace: expected 403, got %d: %s", w.Code, w.Body.String())
}
}
// TestUploadFileResolvesWorkspaceViaSlugHeader is a regression test for the
// v2 workspace URL refactor (#1141). The frontend switched from sending
// X-Workspace-ID (UUID) to X-Workspace-Slug. For endpoints that sit outside
// the workspace middleware — like /api/upload-file — the handler-side
// resolver must accept the slug and translate it to a UUID, otherwise the
// handler silently falls through to the "no workspace context" branch and
// skips creating the DB attachment record. Files end up in S3 with no row
// in the attachment table, invisible to the UI.
func TestUploadFileResolvesWorkspaceViaSlugHeader(t *testing.T) {
origStorage := testHandler.Storage
testHandler.Storage = &mockStorage{}
defer func() { testHandler.Storage = origStorage }()
var body bytes.Buffer
writer := multipart.NewWriter(&body)
part, err := writer.CreateFormFile("file", "slug-upload.txt")
if err != nil {
t.Fatal(err)
}
part.Write([]byte("hello via slug"))
writer.Close()
req := httptest.NewRequest("POST", "/api/upload-file", &body)
req.Header.Set("Content-Type", writer.FormDataContentType())
req.Header.Set("X-User-ID", testUserID)
// Intentionally NOT setting X-Workspace-ID — post-v2 clients only send slug.
req.Header.Set("X-Workspace-Slug", handlerTestWorkspaceSlug)
w := httptest.NewRecorder()
testHandler.UploadFile(w, req)
if w.Code != http.StatusOK {
t.Fatalf("UploadFile with slug header: expected 200, got %d: %s", w.Code, w.Body.String())
}
// The workspace-aware branch returns the full AttachmentResponse (with
// id, workspace_id, uploader, etc.). The no-workspace-context branch
// returns only {filename, link}. Distinguish by checking the shape.
var resp map[string]any
if err := json.Unmarshal(w.Body.Bytes(), &resp); err != nil {
t.Fatalf("decode response: %v; body: %s", err, w.Body.String())
}
if _, ok := resp["id"]; !ok {
t.Fatalf("expected attachment response with 'id' field (DB row created); got fallback link-only response: %s", w.Body.String())
}
if gotWs, _ := resp["workspace_id"].(string); gotWs != testWorkspaceID {
t.Fatalf("attachment workspace_id mismatch: want %s, got %v", testWorkspaceID, resp["workspace_id"])
}
// Verify the row actually exists in the database.
var count int
if err := testPool.QueryRow(
context.Background(),
`SELECT count(*) FROM attachment WHERE workspace_id = $1 AND filename = $2`,
testWorkspaceID,
"slug-upload.txt",
).Scan(&count); err != nil {
t.Fatalf("query attachment count: %v", err)
}
if count != 1 {
t.Fatalf("attachment row count: want 1, got %d", count)
}
// Clean up so reruns don't accumulate rows.
if _, err := testPool.Exec(
context.Background(),
`DELETE FROM attachment WHERE workspace_id = $1 AND filename = $2`,
testWorkspaceID,
"slug-upload.txt",
); err != nil {
t.Fatalf("cleanup attachment: %v", err)
}
}
// TestUploadFileResolvesWorkspaceViaIDHeaderStill confirms the legacy path
// (CLI / daemon clients sending X-Workspace-ID as a UUID) still works after
// the refactor. Prevents a regression in the CLI/daemon compat branch.
func TestUploadFileResolvesWorkspaceViaIDHeaderStill(t *testing.T) {
origStorage := testHandler.Storage
testHandler.Storage = &mockStorage{}
defer func() { testHandler.Storage = origStorage }()
var body bytes.Buffer
writer := multipart.NewWriter(&body)
part, err := writer.CreateFormFile("file", "uuid-upload.txt")
if err != nil {
t.Fatal(err)
}
part.Write([]byte("hello via uuid"))
writer.Close()
req := httptest.NewRequest("POST", "/api/upload-file", &body)
req.Header.Set("Content-Type", writer.FormDataContentType())
req.Header.Set("X-User-ID", testUserID)
req.Header.Set("X-Workspace-ID", testWorkspaceID)
w := httptest.NewRecorder()
testHandler.UploadFile(w, req)
if w.Code != http.StatusOK {
t.Fatalf("UploadFile with UUID header: expected 200, got %d: %s", w.Code, w.Body.String())
}
// Clean up.
if _, err := testPool.Exec(
context.Background(),
`DELETE FROM attachment WHERE workspace_id = $1 AND filename = $2`,
testWorkspaceID,
"uuid-upload.txt",
); err != nil {
t.Fatalf("cleanup attachment: %v", err)
}
}
// TestUploadFile_AttachesToChatSession verifies that a multipart upload with
// a chat_session_id form field creates an attachment row linked to that chat
// session (chat_message_id remains NULL — it is back-filled on send).
func TestUploadFile_AttachesToChatSession(t *testing.T) {
origStorage := testHandler.Storage
testHandler.Storage = &mockStorage{}
defer func() { testHandler.Storage = origStorage }()
agentID := createHandlerTestAgent(t, "ChatUploadAgent", []byte("[]"))
sessionID := createHandlerTestChatSession(t, agentID)
var body bytes.Buffer
writer := multipart.NewWriter(&body)
part, err := writer.CreateFormFile("file", "chat-upload.png")
if err != nil {
t.Fatal(err)
}
// Minimal PNG signature so content-type sniffs as image/png.
part.Write([]byte("\x89PNG\r\n\x1a\nrest-of-bytes"))
if err := writer.WriteField("chat_session_id", sessionID); err != nil {
t.Fatal(err)
}
writer.Close()
req := httptest.NewRequest("POST", "/api/upload-file", &body)
req.Header.Set("Content-Type", writer.FormDataContentType())
req.Header.Set("X-User-ID", testUserID)
req.Header.Set("X-Workspace-ID", testWorkspaceID)
w := httptest.NewRecorder()
testHandler.UploadFile(w, req)
if w.Code != http.StatusOK {
t.Fatalf("UploadFile with chat_session_id: expected 200, got %d: %s", w.Code, w.Body.String())
}
var resp AttachmentResponse
if err := json.Unmarshal(w.Body.Bytes(), &resp); err != nil {
t.Fatalf("decode response: %v; body: %s", err, w.Body.String())
}
if resp.ChatSessionID == nil || *resp.ChatSessionID != sessionID {
t.Fatalf("chat_session_id in response: want %s, got %v", sessionID, resp.ChatSessionID)
}
if resp.ChatMessageID != nil {
t.Fatalf("chat_message_id should be NULL before send, got %v", resp.ChatMessageID)
}
if resp.IssueID != nil || resp.CommentID != nil {
t.Fatalf("issue_id/comment_id should be NULL for chat-only upload: %+v", resp)
}
if resp.URL == "" {
t.Fatal("expected non-empty url")
}
// Verify the DB row directly.
var dbSession, dbMessage *string
if err := testPool.QueryRow(
context.Background(),
`SELECT chat_session_id::text, chat_message_id::text FROM attachment WHERE id = $1`,
resp.ID,
).Scan(&dbSession, &dbMessage); err != nil {
t.Fatalf("query attachment row: %v", err)
}
if dbSession == nil || *dbSession != sessionID {
t.Fatalf("DB chat_session_id mismatch: want %s, got %v", sessionID, dbSession)
}
if dbMessage != nil {
t.Fatalf("DB chat_message_id should be NULL, got %v", dbMessage)
}
t.Cleanup(func() {
testPool.Exec(context.Background(), `DELETE FROM attachment WHERE id = $1`, resp.ID)
})
}
// TestUploadFile_RejectsForeignChatSession verifies a chat_session in another
// workspace (or owned by another user) is rejected with 403/404, preventing
// cross-tenant attachment binding.
func TestUploadFile_RejectsForeignChatSession(t *testing.T) {
origStorage := testHandler.Storage
testHandler.Storage = &mockStorage{}
defer func() { testHandler.Storage = origStorage }()
var body bytes.Buffer
writer := multipart.NewWriter(&body)
part, _ := writer.CreateFormFile("file", "evil.txt")
part.Write([]byte("payload"))
// Random non-existent UUID.
writer.WriteField("chat_session_id", "00000000-0000-0000-0000-0000deadbeef")
writer.Close()
req := httptest.NewRequest("POST", "/api/upload-file", &body)
req.Header.Set("Content-Type", writer.FormDataContentType())
req.Header.Set("X-User-ID", testUserID)
req.Header.Set("X-Workspace-ID", testWorkspaceID)
w := httptest.NewRecorder()
testHandler.UploadFile(w, req)
if w.Code != http.StatusNotFound && w.Code != http.StatusForbidden && w.Code != http.StatusBadRequest {
t.Fatalf("UploadFile with unknown chat_session_id: expected 4xx, got %d: %s", w.Code, w.Body.String())
}
}
// ---------------------------------------------------------------------------
// GetAttachmentContent tests (preview proxy)
// ---------------------------------------------------------------------------
// seedPreviewAttachment inserts an attachment row + writes the bytes into the
// active mockStorage. Returns the new attachment id. Caller is responsible for
// installing the mockStorage on testHandler before calling.
func seedPreviewAttachment(t *testing.T, store *mockStorage, key, filename, contentType string, body []byte) string {
t.Helper()
// Register the body so GetReader can find it via KeyFromURL → key.
url, err := store.Upload(context.Background(), key, body, contentType, filename)
if err != nil {
t.Fatalf("seed Upload: %v", err)
}
var id string
if err := testPool.QueryRow(context.Background(), `
INSERT INTO attachment (workspace_id, uploader_type, uploader_id, filename, url, content_type, size_bytes)
VALUES ($1, 'member', $2, $3, $4, $5, $6)
RETURNING id::text
`, testWorkspaceID, testUserID, filename, url, contentType, len(body)).Scan(&id); err != nil {
t.Fatalf("seed attachment row: %v", err)
}
t.Cleanup(func() {
testPool.Exec(context.Background(), `DELETE FROM attachment WHERE id = $1`, id)
})
return id
}
func seedAttachmentURL(t *testing.T, rawURL, filename, contentType string, sizeBytes int64) string {
t.Helper()
var id string
if err := testPool.QueryRow(context.Background(), `
INSERT INTO attachment (workspace_id, uploader_type, uploader_id, filename, url, content_type, size_bytes)
VALUES ($1, 'member', $2, $3, $4, $5, $6)
RETURNING id::text
`, testWorkspaceID, testUserID, filename, rawURL, contentType, sizeBytes).Scan(&id); err != nil {
t.Fatalf("seed attachment row: %v", err)
}
t.Cleanup(func() {
testPool.Exec(context.Background(), `DELETE FROM attachment WHERE id = $1`, id)
})
return id
}
func newPreviewRequest(t *testing.T, attachmentID, workspaceID string) (*http.Request, *httptest.ResponseRecorder) {
t.Helper()
req := httptest.NewRequest("GET", "/api/attachments/"+attachmentID+"/content", nil)
req.Header.Set("X-User-ID", testUserID)
req.Header.Set("X-Workspace-ID", workspaceID)
rctx := chi.NewRouteContext()
rctx.URLParams.Add("id", attachmentID)
req = req.WithContext(context.WithValue(req.Context(), chi.RouteCtxKey, rctx))
return req, httptest.NewRecorder()
}
func newDownloadRequest(t *testing.T, attachmentID, workspaceID string) (*http.Request, *httptest.ResponseRecorder) {
t.Helper()
req := httptest.NewRequest("GET", "/api/attachments/"+attachmentID+"/download", nil)
req.Header.Set("X-User-ID", testUserID)
req.Header.Set("X-Workspace-ID", workspaceID)
rctx := chi.NewRouteContext()
rctx.URLParams.Add("id", attachmentID)
req = req.WithContext(context.WithValue(req.Context(), chi.RouteCtxKey, rctx))
return req, httptest.NewRecorder()
}
// requireAttachmentDownloadHeaders asserts the security / disposition headers
// that proxyAttachmentDownload sets before choosing a Range branch are preserved
// on the final response, whether it went out as a 206 (partial) or a full 200.
// The Range/206 work must not drop Content-Type / Content-Disposition /
// Cache-Control: no-store / X-Content-Type-Options / preview CSP — the seekable
// (http.ServeContent) path in particular could silently clobber them.
func requireAttachmentDownloadHeaders(t *testing.T, header http.Header, wantFilename string) {
t.Helper()
if got, want := header.Get("Content-Type"), "application/octet-stream"; got != want {
t.Fatalf("Content-Type = %q, want %q", got, want)
}
if got := header.Get("Content-Disposition"); got == "" || !strings.Contains(got, wantFilename) {
t.Fatalf("Content-Disposition = %q, want non-empty containing %q", got, wantFilename)
}
if got, want := header.Get("Cache-Control"), "no-store"; got != want {
t.Fatalf("Cache-Control = %q, want %q", got, want)
}
if got, want := header.Get("X-Content-Type-Options"), "nosniff"; got != want {
t.Fatalf("X-Content-Type-Options = %q, want %q", got, want)
}
requireAttachmentPreviewCSP(t, header)
}
func requireAttachmentPreviewCSP(t *testing.T, header http.Header, extraAncestors ...string) {
t.Helper()
csp := header.Get("Content-Security-Policy")
if csp == "" {
t.Fatal("Content-Security-Policy header is missing")
}
for _, directive := range []string{
"default-src 'none'",
"frame-ancestors 'self'",
"object-src 'none'",
"base-uri 'none'",
"form-action 'none'",
} {
if !strings.Contains(csp, directive) {
t.Fatalf("Content-Security-Policy missing %q; got %q", directive, csp)
}
}
for _, ancestor := range extraAncestors {
if !strings.Contains(csp, ancestor) {
t.Fatalf("Content-Security-Policy missing frame ancestor %q; got %q", ancestor, csp)
}
}
if strings.Contains(csp, "frame-ancestors 'none'") {
t.Fatalf("Content-Security-Policy still blocks same-origin previews: %q", csp)
}
}
func TestAttachmentPreviewCSPHeader_AllowsConfiguredFrontendOrigins(t *testing.T) {
csp := attachmentPreviewCSPHeader([]string{
"https://app.example.test",
" https://App.Example.Test/some/path ",
"http://localhost:3000",
"*",
"javascript:alert(1)",
"not a url",
})
for _, want := range []string{
"frame-ancestors 'self' https://app.example.test http://localhost:3000",
"default-src 'none'",
"object-src 'none'",
} {
if !strings.Contains(csp, want) {
t.Fatalf("Content-Security-Policy missing %q; got %q", want, csp)
}
}
for _, reject := range []string{"*", "javascript:", "not a url", "some/path"} {
if strings.Contains(csp, reject) {
t.Fatalf("Content-Security-Policy includes rejected source %q; got %q", reject, csp)
}
}
if strings.Count(csp, "https://app.example.test") != 1 {
t.Fatalf("Content-Security-Policy should dedupe origins; got %q", csp)
}
}
func newDownloadRouter() http.Handler {
// Mirrors the production router after MUL-3130: the download
// route is registered under Auth-only with no
// RequireWorkspaceMember wrapper. The handler self-resolves the
// workspace from the attachment row and enforces membership
// internally, so a native browser <img>/<video> resource load
// with no X-Workspace-* headers is the supported call shape.
r := chi.NewRouter()
r.Get("/api/attachments/{id}/download", testHandler.DownloadAttachment)
return r
}
func testCloudFrontSigner(t *testing.T) *auth.CloudFrontSigner {
t.Helper()
key, err := rsa.GenerateKey(rand.Reader, 2048)
if err != nil {
t.Fatalf("generate CloudFront test key: %v", err)
}
pemBytes := pem.EncodeToMemory(&pem.Block{
Type: "RSA PRIVATE KEY",
Bytes: x509.MarshalPKCS1PrivateKey(key),
})
t.Setenv("CLOUDFRONT_KEY_PAIR_ID", "KTEST")
t.Setenv("CLOUDFRONT_DOMAIN", "static.example.test")
t.Setenv("COOKIE_DOMAIN", ".example.test")
t.Setenv("CLOUDFRONT_PRIVATE_KEY", base64.StdEncoding.EncodeToString(pemBytes))
t.Setenv("CLOUDFRONT_PRIVATE_KEY_SECRET", "")
signer := auth.NewCloudFrontSignerFromEnv()
if signer == nil {
t.Fatal("expected CloudFront signer")
}
return signer
}
func TestAttachmentToResponse_NonCloudFrontUsesDownloadEndpoint(t *testing.T) {
origSigner := testHandler.CFSigner
testHandler.CFSigner = nil
t.Cleanup(func() { testHandler.CFSigner = origSigner })
id := seedAttachmentURL(t, "http://rustfs:9000/test-bucket/private.txt", "private.txt", "text/plain", 5)
att, err := testHandler.Queries.GetAttachment(context.Background(), db.GetAttachmentParams{
ID: parseUUID(id),
WorkspaceID: parseUUID(testWorkspaceID),
})
if err != nil {
t.Fatalf("GetAttachment: %v", err)
}
resp := testHandler.attachmentToResponse(att)
if resp.URL != "http://rustfs:9000/test-bucket/private.txt" {
t.Fatalf("stored url changed: %q", resp.URL)
}
if resp.DownloadURL != "/api/attachments/"+id+"/download" {
t.Fatalf("download_url = %q, want unified endpoint", resp.DownloadURL)
}
}
func TestGetAttachmentByID_AutoPublicEndpointReturnsPresignedDownloadURL(t *testing.T) {
store := &mockStorageNoCdn{}
origStorage := testHandler.Storage
origCfg := testHandler.cfg
origSigner := testHandler.CFSigner
testHandler.Storage = store
testHandler.cfg.AttachmentDownloadMode = "auto"
testHandler.CFSigner = nil
t.Cleanup(func() {
testHandler.Storage = origStorage
testHandler.cfg = origCfg
testHandler.CFSigner = origSigner
})
key := "downloads/desktop-inline.png"
id := seedAttachmentURL(
t,
"https://s3.example.com/test-bucket/"+key,
"desktop-inline.png",
"image/png",
10,
)
req := httptest.NewRequest("GET", "/api/attachments/"+id, nil)
req.Header.Set("X-User-ID", testUserID)
req.Header.Set("X-Workspace-ID", testWorkspaceID)
rctx := chi.NewRouteContext()
rctx.URLParams.Add("id", id)
req = req.WithContext(context.WithValue(req.Context(), chi.RouteCtxKey, rctx))
w := httptest.NewRecorder()
testHandler.GetAttachmentByID(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200; body=%s", w.Code, w.Body.String())
}
var resp AttachmentResponse
if err := json.Unmarshal(w.Body.Bytes(), &resp); err != nil {
t.Fatalf("decode response: %v; body=%s", err, w.Body.String())
}
parsed, err := url.Parse(resp.DownloadURL)
if err != nil {
t.Fatalf("parse download_url: %v", err)
}
if got := parsed.Query().Get("X-Amz-Signature"); got != "mock" {
t.Fatalf("download_url = %q, want S3 presigned URL", resp.DownloadURL)
}
if got := parsed.Query().Get("response-content-disposition"); got != "" {
t.Fatalf("response-content-disposition = %q, want inline-loadable URL", got)
}
if want := "/api/attachments/" + id + "/download"; resp.MarkdownURL != want {
t.Fatalf("markdown_url = %q, want stable URL %q", resp.MarkdownURL, want)
}
if len(store.presignCalls) != 1 || store.presignCalls[0] != key {
t.Fatalf("presign calls = %v, want [%s]", store.presignCalls, key)
}
}
func TestDownloadAttachment_CloudFrontRedirectSignsAttachmentDisposition(t *testing.T) {
origStorage := testHandler.Storage
origCfg := testHandler.cfg
origSigner := testHandler.CFSigner
testHandler.Storage = &mockStorage{}
testHandler.cfg.AttachmentDownloadMode = "cloudfront"
testHandler.cfg.AttachmentFrameAncestors = []string{"https://app.example.test"}
testHandler.CFSigner = testCloudFrontSigner(t)
t.Cleanup(func() {
testHandler.Storage = origStorage
testHandler.cfg = origCfg
testHandler.CFSigner = origSigner
})
id := seedAttachmentURL(t, "https://static.example.test/downloads/cloudfront.md", "cloud front.md", "text/markdown", 10)
req, w := newDownloadRequest(t, id, testWorkspaceID)
w.Header().Set("Content-Security-Policy", "default-src 'self'; frame-ancestors 'none'")
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusFound {
t.Fatalf("status = %d, want 302; body=%s", w.Code, w.Body.String())
}
loc := w.Header().Get("Location")
parsed, err := url.Parse(loc)
if err != nil {
t.Fatalf("parse Location: %v", err)
}
if got := parsed.Query().Get("response-content-disposition"); got != `attachment; filename="cloud front.md"` {
t.Fatalf("response-content-disposition = %q", got)
}
if got := parsed.Query().Get("Key-Pair-Id"); got != "KTEST" {
t.Fatalf("Key-Pair-Id = %q", got)
}
requireAttachmentPreviewCSP(t, w.Header(), "https://app.example.test")
}
func TestDownloadAttachment_BareNavigationWithWorkspaceSlugQueryPassesMiddleware(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
origCfg := testHandler.cfg
origSigner := testHandler.CFSigner
testHandler.Storage = store
testHandler.cfg.AttachmentDownloadMode = "proxy"
testHandler.CFSigner = nil
t.Cleanup(func() {
testHandler.Storage = origStorage
testHandler.cfg = origCfg
testHandler.CFSigner = origSigner
})
key := "downloads/bare-nav.txt"
body := []byte("download body")
store.put(key, body)
id := seedAttachmentURL(t, "https://s3.example.com/test-bucket/"+key, "bare-nav.txt", "text/plain", int64(len(body)))
req := httptest.NewRequest("GET", "/api/attachments/"+id+"/download?workspace_slug="+url.QueryEscape(handlerTestWorkspaceSlug), nil)
req.Header.Set("X-User-ID", testUserID)
w := httptest.NewRecorder()
newDownloadRouter().ServeHTTP(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200; body=%s", w.Code, w.Body.String())
}
if got := w.Body.String(); got != string(body) {
t.Fatalf("body = %q, want %q", got, body)
}
if req.Header.Get("X-Workspace-ID") != "" || req.Header.Get("X-Workspace-Slug") != "" {
t.Fatalf("bare navigation test must not set custom workspace headers")
}
}
// TestDownloadAttachment_BareNavigationServesMemberWithoutWorkspaceHeaders
// is the regression test for MUL-3130: a markdown image rendered as
// `<img src="/api/attachments/<id>/download">` produces a native browser
// resource load that cannot attach X-Workspace-Slug / X-Workspace-ID
// headers. After the fix the handler self-resolves the workspace from
// the attachment row, so a bare URL succeeds for a workspace member.
func TestDownloadAttachment_BareNavigationServesMemberWithoutWorkspaceHeaders(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
origCfg := testHandler.cfg
origSigner := testHandler.CFSigner
testHandler.Storage = store
testHandler.cfg.AttachmentDownloadMode = "proxy"
testHandler.CFSigner = nil
t.Cleanup(func() {
testHandler.Storage = origStorage
testHandler.cfg = origCfg
testHandler.CFSigner = origSigner
})
key := "downloads/bare-nav.txt"
body := []byte("download body")
store.put(key, body)
id := seedAttachmentURL(t, "https://s3.example.com/test-bucket/"+key, "bare-nav.txt", "text/plain", int64(len(body)))
// Bare URL — no workspace_slug / workspace_id query, no
// X-Workspace-* headers. This is what a browser <img> tag emits
// when the markdown stores `/api/attachments/<id>/download`.
req := httptest.NewRequest("GET", "/api/attachments/"+id+"/download", nil)
req.Header.Set("X-User-ID", testUserID)
w := httptest.NewRecorder()
newDownloadRouter().ServeHTTP(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200; body=%s", w.Code, w.Body.String())
}
if got := w.Body.String(); got != string(body) {
t.Fatalf("body = %q, want %q", got, body)
}
if req.Header.Get("X-Workspace-ID") != "" || req.Header.Get("X-Workspace-Slug") != "" {
t.Fatalf("bare navigation test must not set custom workspace headers")
}
}
// TestDownloadAttachment_BareNavigationDeniesNonMemberWith404 covers the
// IDOR boundary: a stray attachment ID belonging to a workspace the
// requester is NOT a member of must return 404, not 200 (would leak
// bytes) and not 403 (would confirm the ID exists). Mirrors
// ServeLocalUpload's deny shape.
func TestDownloadAttachment_BareNavigationDeniesNonMemberWith404(t *testing.T) {
if testPool == nil {
t.Skip("test database not available")
}
store := &mockStorage{}
origStorage := testHandler.Storage
origCfg := testHandler.cfg
origSigner := testHandler.CFSigner
testHandler.Storage = store
testHandler.cfg.AttachmentDownloadMode = "proxy"
testHandler.CFSigner = nil
t.Cleanup(func() {
testHandler.Storage = origStorage
testHandler.cfg = origCfg
testHandler.CFSigner = origSigner
})
// Seed an attachment that lives in a workspace testUserID is NOT
// a member of. The workspace row has to exist so the FK on
// attachment.workspace_id resolves; we tear both down on
// cleanup.
ctx := context.Background()
var foreignWorkspaceID string
if err := testPool.QueryRow(ctx, `
INSERT INTO workspace (name, slug, description, issue_prefix)
VALUES ('Bare-Nav Foreign', 'bare-nav-foreign', '', 'BNF')
RETURNING id::text
`).Scan(&foreignWorkspaceID); err != nil {
t.Fatalf("seed foreign workspace: %v", err)
}
t.Cleanup(func() { testPool.Exec(ctx, `DELETE FROM workspace WHERE id = $1`, foreignWorkspaceID) })
key := "downloads/bare-nav-foreign.txt"
store.put(key, []byte("foreign-body"))
var id string
if err := testPool.QueryRow(ctx, `
INSERT INTO attachment (workspace_id, uploader_type, uploader_id, filename, url, content_type, size_bytes)
VALUES ($1, 'member', $2, $3, $4, $5, $6)
RETURNING id::text
`, foreignWorkspaceID, testUserID, "foreign.txt", "https://s3.example.com/test-bucket/"+key, "text/plain", 12).Scan(&id); err != nil {
t.Fatalf("seed foreign attachment: %v", err)
}
t.Cleanup(func() { testPool.Exec(ctx, `DELETE FROM attachment WHERE id = $1`, id) })
req := httptest.NewRequest("GET", "/api/attachments/"+id+"/download", nil)
req.Header.Set("X-User-ID", testUserID)
w := httptest.NewRecorder()
newDownloadRouter().ServeHTTP(w, req)
if w.Code != http.StatusNotFound {
t.Fatalf("status = %d, want 404 for non-member; body=%s", w.Code, w.Body.String())
}
if strings.Contains(w.Body.String(), "foreign-body") {
t.Fatalf("response body leaked file contents: %q", w.Body.String())
}
}
func TestDownloadAttachment_AutoInternalEndpointProxies(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
origCfg := testHandler.cfg
origSigner := testHandler.CFSigner
testHandler.Storage = store
testHandler.cfg.AttachmentDownloadMode = "auto"
testHandler.cfg.AttachmentFrameAncestors = []string{"https://app.example.test"}
testHandler.CFSigner = nil
t.Cleanup(func() {
testHandler.Storage = origStorage
testHandler.cfg = origCfg
testHandler.CFSigner = origSigner
})
key := "downloads/proxy-private.txt"
body := []byte("private object")
store.put(key, body)
id := seedAttachmentURL(t, "http://rustfs:9000/test-bucket/"+key, "report.txt", "text/plain", int64(len(body)))
req, w := newDownloadRequest(t, id, testWorkspaceID)
w.Header().Set("Content-Security-Policy", "default-src 'self'; frame-ancestors 'none'")
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200; body=%s", w.Code, w.Body.String())
}
if got := w.Body.String(); got != string(body) {
t.Fatalf("body = %q, want %q", got, body)
}
if got := w.Header().Get("Location"); got != "" {
t.Fatalf("Location should be empty for proxy download, got %q", got)
}
if got := w.Header().Get("Content-Disposition"); got != `attachment; filename="report.txt"` {
t.Fatalf("Content-Disposition = %q", got)
}
if got := w.Header().Get("Cache-Control"); got != "no-store" {
t.Fatalf("Cache-Control = %q, want no-store", got)
}
if got := w.Header().Get("X-Content-Type-Options"); got != "nosniff" {
t.Fatalf("X-Content-Type-Options = %q, want nosniff", got)
}
requireAttachmentPreviewCSP(t, w.Header(), "https://app.example.test")
if len(store.presignCalls) != 0 {
t.Fatalf("internal endpoint should not presign, calls=%v", store.presignCalls)
}
}
func TestDownloadAttachment_AutoPublicEndpointPresigns(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
origCfg := testHandler.cfg
origSigner := testHandler.CFSigner
testHandler.Storage = store
testHandler.cfg.AttachmentDownloadMode = "auto"
testHandler.cfg.AttachmentFrameAncestors = []string{"https://app.example.test"}
testHandler.CFSigner = nil
t.Cleanup(func() {
testHandler.Storage = origStorage
testHandler.cfg = origCfg
testHandler.CFSigner = origSigner
})
key := "downloads/public-private.txt"
id := seedAttachmentURL(t, "https://s3.example.com/test-bucket/"+key, "public.txt", "text/plain", 10)
req, w := newDownloadRequest(t, id, testWorkspaceID)
w.Header().Set("Content-Security-Policy", "default-src 'self'; frame-ancestors 'none'")
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusFound {
t.Fatalf("status = %d, want 302; body=%s", w.Code, w.Body.String())
}
loc := w.Header().Get("Location")
if !strings.Contains(loc, "X-Amz-Signature=mock") {
t.Fatalf("Location = %q, want fake S3 signature", loc)
}
parsed, err := url.Parse(loc)
if err != nil {
t.Fatalf("parse Location: %v", err)
}
if got := parsed.Query().Get("response-content-disposition"); got != `attachment; filename="public.txt"` {
t.Fatalf("response-content-disposition = %q", got)
}
if len(store.presignCalls) != 1 || store.presignCalls[0] != key {
t.Fatalf("presign calls = %v, want [%s]", store.presignCalls, key)
}
if len(store.presignDispositions) != 1 || store.presignDispositions[0] != `attachment; filename="public.txt"` {
t.Fatalf("presign dispositions = %v", store.presignDispositions)
}
requireAttachmentPreviewCSP(t, w.Header(), "https://app.example.test")
}
func TestDownloadAttachment_ExplicitProxyStreamsPublicEndpoint(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
origCfg := testHandler.cfg
origSigner := testHandler.CFSigner
testHandler.Storage = store
testHandler.cfg.AttachmentDownloadMode = "proxy"
testHandler.CFSigner = nil
t.Cleanup(func() {
testHandler.Storage = origStorage
testHandler.cfg = origCfg
testHandler.CFSigner = origSigner
})
key := "downloads/forced-proxy.png"
body := []byte("\x89PNG\r\n\x1a\nimage")
store.put(key, body)
id := seedAttachmentURL(t, "https://s3.example.com/test-bucket/"+key, "image.png", "image/png", int64(len(body)))
req, w := newDownloadRequest(t, id, testWorkspaceID)
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200; body=%s", w.Code, w.Body.String())
}
if got := w.Body.Bytes(); !bytes.Equal(got, body) {
t.Fatalf("body mismatch: got %q want %q", got, body)
}
if got := w.Header().Get("Content-Disposition"); got != `inline; filename="image.png"` {
t.Fatalf("Content-Disposition = %q", got)
}
requireAttachmentPreviewCSP(t, w.Header())
if len(store.presignCalls) != 0 {
t.Fatalf("forced proxy should not presign, calls=%v", store.presignCalls)
}
}
// setProxyDownloadHandler swaps the handler into forced-proxy mode with the
// given storage backend and returns the seeded attachment id + body. It wires
// t.Cleanup to restore the original handler fields.
func setProxyDownloadHandler(t *testing.T, store storage.Storage, body []byte) string {
t.Helper()
origStorage := testHandler.Storage
origCfg := testHandler.cfg
origSigner := testHandler.CFSigner
testHandler.Storage = store
testHandler.cfg.AttachmentDownloadMode = "proxy"
testHandler.CFSigner = nil
t.Cleanup(func() {
testHandler.Storage = origStorage
testHandler.cfg = origCfg
testHandler.CFSigner = origSigner
})
key := "downloads/range-sample.bin"
if putter, ok := store.(interface{ put(string, []byte) }); ok {
putter.put(key, body)
} else {
t.Fatalf("store %T does not support put()", store)
}
return seedAttachmentURL(t, "https://s3.example.com/test-bucket/"+key, "sample.bin", "application/octet-stream", int64(len(body)))
}
// rangeBody is a deterministic 4 KiB payload used by the Range tests so we can
// assert that a partial response's bytes match the corresponding slice of the
// full object.
func rangeBody() []byte {
b := make([]byte, 4096)
for i := range b {
b[i] = byte(i % 251) // 251 is prime → no alignment with 256 boundaries
}
return b
}
// runProxyRangeMatrix exercises the three acceptance paths (Range hit, 416,
// no-Range) against whichever storage backend is supplied, so both the
// http.ServeContent (seekable) and manual (non-seekable) branches are covered
// by identical assertions.
func runProxyRangeMatrix(t *testing.T, newStore func() storage.Storage) {
body := rangeBody()
t.Run("RangeHitReturns206", func(t *testing.T) {
id := setProxyDownloadHandler(t, newStore(), body)
req, w := newDownloadRequest(t, id, testWorkspaceID)
req.Header.Set("Range", "bytes=0-1023")
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusPartialContent {
t.Fatalf("status = %d, want 206; body=%s", w.Code, w.Body.String())
}
if got, want := w.Header().Get("Accept-Ranges"), "bytes"; got != want {
t.Fatalf("Accept-Ranges = %q, want %q", got, want)
}
if got, want := w.Header().Get("Content-Range"), fmt.Sprintf("bytes 0-1023/%d", len(body)); got != want {
t.Fatalf("Content-Range = %q, want %q", got, want)
}
if got, want := w.Header().Get("Content-Length"), "1024"; got != want {
t.Fatalf("Content-Length = %q, want %q", got, want)
}
if got := w.Body.Bytes(); !bytes.Equal(got, body[:1024]) {
t.Fatalf("partial body mismatch: len(got)=%d, want first 1024 bytes of object", len(got))
}
requireAttachmentDownloadHeaders(t, w.Header(), "sample.bin")
})
t.Run("MidStreamRangeMatchesFullSlice", func(t *testing.T) {
id := setProxyDownloadHandler(t, newStore(), body)
req, w := newDownloadRequest(t, id, testWorkspaceID)
req.Header.Set("Range", "bytes=1000-1999")
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusPartialContent {
t.Fatalf("status = %d, want 206; body=%s", w.Code, w.Body.String())
}
if got, want := w.Header().Get("Content-Range"), fmt.Sprintf("bytes 1000-1999/%d", len(body)); got != want {
t.Fatalf("Content-Range = %q, want %q", got, want)
}
if got := w.Body.Bytes(); !bytes.Equal(got, body[1000:2000]) {
t.Fatalf("mid-stream range bytes do not match object[1000:2000]")
}
})
t.Run("SuffixRangeReturnsTail", func(t *testing.T) {
id := setProxyDownloadHandler(t, newStore(), body)
req, w := newDownloadRequest(t, id, testWorkspaceID)
req.Header.Set("Range", "bytes=-500")
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusPartialContent {
t.Fatalf("status = %d, want 206; body=%s", w.Code, w.Body.String())
}
start := len(body) - 500
if got, want := w.Header().Get("Content-Range"), fmt.Sprintf("bytes %d-%d/%d", start, len(body)-1, len(body)); got != want {
t.Fatalf("Content-Range = %q, want %q", got, want)
}
if got := w.Body.Bytes(); !bytes.Equal(got, body[start:]) {
t.Fatalf("suffix range bytes do not match object tail")
}
})
t.Run("UnsatisfiableRangeReturns416", func(t *testing.T) {
id := setProxyDownloadHandler(t, newStore(), body)
req, w := newDownloadRequest(t, id, testWorkspaceID)
req.Header.Set("Range", fmt.Sprintf("bytes=%d-%d", len(body)+10, len(body)+20))
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusRequestedRangeNotSatisfiable {
t.Fatalf("status = %d, want 416; body=%s", w.Code, w.Body.String())
}
if got, want := w.Header().Get("Content-Range"), fmt.Sprintf("bytes */%d", len(body)); got != want {
t.Fatalf("Content-Range = %q, want %q", got, want)
}
})
t.Run("NoRangeReturnsFull200", func(t *testing.T) {
id := setProxyDownloadHandler(t, newStore(), body)
req, w := newDownloadRequest(t, id, testWorkspaceID)
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200; body=%s", w.Code, w.Body.String())
}
if got, want := w.Header().Get("Accept-Ranges"), "bytes"; got != want {
t.Fatalf("Accept-Ranges = %q, want %q", got, want)
}
if got := w.Body.Bytes(); !bytes.Equal(got, body) {
t.Fatalf("full body mismatch: len(got)=%d, want %d", len(got), len(body))
}
requireAttachmentDownloadHeaders(t, w.Header(), "sample.bin")
})
}
// TestDownloadAttachment_ProxyRange_Seekable covers the http.ServeContent path
// taken when the storage backend returns a seekable reader (local disk).
func TestDownloadAttachment_ProxyRange_Seekable(t *testing.T) {
runProxyRangeMatrix(t, func() storage.Storage { return &seekableMockStorage{} })
}
// TestDownloadAttachment_ProxyRange_NonSeekable covers the manual single-range
// fallback taken when the backend returns a forward-only stream (S3/MinIO).
func TestDownloadAttachment_ProxyRange_NonSeekable(t *testing.T) {
runProxyRangeMatrix(t, func() storage.Storage { return &mockStorage{} })
}
// TestDownloadAttachment_NonSeekableRangeSkipFailureReturns502 is the must-fix
// regression (RAS-31 ①). When the storage read fails while the handler is
// skipping forward to the Range start — before any response header is written —
// it must reply with an honest 502, NOT net/http's default 200 OK + empty body.
// A resuming client reads that default 200 as "Range ignored, this is the full
// object", silently turning a transient storage error into a corrupt/empty
// download and defeating the resume feature itself.
func TestDownloadAttachment_NonSeekableRangeSkipFailureReturns502(t *testing.T) {
body := rangeBody()
// Die at offset 500, well before the Range start (1000), so the failure lands
// inside the skip CopyN rather than the payload copy.
store := &failingMockStorage{failAfter: 500}
id := setProxyDownloadHandler(t, store, body)
req, w := newDownloadRequest(t, id, testWorkspaceID)
req.Header.Set("Range", "bytes=1000-1999")
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusBadGateway {
t.Fatalf("status = %d, want 502; body=%s", w.Code, w.Body.String())
}
if !strings.Contains(w.Body.String(), "failed to read attachment range") {
t.Fatalf("502 body should carry the honest error, got %q", w.Body.String())
}
if got := w.Body.Bytes(); bytes.Contains(got, body[:64]) {
t.Fatalf("502 response must not leak object bytes; got %q", got)
}
}
// TestDownloadAttachment_NonSeekableMultiRangeServesFull200 is the should-fix
// (RAS-31 ②). A multi-range request the non-seekable path cannot serve must be
// ignored and answered with a full 200 (RFC 7233), not 416. Paired with the
// seekable test below, this proves the two backends no longer disagree on
// success vs failure for the same multi-range request.
func TestDownloadAttachment_NonSeekableMultiRangeServesFull200(t *testing.T) {
body := rangeBody()
id := setProxyDownloadHandler(t, &mockStorage{}, body)
req, w := newDownloadRequest(t, id, testWorkspaceID)
req.Header.Set("Range", "bytes=0-10,20-30")
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200 (multi-range ignored); body=%s", w.Code, w.Body.String())
}
if got := w.Body.Bytes(); !bytes.Equal(got, body) {
t.Fatalf("multi-range should serve full body: len(got)=%d, want %d", len(got), len(body))
}
if got := w.Header().Get("Content-Range"); got != "" {
t.Fatalf("full 200 response must not carry Content-Range, got %q", got)
}
if got, want := w.Header().Get("Accept-Ranges"), "bytes"; got != want {
t.Fatalf("Accept-Ranges = %q, want %q", got, want)
}
requireAttachmentDownloadHeaders(t, w.Header(), "sample.bin")
}
// TestDownloadAttachment_SeekableMultiRangeServes206 documents the other half of
// the ②-divergence: the seekable (http.ServeContent) path answers the same
// multi-range request with a successful 206 multipart. Together with the
// non-seekable 200 test above, it shows neither backend fails the request with
// 416 — the divergence the maintainer flagged is gone.
func TestDownloadAttachment_SeekableMultiRangeServes206(t *testing.T) {
body := rangeBody()
id := setProxyDownloadHandler(t, &seekableMockStorage{}, body)
req, w := newDownloadRequest(t, id, testWorkspaceID)
req.Header.Set("Range", "bytes=0-10,20-30")
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusPartialContent {
t.Fatalf("status = %d, want 206 (multipart); body len=%d", w.Code, w.Body.Len())
}
if ct := w.Header().Get("Content-Type"); !strings.HasPrefix(ct, "multipart/byteranges") {
t.Fatalf("Content-Type = %q, want multipart/byteranges", ct)
}
}
// TestDownloadAttachment_NonSeekableEmptyObjectRangeServesFull200 covers the
// zero-length nit the maintainer flagged during re-review (same class as RAS-31
// ②). A WELL-FORMED Range against a 0-byte attachment used to collapse to 416 on
// the non-seekable path (parseSingleByteRange → rangeUnsatisfiable) while the
// seekable http.ServeContent path ignores the Range and returns an empty 200.
// parseSingleByteRange now classifies a well-formed range against an empty object
// as rangeUnsupported (→ empty 200), so neither backend hard-fails it. Malformed
// ranges are covered by the 416 companion test below.
func TestDownloadAttachment_NonSeekableEmptyObjectRangeServesFull200(t *testing.T) {
// Both a start-based range and a suffix range must be ignored on an empty
// object; the suffix form ("bytes=-100" against size 0) is the exact case the
// maintainer referenced.
for _, rangeHeader := range []string{"bytes=0-", "bytes=-100", "bytes=0-99"} {
t.Run(rangeHeader, func(t *testing.T) {
id := setProxyDownloadHandler(t, &mockStorage{}, []byte{})
req, w := newDownloadRequest(t, id, testWorkspaceID)
req.Header.Set("Range", rangeHeader)
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200 (Range ignored on empty object); body=%s", w.Code, w.Body.String())
}
if got := w.Body.Len(); got != 0 {
t.Fatalf("empty object must serve an empty body, got %d bytes", got)
}
if got := w.Header().Get("Content-Length"); got != "0" {
t.Fatalf("Content-Length = %q, want %q for an empty 200", got, "0")
}
if got := w.Header().Get("Content-Range"); got != "" {
t.Fatalf("full 200 response must not carry Content-Range, got %q", got)
}
if got, want := w.Header().Get("Accept-Ranges"), "bytes"; got != want {
t.Fatalf("Accept-Ranges = %q, want %q", got, want)
}
requireAttachmentDownloadHeaders(t, w.Header(), "sample.bin")
})
}
}
// TestDownloadAttachment_NonSeekableEmptyObjectMalformedRangeReturns416 is the
// guard against over-correcting the zero-length nit: a MALFORMED / unknown-unit
// Range against a 0-byte object must still return 416, not be swallowed into a
// 200. stdlib http.ServeContent returns 416 for these even on an empty object
// (it rejects the range before the empty-content short-circuit), so returning
// 200 here would introduce a NEW backend-dependent divergence — the exact
// regression a blanket "total == 0 → 200" would cause.
func TestDownloadAttachment_NonSeekableEmptyObjectMalformedRangeReturns416(t *testing.T) {
for _, rangeHeader := range []string{"items=0-10", "bytes=abc-def", "bytes=100"} {
t.Run(rangeHeader, func(t *testing.T) {
id := setProxyDownloadHandler(t, &mockStorage{}, []byte{})
req, w := newDownloadRequest(t, id, testWorkspaceID)
req.Header.Set("Range", rangeHeader)
testHandler.DownloadAttachment(w, req)
if w.Code != http.StatusRequestedRangeNotSatisfiable {
t.Fatalf("status = %d, want 416 for malformed range on empty object; body=%s", w.Code, w.Body.String())
}
})
}
}
func TestParseSingleByteRange(t *testing.T) {
const size = 1000
tests := []struct {
name string
header string
size int64
wantOutcome rangeParseOutcome
wantStart int64
wantLength int64
}{
{"full closed range", "bytes=0-1023", 2000, rangeSatisfiable, 0, 1024},
{"open-ended range", "bytes=500-", size, rangeSatisfiable, 500, 500},
{"end clamped to eof", "bytes=900-5000", size, rangeSatisfiable, 900, 100},
{"suffix range", "bytes=-200", size, rangeSatisfiable, 800, 200},
{"suffix larger than size clamps", "bytes=-5000", size, rangeSatisfiable, 0, 1000},
{"single byte", "bytes=0-0", size, rangeSatisfiable, 0, 1},
{"last byte", "bytes=999-999", size, rangeSatisfiable, 999, 1},
{"start at eof unsatisfiable", "bytes=1000-1001", size, rangeUnsatisfiable, 0, 0},
{"start past eof unsatisfiable", "bytes=2000-", size, rangeUnsatisfiable, 0, 0},
{"end before start", "bytes=500-499", size, rangeUnsatisfiable, 0, 0},
{"missing unit", "0-100", size, rangeUnsatisfiable, 0, 0},
// Multi-range is a valid-but-unsupported form: ignored → full body (200),
// not 416. This is the must-fix that stops the non-seekable path from
// diverging from the seekable (ServeContent) path on multi-range.
{"multi-range ignored", "bytes=0-10,20-30", size, rangeUnsupported, 0, 0},
{"empty spec", "bytes=", size, rangeUnsatisfiable, 0, 0},
{"no dash", "bytes=100", size, rangeUnsatisfiable, 0, 0},
{"suffix zero", "bytes=-0", size, rangeUnsatisfiable, 0, 0},
{"negative garbage", "bytes=abc-def", size, rangeUnsatisfiable, 0, 0},
// Empty object (size == 0): a WELL-FORMED range has no bytes to satisfy it,
// so it is ignored → full (empty) 200, matching stdlib http.ServeContent
// rather than diverging with a 416. This covers the maintainer's
// zero-length nit. Genuinely malformed / unknown-unit ranges against an
// empty object still yield rangeUnsatisfiable (416), also matching stdlib.
{"suffix on empty object", "bytes=-100", 0, rangeUnsupported, 0, 0},
{"start range on empty object", "bytes=0-", 0, rangeUnsupported, 0, 0},
{"mid range on empty object", "bytes=5-10", 0, rangeUnsupported, 0, 0},
{"closed range on empty object", "bytes=0-99", 0, rangeUnsupported, 0, 0},
// stdlib checks start-vs-size before validating the end, so on an empty
// object it ignores (200) even a range with a bad end or reversed bounds;
// mirror that. The paired non-empty rows below lock the other half of the
// contract: once the start is IN range, the end IS validated → 416.
{"bad end on empty object", "bytes=0-abc", 0, rangeUnsupported, 0, 0},
{"reversed on empty object", "bytes=10-5", 0, rangeUnsupported, 0, 0},
{"bad end on non-empty", "bytes=0-abc", size, rangeUnsatisfiable, 0, 0},
{"reversed on non-empty", "bytes=10-5", size, rangeUnsatisfiable, 0, 0},
{"unknown unit on empty object", "items=0-10", 0, rangeUnsatisfiable, 0, 0},
{"malformed start on empty object", "bytes=abc-def", 0, rangeUnsatisfiable, 0, 0},
{"no dash on empty object", "bytes=100", 0, rangeUnsatisfiable, 0, 0},
}
for _, tc := range tests {
t.Run(tc.name, func(t *testing.T) {
start, length, outcome := parseSingleByteRange(tc.header, tc.size)
if outcome != tc.wantOutcome {
t.Fatalf("outcome = %v, want %v (start=%d length=%d)", outcome, tc.wantOutcome, start, length)
}
if outcome != rangeSatisfiable {
return
}
if start != tc.wantStart || length != tc.wantLength {
t.Fatalf("got (start=%d, length=%d), want (start=%d, length=%d)", start, length, tc.wantStart, tc.wantLength)
}
})
}
}
func TestShouldProxyAttachmentURL(t *testing.T) {
cases := []struct {
raw string
want bool
}{
{"http://rustfs:9000/test-bucket/file.txt", true},
{"http://localhost:9000/test-bucket/file.txt", true},
{"http://127.0.0.1:9000/test-bucket/file.txt", true},
{"http://10.0.2.15/test-bucket/file.txt", true},
{"https://minio.internal/test-bucket/file.txt", true},
{"/uploads/workspaces/abc/file.txt", true},
{"https://s3.example.com/test-bucket/file.txt", false},
{"https://bucket.s3.us-east-1.amazonaws.com/file.txt", false},
}
for _, tc := range cases {
t.Run(tc.raw, func(t *testing.T) {
if got := shouldProxyAttachmentURL(tc.raw); got != tc.want {
t.Fatalf("shouldProxyAttachmentURL(%q) = %v, want %v", tc.raw, got, tc.want)
}
})
}
}
func TestGetAttachmentContent_HappyPath_Markdown(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
testHandler.Storage = store
defer func() { testHandler.Storage = origStorage }()
body := []byte("# heading\n\nbody text\n")
id := seedPreviewAttachment(t, store, "preview-md-key.md", "preview.md", "text/markdown", body)
req, w := newPreviewRequest(t, id, testWorkspaceID)
w.Header().Set("Content-Security-Policy", "default-src 'self'; frame-ancestors 'none'")
testHandler.GetAttachmentContent(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200; body=%s", w.Code, w.Body.String())
}
if got := w.Body.String(); got != string(body) {
t.Errorf("body = %q, want %q", got, body)
}
if got := w.Header().Get("Content-Type"); got != "text/plain; charset=utf-8" {
t.Errorf("Content-Type = %q, want text/plain; charset=utf-8", got)
}
if got := w.Header().Get("X-Original-Content-Type"); got != "text/markdown" {
t.Errorf("X-Original-Content-Type = %q, want text/markdown", got)
}
if got := w.Header().Get("X-Content-Type-Options"); got != "nosniff" {
t.Errorf("X-Content-Type-Options = %q, want nosniff", got)
}
requireAttachmentPreviewCSP(t, w.Header())
}
// Even when http.DetectContentType returned "text/plain" instead of "text/markdown"
// (a known sniffer quirk), the extension whitelist still grants access.
func TestGetAttachmentContent_AcceptsByExtensionWhenContentTypeIsGeneric(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
testHandler.Storage = store
defer func() { testHandler.Storage = origStorage }()
body := []byte("package main\n")
id := seedPreviewAttachment(t, store, "main-go-key.go", "main.go", "application/octet-stream", body)
req, w := newPreviewRequest(t, id, testWorkspaceID)
testHandler.GetAttachmentContent(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200; body=%s", w.Code, w.Body.String())
}
}
func TestGetAttachmentContent_Unsupported_PDF(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
testHandler.Storage = store
defer func() { testHandler.Storage = origStorage }()
id := seedPreviewAttachment(t, store, "pdf-key.pdf", "manual.pdf", "application/pdf", []byte("%PDF-1.4\n"))
req, w := newPreviewRequest(t, id, testWorkspaceID)
testHandler.GetAttachmentContent(w, req)
if w.Code != http.StatusUnsupportedMediaType {
t.Fatalf("status = %d, want 415; body=%s", w.Code, w.Body.String())
}
}
func TestGetAttachmentContent_TooLarge(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
testHandler.Storage = store
defer func() { testHandler.Storage = origStorage }()
// One byte over the limit. Allocate ASCII so io.ReadAll has work to do.
big := bytes.Repeat([]byte("a"), maxPreviewTextSize+1)
id := seedPreviewAttachment(t, store, "huge-key.txt", "huge.txt", "text/plain", big)
req, w := newPreviewRequest(t, id, testWorkspaceID)
testHandler.GetAttachmentContent(w, req)
if w.Code != http.StatusRequestEntityTooLarge {
t.Fatalf("status = %d, want 413; body=%s", w.Code, w.Body.String())
}
}
func TestGetAttachmentContent_ForeignWorkspace(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
testHandler.Storage = store
defer func() { testHandler.Storage = origStorage }()
id := seedPreviewAttachment(t, store, "ws-mismatch.md", "note.md", "text/markdown", []byte("# secret\n"))
// Same attachment id, but request comes in scoped to a different workspace.
foreign := "00000000-0000-0000-0000-000000000099"
req, w := newPreviewRequest(t, id, foreign)
testHandler.GetAttachmentContent(w, req)
if w.Code != http.StatusNotFound {
t.Fatalf("status = %d, want 404; body=%s", w.Code, w.Body.String())
}
}
func TestGetAttachmentContent_NotFound(t *testing.T) {
store := &mockStorage{}
origStorage := testHandler.Storage
testHandler.Storage = store
defer func() { testHandler.Storage = origStorage }()
req, w := newPreviewRequest(t, "00000000-0000-0000-0000-000000000abc", testWorkspaceID)
testHandler.GetAttachmentContent(w, req)
if w.Code != http.StatusNotFound {
t.Fatalf("status = %d, want 404; body=%s", w.Code, w.Body.String())
}
}
// isTextPreviewable is the whitelist linkpin between the proxy and the
// client-side dispatcher. Regress against the most common content types so
// drifting one of the lists alone fails loud.
func TestIsTextPreviewable(t *testing.T) {
t.Helper()
cases := []struct {
name string
contentType string
filename string
want bool
}{
{"markdown by ext", "application/octet-stream", "README.md", true},
{"markdown by mime", "text/markdown", "README", true},
{"plain text", "text/plain", "log.txt", true},
{"json by mime", "application/json", "data.json", true},
{"yaml by ext", "application/octet-stream", "config.yml", true},
{"go source", "text/plain", "main.go", true},
{"typescript", "application/octet-stream", "index.ts", true},
{"html", "text/html", "page.html", true},
{"dockerfile no ext", "application/octet-stream", "Dockerfile", true},
{"makefile no ext", "application/octet-stream", "Makefile", true},
{"env dotfile", "application/octet-stream", ".env", true},
{"gitignore dotfile", "application/octet-stream", ".gitignore", true},
{"dockerfile extension", "application/octet-stream", "service.dockerfile", true},
{"makefile extension", "application/octet-stream", "rules.makefile", true},
{"pdf rejected", "application/pdf", "doc.pdf", false},
{"png rejected", "image/png", "shot.png", false},
{"video rejected", "video/mp4", "clip.mp4", false},
{"binary fallthrough", "application/octet-stream", "blob.bin", false},
{"docx rejected", "application/vnd.openxmlformats-officedocument.wordprocessingml.document", "report.docx", false},
}
for _, tc := range cases {
t.Run(tc.name, func(t *testing.T) {
if got := isTextPreviewable(tc.contentType, tc.filename); got != tc.want {
t.Errorf("isTextPreviewable(%q, %q) = %v, want %v", tc.contentType, tc.filename, got, tc.want)
}
})
}
}
// MUL-3192 — buildMarkdownURL must emit a durable, absolute-when-possible
// URL that loads natively in any client (web, desktop, mobile webview).
// `download_url` may be a short-lived signed URL and is unsafe to persist;
// `markdown_url` is the contract for "ok to embed in markdown body".
//
// Matrix:
//
// - public CDN durable URL ............... reuse a.Url verbatim
// - LocalStorage with PublicURL set ....... reuse a.Url (already absolute)
// - CloudFront-signed mode ................ never reuse a.Url (raw S3),
// prefer absolute API endpoint
// - LocalStorage relative + PublicURL set . prefix to absolute API endpoint
// - PublicURL unset ....................... fall back to site-relative
// (web's Next rewrite handles it)
// - signed URL (CloudFront-signed leaked
// into a.Url somehow) ................... reject as durable, fall through
// to API endpoint to avoid
// re-opening MUL-3130
func TestBuildMarkdownURL_PublicCdnAbsoluteURLReusedVerbatim(t *testing.T) {
origPublic := testHandler.cfg.PublicURL
origSigner := testHandler.CFSigner
origStorage := testHandler.Storage
t.Cleanup(func() {
testHandler.cfg.PublicURL = origPublic
testHandler.CFSigner = origSigner
testHandler.Storage = origStorage
})
testHandler.cfg.PublicURL = "https://api.multica.test"
testHandler.CFSigner = nil
// mockStorage.CdnDomain() returns "cdn.example.com" — that's the
// operator-set signal that the URL host serves content publicly
// without per-request auth. Without this, the new gate routes
// through the API endpoint to be safe.
testHandler.Storage = &mockStorage{}
id := seedAttachmentURL(t, "https://cdn.multica.test/uploads/abc.png", "abc.png", "image/png", 1)
att, err := testHandler.Queries.GetAttachment(context.Background(), db.GetAttachmentParams{
ID: parseUUID(id),
WorkspaceID: parseUUID(testWorkspaceID),
})
if err != nil {
t.Fatalf("GetAttachment: %v", err)
}
resp := testHandler.attachmentToResponse(att)
if resp.MarkdownURL != "https://cdn.multica.test/uploads/abc.png" {
t.Fatalf("markdown_url = %q, want raw a.Url passthrough", resp.MarkdownURL)
}
}
// MUL-3192 review must-fix 1 — `att.url` for a private S3 / R2 / MinIO
// bucket is absolute https + unsigned but is NOT publicly readable. The
// generic "absolute http(s) without signature" check would have wrongly
// persisted it; the gate now also requires `Storage.CdnDomain()` to be
// set so the operator has explicitly opted into "URLs from this storage
// load directly".
func TestBuildMarkdownURL_PrivateBucketWithoutCdnDomainRoutesThroughAPIEndpoint(t *testing.T) {
origPublic := testHandler.cfg.PublicURL
origSigner := testHandler.CFSigner
origStorage := testHandler.Storage
t.Cleanup(func() {
testHandler.cfg.PublicURL = origPublic
testHandler.CFSigner = origSigner
testHandler.Storage = origStorage
})
testHandler.cfg.PublicURL = "https://api.multica.test"
testHandler.CFSigner = nil
testHandler.Storage = &mockStorageNoCdn{}
id := seedAttachmentURL(t, "https://prod.s3.amazonaws.com/key.png", "key.png", "image/png", 1)
att, err := testHandler.Queries.GetAttachment(context.Background(), db.GetAttachmentParams{
ID: parseUUID(id),
WorkspaceID: parseUUID(testWorkspaceID),
})
if err != nil {
t.Fatalf("GetAttachment: %v", err)
}
resp := testHandler.attachmentToResponse(att)
want := "https://api.multica.test/api/attachments/" + id + "/download"
if resp.MarkdownURL != want {
t.Fatalf("markdown_url = %q, want absolute API endpoint %q (private bucket without explicit CDN must not persist raw S3 URL)", resp.MarkdownURL, want)
}
}
func TestBuildMarkdownURL_CloudFrontSignedModeNeverPersistsRawStorageURL(t *testing.T) {
origPublic := testHandler.cfg.PublicURL
origSigner := testHandler.CFSigner
t.Cleanup(func() {
testHandler.cfg.PublicURL = origPublic
testHandler.CFSigner = origSigner
})
testHandler.cfg.PublicURL = "https://api.multica.test"
testHandler.CFSigner = testCloudFrontSigner(t)
// Raw S3 URL — private bucket, not loadable directly by clients.
id := seedAttachmentURL(t, "https://prod.s3.amazonaws.com/key.png", "key.png", "image/png", 1)
att, err := testHandler.Queries.GetAttachment(context.Background(), db.GetAttachmentParams{
ID: parseUUID(id),
WorkspaceID: parseUUID(testWorkspaceID),
})
if err != nil {
t.Fatalf("GetAttachment: %v", err)
}
resp := testHandler.attachmentToResponse(att)
want := "https://api.multica.test/api/attachments/" + id + "/download"
if resp.MarkdownURL != want {
t.Fatalf("markdown_url = %q, want absolute API endpoint %q", resp.MarkdownURL, want)
}
// download_url is allowed to carry a TTL (CloudFront-signed); it's NOT
// what the client persists, but it IS what the renderer uses for this
// response. The two are intentionally distinct.
if resp.DownloadURL == resp.MarkdownURL {
t.Fatalf("download_url and markdown_url must differ in CloudFront-signed mode (got identical %q)", resp.DownloadURL)
}
}
func TestBuildMarkdownURL_RelativeStorageURLPrefixedWithPublicURL(t *testing.T) {
origPublic := testHandler.cfg.PublicURL
origSigner := testHandler.CFSigner
t.Cleanup(func() {
testHandler.cfg.PublicURL = origPublic
testHandler.CFSigner = origSigner
})
testHandler.cfg.PublicURL = "https://api.multica.test"
testHandler.CFSigner = nil
// LocalStorage without LOCAL_UPLOAD_BASE_URL stores a site-relative URL.
id := seedAttachmentURL(t, "/uploads/abc.png", "abc.png", "image/png", 1)
att, err := testHandler.Queries.GetAttachment(context.Background(), db.GetAttachmentParams{
ID: parseUUID(id),
WorkspaceID: parseUUID(testWorkspaceID),
})
if err != nil {
t.Fatalf("GetAttachment: %v", err)
}
resp := testHandler.attachmentToResponse(att)
want := "https://api.multica.test/api/attachments/" + id + "/download"
if resp.MarkdownURL != want {
t.Fatalf("markdown_url = %q, want absolute API endpoint %q", resp.MarkdownURL, want)
}
}
func TestBuildMarkdownURL_PublicURLUnsetFallsBackToSiteRelative(t *testing.T) {
origPublic := testHandler.cfg.PublicURL
origSigner := testHandler.CFSigner
t.Cleanup(func() {
testHandler.cfg.PublicURL = origPublic
testHandler.CFSigner = origSigner
})
testHandler.cfg.PublicURL = ""
testHandler.CFSigner = nil
id := seedAttachmentURL(t, "/uploads/abc.png", "abc.png", "image/png", 1)
att, err := testHandler.Queries.GetAttachment(context.Background(), db.GetAttachmentParams{
ID: parseUUID(id),
WorkspaceID: parseUUID(testWorkspaceID),
})
if err != nil {
t.Fatalf("GetAttachment: %v", err)
}
resp := testHandler.attachmentToResponse(att)
want := "/api/attachments/" + id + "/download"
if resp.MarkdownURL != want {
t.Fatalf("markdown_url = %q, want site-relative fallback %q", resp.MarkdownURL, want)
}
}
func TestBuildMarkdownURL_StripsTrailingSlashOnPublicURL(t *testing.T) {
origPublic := testHandler.cfg.PublicURL
origSigner := testHandler.CFSigner
t.Cleanup(func() {
testHandler.cfg.PublicURL = origPublic
testHandler.CFSigner = origSigner
})
testHandler.cfg.PublicURL = "https://api.multica.test/"
testHandler.CFSigner = nil
id := seedAttachmentURL(t, "/uploads/abc.png", "abc.png", "image/png", 1)
att, err := testHandler.Queries.GetAttachment(context.Background(), db.GetAttachmentParams{
ID: parseUUID(id),
WorkspaceID: parseUUID(testWorkspaceID),
})
if err != nil {
t.Fatalf("GetAttachment: %v", err)
}
resp := testHandler.attachmentToResponse(att)
want := "https://api.multica.test/api/attachments/" + id + "/download"
if resp.MarkdownURL != want {
t.Fatalf("markdown_url = %q, want exactly one separator %q", resp.MarkdownURL, want)
}
}
func TestIsDurablePublicURL(t *testing.T) {
cases := []struct {
name string
url string
want bool
}{
{"absolute https no signature", "https://cdn.multica.test/foo.png", true},
{"absolute http no signature", "http://cdn.multica.test/foo.png", true},
{"absolute with port + path", "https://cdn.example.test:8080/a/b/c.png", true},
{"empty string", "", false},
{"site-relative", "/uploads/abc.png", false},
{"protocol-relative", "//cdn.example/foo.png", false},
{"data URL", "data:image/png;base64,abc", false},
{"blob URL", "blob:https://app/abc", false},
{"unsupported scheme", "ftp://server/foo", false},
{"cloudfront-signed Signature", "https://cdn.example/foo.png?Signature=abc&Key-Pair-Id=K1", false},
{"cloudfront-signed Key-Pair-Id alone", "https://cdn.example/foo.png?Key-Pair-Id=K1", false},
{"s3-presigned X-Amz-Signature", "https://bucket.s3/foo.png?X-Amz-Signature=abc", false},
{"s3-presigned X-Amz-Expires alone", "https://bucket.s3/foo.png?X-Amz-Expires=900", false},
{"plain Expires query", "https://cdn.example/foo.png?Expires=99", false},
{"unrelated query", "https://cdn.example/foo.png?cache=1", true},
}
for _, tc := range cases {
t.Run(tc.name, func(t *testing.T) {
if got := isDurablePublicURL(tc.url); got != tc.want {
t.Errorf("isDurablePublicURL(%q) = %v, want %v", tc.url, got, tc.want)
}
})
}
}
// TestServeLocalUpload_RelaxesFrameAncestorsForPreview covers the self-hosted
// local-disk case where document previews (PDF/HTML) are fetched straight from
// the public /uploads/* static route. That route inherits the global
// "frame-ancestors 'none'" CSP from the middleware, which blocks iframe
// previews; ServeLocalUpload must overwrite it with the same relaxed preview
// policy the /api/attachments download endpoint uses. See MUL-3821 / #4477.
func TestServeLocalUpload_RelaxesFrameAncestorsForPreview(t *testing.T) {
dir := t.TempDir()
key := "workspaces/ws-1/preview.pdf"
full := filepath.Join(dir, filepath.FromSlash(key))
if err := os.MkdirAll(filepath.Dir(full), 0o755); err != nil {
t.Fatalf("mkdir: %v", err)
}
const body = "%PDF-1.7 local-disk preview"
if err := os.WriteFile(full, []byte(body), 0o644); err != nil {
t.Fatalf("write file: %v", err)
}
t.Setenv("LOCAL_UPLOAD_DIR", dir)
t.Setenv("LOCAL_UPLOAD_BASE_URL", "")
local := storage.NewLocalStorageFromEnv()
if local == nil {
t.Fatal("NewLocalStorageFromEnv returned nil")
}
h := &Handler{
Storage: local,
cfg: Config{AttachmentFrameAncestors: []string{"https://app.example.test"}},
}
req := httptest.NewRequest(http.MethodGet, "/uploads/"+key, nil)
w := httptest.NewRecorder()
// Simulate the global CSP middleware having already stamped the strict
// policy on the response before the static route runs.
w.Header().Set("Content-Security-Policy", "default-src 'self'; frame-ancestors 'none'")
h.ServeLocalUpload(w, req)
if w.Code != http.StatusOK {
t.Fatalf("status = %d, want 200; body=%q", w.Code, w.Body.String())
}
if got := w.Body.String(); got != body {
t.Fatalf("body = %q, want %q", got, body)
}
requireAttachmentPreviewCSP(t, w.Header(), "https://app.example.test")
}
// TestServeLocalUpload_NonLocalStorage404 guards the defensive branch: the
// /uploads/* route is only registered under local storage, but the handler
// must not serve anything when the backing store is not local disk.
func TestServeLocalUpload_NonLocalStorage404(t *testing.T) {
h := &Handler{Storage: &mockStorage{}}
req := httptest.NewRequest(http.MethodGet, "/uploads/workspaces/ws-1/x.png", nil)
w := httptest.NewRecorder()
h.ServeLocalUpload(w, req)
if w.Code != http.StatusNotFound {
t.Fatalf("status = %d, want 404", w.Code)
}
}