WPB-24669 [fix] upload of files with umlaut when audit log enabled

[battermann]

https://wearezeta.atlassian.net/browse/WPB-24669

Checklist

• Add a new entry in an appropriate subdirectory of changelog.d
• Read and follow the PR guidelines

[Copilot]

Pull request overview

This PR fixes asset uploads failing for non-ASCII filenames (e.g. umlauts) when audit logging is enabled by percent-encoding the audit-log JSON metadata before putting it into S3 user-metadata headers, and decoding it when reading metadata back.

Changes:

• Percent-encode audit-log metadata for S3 headers and URL-decode it on read-back.
• Add a unit (QuickCheck) regression test ensuring produced S3 metadata header values are ASCII even with umlaut filenames.
• Wire the new test module into the cargohold unit test suite and add a changelog entry.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
services/cargohold/src/CargoHold/S3.hs	Encode audit-log metadata as percent-encoded ASCII for S3 headers; decode when reading metadata back.
services/cargohold/test/unit/Test/CargoHold/S3Test.hs	Adds a QuickCheck property asserting the metadata header value remains ASCII for umlaut filenames.
services/cargohold/test/unit/Main.hs	Registers the new S3 unit tests in the test runner.
services/cargohold/cargohold.cabal	Adds the new test module to the test-suite’s other-modules.
changelog.d/3-bug-fixes/WPB-24669	Documents the bug fix and the encoding/decoding approach.

[Copilot]

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated 1 comment.

Comments suppressed due to low confidence (1)

services/cargohold/src/CargoHold/S3.hs:431

• decodePercentEncoded uses decodeLatin1 on the percent-decoded bytes and then parseJSON re-encodes with encodeUtf8. For non-ASCII filenames this corrupts the original UTF-8 JSON bytes (bytes >= 0x80 become multi-byte UTF-8), so percent-encoded audit-log metadata cannot be decoded correctly.

A simpler and correct approach is to parse JSON directly from the strict ByteString (either the raw header bytes, or the urlDecoded bytes) and avoid the Text roundtrip.

    parseJSON :: Text -> Maybe AssetAuditLogMetadata
    parseJSON = A.decode . fromStrict . encodeUtf8

    decodePercentEncoded :: Text -> Text
    decodePercentEncoded = decodeLatin1 . HTTPURI.urlDecode False . encodeUtf8

[Copilot]

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated 2 comments.

[Copilot]

Pull request overview

Copilot reviewed 5 out of 5 changed files in this pull request and generated no new comments.