Bump zip_kit from 6.3.4 to 6.4.1

[dependabot]

Bumps zip_kit from 6.3.4 to 6.4.1.

Changelog

Sourced from zip_kit's changelog.

6.4.1

• Use lowercase names for the HTTP headers ZipKit produces, since Rack 3 requires them (Rack::Lint rejects uppercase header names). OutputEnumerator.streaming_http_headers now returns "content-type", "last-modified", "content-encoding" and "x-accel-buffering". This is a breaking change if you look those headers up or override them in the returned Hash using the capitalised names: headers["Content-Type"] is now nil, and headers["Content-Type"] = "application/epub+zip" adds a second content type instead of replacing ours. Use the lowercase names instead. zip_kit_stream in Rails is not affected: it now merges our headers into the response regardless of case, so headers Rails sets itself (like the content type) still take precedence and do not get sent twice.

6.4.0

• Add Streamer.new(io, ocf: true) for writing EPUB (OCF) and OpenDocument containers. The first entry must then be a stored mimetype file with its sizes in the local header, and Streamer::OCFViolation is raised otherwise. That entry is written without the extended timestamp extra field, which those formats forbid - before, they could not be produced with ZipKit at all. File names get checked against the OCF rules as well: valid UTF-8, no forbidden characters, segments of at most 255 bytes not ending with a ., and no two names in a directory which only differ by case or Unicode normalization.
• Add Streamer#write_mimetype_file and SizeEstimator#add_mimetype_entry
• Add the extended_timestamp: keyword argument to ZipWriter#write_local_file_header and ZipWriter#write_central_directory_file_header

6.3.6

• Fix the end of central directory record counting entries discarded by rollback!, which made archives with a rolled back entry unreadable for readers which check the entry count (including ZipKit::FileReader).
• Make ZipKit usable from within Ractors: freeze the computed String constants in ZipWriter, BlockDeflate and RackChunkedBody, and stop defining the raising BlockWrite methods with define_method. Ractor tests run on Ruby 4.0 in CI.

6.3.5

• Make WriteBuffer#<< much faster for lots of tiny writes, like the XML fragments written by libraries such as caxlsx. Strings are no longer copied with String#b before being appended: on Ruby 3.4+ String#append_as_bytes is used, on older Rubies the string is converted only if Ruby refuses to append it. Mixing binary strings and non-ASCII strings still works, and the writable still always receives binary strings.
• WriteBuffer now flushes before a write would make the buffer exceed its size (instead of after), so that it never outputs chunks larger than the buffer size, and writes which are larger than the buffer size are still passed through without getting copied.
• WriteBuffer no longer preallocates (and immediately discards) a zero-filled String of twice its buffer size, which makes writing ZIPs with lots of small files faster.
• Require stringio in ZipWriter, which uses it. Previously writing a ZIP could fail with a NameError if nothing else in the process had loaded stringio.
• Add bench/write_buffer_bench.rb

Commits

• 3537e44 Bump version
• 8f4b351 Use lowercase names for HTTP headers (#35)
• d8e0234 Bump version
• d5bde00 Allow writing EPUB and OpenDocument containers (#34)
• 73b6510 Avoid typedef regen on VERSION change
• 3c38e68 Update typedefs
• 40d48df Bump version
• ca3f2f2 Make zip_kit usable from within Ractors (#32)
• b92bf72 Do not count rolled back entries in the EOCD (#31)
• 544f314 Bump typedefs
• Additional commits viewable in compare view

[Copilot]

Copilot wasn't able to review any files in this pull request. Check if the Files changed in this pull request are included in default exclusions.

[flavorjones]

@dependabot recreate