v2.0.2

• fix: improve encoding handling for specific files (#249) 1d01fe6
• chore(deps): update node.js to v22 (#245) e22cf6e
• chore(deps): update dependency yauzl to v3.2.0 (#243) 9c05bc6

v2.0.1...v2.0.2

v2.0.1

• chore: fix peer dependency range bd385d7

v2.0.0...v2.0.1

v2.0.0

requires Node v18 and bumps all possible dependencies

v1.5.0

• feat(run): Allow zip path as first argument (#3) bdaedb5
• ci: Add github action config (#2) f030605

v1.4.0...v1.5.0

Feature: Add `toString` to result of `combineFilters`, first tests

• chore: Pin node version with .nvmrc 4451d69
• feat(combineFilters): Implement toString 1dbbbf2
• chore: Add test script and first tests 93422a3
• chore: Add jest as devDependency 8dc06f4

v1.3.1...v1.4.0

First version published to npm

• chore: Add devDependency np 3aabf09

v1.3.0...v1.3.1

Add `extract` script, improved docs

docs: Improve Usage & Related Resources sections

- describe cli
- describe `extract` script
- move API into Usage section

feat: Mark package main as executable/binary

v1.2.0

feat: Mark package main as executable/binary

fix: Replace only known chars, display as hex

v1.1.1

fix(replaceNonTextChars): Only replace known characters and display h…

v1.1.0: feat: Provide `replaceNonTextChars`

Some xml documents (purposely) contain characters that are not visible and make it hard to reason about a test result.
Those characters also cause git to think a file is binary,
when the test output is being committed (e.g. in a jest snapshot).
By replacing them with a visual replacement using `codePointAt` (e.g. replacing NUL/`�` with `{!0!}`,
it's makes those more obvious.