Repository navigation
What's Changed
Writer encodes each Nacha record as 94 ISO-8859-1 bytes. In-memory Go strings stay UTF-8. ASCII records are byte-identical. Extended characters such as é (U+00E9) are still é: the Writer emits Latin-1 0xE9 instead of UTF-8 C3 A9, so the line stays 94 bytes. It does not flatten é to e and it does not truncate a 95-rune record. Runes above U+00FF fail Write unless BypassValidation is set. The Reader already charset-decodes Latin-1/windows-1252, so round-trip through this library is unchanged.
File skips building the domestic batch-number map when there are no IAT batches.
Tests pin the inclusive $25,000 ARC entry limit.
- fix: encode Nacha records as ISO-8859-1 bytes by @adamdecaf in #1915
- perf: skip domesticNumbers map when File has no IAT batches by @adamdecaf in #1914
- test: pin the inclusive $25,000 limit on ARC entries by @Trenyx in #1913
Full Changelog: v1.67.0...v1.68.0