v1.0.4

v1.0.4 — 2025-06-21 “Unicode Whitespaces & Docs”

Highlights

• 🔒 Robust‑whitespace support

  • Added hard‑break handling for \r, U+0085 (NEL), U+2028 (LS) and U+2029 (PS).
  • Zero‑width control blanks \v and \f are now ignored (no stray control bytes in output).
  • All other Unicode space separators (U+1680, U+2000–U+200A, U+202F, U+205F, U+3000) are preserved as visible blanks with correct runewidth accounting (U+3000 = 2 columns).

• 🛡️ Safety fixes

  • writeTabToLine now guards against tabSize == 0; prevents modulo‑zero panic.
  • isWordyGrapheme correctly classifies combining‑mark‑only clusters.

• 📝 Documentation & API comments

  • Expanded docstrings for StringWrap and StringWrapSplit, clarifying continued use of grapheme parsing for accurate width (e.g., ZWJ emojis).
  • Added inline commentary detailing whitespace classification logic.

• 🧪 New test coverage

  • Four fixtures covering legacy whitespace (\r, U+2028, \v, and long‑word split with U+000B).

• 🔧 Small clean‑ups

  • Consolidated whitespace switch: special‑cases first, single default for printable blanks.
  • Eliminated redundant runewidth.StringWidth call in the iter() word‑split helper, reducing extra scans on oversized words.
  • Added performance notes; core algorithm unchanged.

No breaking API changes. Previous tag checksums remain valid.