No zero-width characters in docs / prose
Warn on zero-width Unicode runes (U+200B/U+200C/U+200D/U+FEFF) in documentation and prose files. These render invisibly, but:
- Word-count / readability tools mis-count
- Search / grep miss the highlighted word
- Diff renders fine but actual byte diff shows the smuggled rune
This is the docs-side companion to security/no-zero-width-in-source.
WARN (not BLOCK) because some i18n libraries legitimately use ZWJ for
Indic / Arabic shaping in user-facing strings.
Leading U+FEFF (BOM) at file position 0 is ignored - encoding/no-bom
owns that case.
Fix#
Strip the offending characters:
LC_ALL=C sed -i 's/[\xE2\x80\x8B-\xE2\x80\x8D]//g' <file>
or in editors with “show invisible characters” turned on, delete by hand.
Deliberately not flagged#
- ZWJ between two emoji (👩💻) - the standard way to compose one glyph from two. A joiner between letters is still reported.
- Code examples in Markdown - fenced blocks and inline code spans. A
doc that demonstrates the attack (a ZWNJ inside
varname) has to contain the character it describes.
WARN lets the push through and records the finding. Turn frames on per repo on the dashboard's Policy page - see choosing what the gate checks.
Source on GitHub Live demo How it works Questions: contact@nimblegate.com