LaTeX Regex Lint

A VS Code extension that surfaces custom, user-defined regex diagnostics for LaTeX. It supports per-rule severity, multiline matching, Quick Fixes, inline ignore directives, range disable/enable blocks, file-level ignores, and whitelist dictionaries to keep patterns readable.

Features

• Custom regex rules from settings (no rules are hardcoded).
• Diagnostics with per-rule severity: error / warning / info / hint.
• Multiline rules with optional maxLines control.
• Quick Fixes for regex replacements.
• Inline ignore directives (line, range, or file-level).
• Ignore by file pattern via glob list.
• Whitelist dictionaries (global + workspace) and external word lists.
• One-click ignore removal across the workspace.

How It Works

• The extension scans LaTeX documents (languageId latex) and common TeX extensions (.tex, .ltx, .ctx, .sty).
• Each rule is compiled into a JavaScript RegExp and matched across the whole document.
• Diagnostics are created for each match. A Quick Fix is shown if a rule defines a fix.

Multiline behavior (maxLines)

• If maxLines is not set: matches that span multiple lines are ignored.
• If maxLines = 0 or < 0: unlimited span allowed.
• If maxLines > 0: allow matches spanning up to that many lines.

Configuration

All configuration lives in settings.json.

Basic rule format

"latexRegexLint.rules": [
  {
    "id": "double-spaces",
    "pattern": "\\b([A-Za-z]+)\\s{2,}([A-Za-z]+)\\b",
    "message": "Double space detected: $1  $2",
    "severity": "Warning",
    "fix": "$1 $2"
  }
]

Fix formats

• String fix (uses rule pattern):

"fix": "$1 $2"

• Object fix:

"fix": {
  "pattern": "\\s{2,}",
  "replace": " ",
  "flags": "g"
}

Inline ignore directives

% lint-ignore                 % ignore all rules on this line
% lint-ignore: ruleA, ruleB    % ignore only specific rules on this line

Range disable/enable

% lint-disable
... ignored lines ...
% lint-enable

% lint-disable: ruleA, ruleB
... ignored for ruleA/ruleB ...
% lint-enable: ruleA

File ignore

% lint-ignore-file

Ignore by glob pattern

"latexRegexLint.ignoreFiles": [
  "**/appendix.tex",
  "**/draft/**"
]

Whitelist dictionaries (external or inline)

Use the {{WHITELIST}} token in a pattern. It will be replaced by a word list.

"latexRegexLint.rules": [
  {
    "id": "comma-followed",
    "pattern": ",(?=\\s+(?!(?:{{WHITELIST}})\\.?\\b)[A-Z][a-z])",
    "message": "Comma followed by unexpected capitalized word",
    "severity": "Warning",
    "maxLines": 0
  }
]

Global whitelist:

"latexRegexLint.whitelistWords": ["Fig", "Algorithm", "Section"],
"latexRegexLint.whitelistFiles": [".vscode/latex-whitelist.txt"]

Rule-specific whitelist:

"whitelistWords": ["Abbas", "Zhao"]

Merge behavior:

"latexRegexLint.whitelistMerge": true

Set to false in a workspace to override global lists instead of merging.

Commands

• LaTeX Regex Lint: Run — manual re-scan.
• LaTeX Regex Lint: Remove Ignore Directives — strips lint-ignore, lint-disable, lint-enable, lint-ignore-file comments from matched files.

Quick Fixes

• Fix: <ruleId> — applies the regex replacement for the matched range.
• Ignore: <ruleId> — appends % lint-ignore: <ruleId> on the same line.

Tips

• Use maxLines for multiline rules to avoid large spans.
• Keep {{WHITELIST}} lists in external files to avoid massive regex literals.
• Use workspace settings to scope special-case vocab.

Troubleshooting

• If a rule doesn't trigger, confirm the file language mode is LaTeX.
• Make sure the regex is valid JavaScript RegExp syntax.
• If patterns look correct but nothing matches, try removing maxLines first.