# Changelog
All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [Unreleased]
## [0.1.21] - 2026-02-14
### Added
- **CLI**: New `full` output format with ruff-style source line display showing
offending lines with caret underlines
([#425](https://github.com/rvben/rumdl/issues/425))
- **GitHub Action**: Add generic `args` input for passing extra CLI flags like
`--output-format json`
([#406](https://github.com/rvben/rumdl/issues/406))
- **MD060**: `loose-last-column` now caps last column width at header text width —
body cells shorter than header are padded, longer cells extend beyond
([#424](https://github.com/rvben/rumdl/issues/424))
### Changed
- **CLI**: `--output-format` help text now documents all available formats with
descriptions ([#425](https://github.com/rvben/rumdl/issues/425))
### Fixed
- **CI**: SchemaStore sync workflow now formats schema with Prettier before
committing
## [0.1.20] - 2026-02-13
### Added
- **Config**: Add `extends` key for config file inheritance — config files can
declare a base config to inherit from, with child settings merging on top.
Supports relative paths, absolute paths, `~/` expansion, recursive chains
(up to 10 levels), and circular reference detection. Works in both
`.rumdl.toml` and `pyproject.toml` ([#390](https://github.com/rvben/rumdl/issues/390))
- **Code block tools**: Detect fenced code blocks inside MkDocs admonitions
and tab containers ([#423](https://github.com/rvben/rumdl/issues/423))
### Fixed
- **MD013**: Detect mkdocstrings autodoc blocks regardless of flavor setting
([#396](https://github.com/rvben/rumdl/issues/396))
- **MD013**: Enforce strict line-length limit in semantic-line-breaks merge
([#414](https://github.com/rvben/rumdl/issues/414))
- **MD060**: Detect misaligned content in center/right-aligned table columns
([#426](https://github.com/rvben/rumdl/issues/426))
### Changed
- **Config**: `disable` now uses replace semantics (matching Ruff's `ignore`),
giving `extend-disable` a distinct purpose as the additive variant that
survives CLI overrides. Document the rule selection model with
`extend-enable` / `extend-disable` in global settings
## [0.1.19] - 2026-02-13
### Added
- **Kramdown flavor**: Add kramdown/Jekyll markdown flavor support with block
IAL (`{:.class}`, `{:#id}`), span IAL, and extension block
(`{::comment}`, `{::nomarkdown}`, `{::options}`) detection. Enable with
`flavor = "kramdown"` or `flavor = "jekyll"`. MD041 skips IALs and
extension blocks as preamble; MD051 defaults to kramdown anchor style
- **Config**: Add `extend-enable` and `extend-disable` global config keys
for additive rule activation, following Ruff's extend-select pattern.
Five rules are now opt-in by default: MD060, MD063, MD072, MD073, MD074.
Use `extend-enable = ["MD060"]` to enable individual opt-in rules without
overriding the full `enable` list
### Fixed
- **MD013**: Prevent reflow from splitting autolinks (`<https://...>`) at the
colon in URL schemes by treating them as atomic elements
([#416](https://github.com/rvben/rumdl/issues/416),
[#417](https://github.com/rvben/rumdl/issues/417))
- **MD013**: Prevent reflow from splitting inside markdown link text, code
spans, and HTML tag attributes by tracking element byte spans during
word-wrap fallback
([#412](https://github.com/rvben/rumdl/issues/412),
[#413](https://github.com/rvben/rumdl/issues/413))
- **MD013**: Thread `length-mode` (chars/visual/bytes) through all reflow
length calculations so CJK and other wide characters are measured correctly
([#414](https://github.com/rvben/rumdl/issues/414))
- **MD013**: Abbreviations like `e.g.` and `i.e.` inside parentheses no longer
trigger false sentence boundary detection in `sentence-per-line` mode
([#422](https://github.com/rvben/rumdl/issues/422))
- **MkDocs**: Track fenced code blocks inside admonitions to prevent false
`in_code_block` overrides that caused reflow corruption
([#415](https://github.com/rvben/rumdl/issues/415))
- **MD060**: Fix table formatter idempotency bug where ATX headings containing
pipe characters were misidentified as table rows and reformatted
- **Kramdown**: Fix self-closing extension blocks (`{::options ... /}`)
incorrectly disabling linting for all subsequent lines
- **Kramdown**: Fix extension fence leakage causing MD022 false positives
### Changed
- **Kramdown**: Add three-layer architectural filtering for extension blocks
(LineInfo sanitization, collection filtering, warning pipeline safety net)
so rules automatically skip kramdown extension blocks without per-rule
opt-in
## [0.1.18] - 2026-02-10
### Added
- **Config**: `enable = ["ALL"]` keyword to explicitly enable all rules
(equivalent to omitting `enable`), with `disable` still applied on top
- **MD013**: In non-strict mode, suppress line-length warnings when the excess
comes entirely from inline link/image URLs — if replacing `[text](url)` with
`[text]` would bring the line within the limit, the warning is not raised
([#393](https://github.com/rvben/rumdl/issues/393))
- **MD057**: `compact-paths` config option to warn about unnecessary path
traversal in wiki links (e.g., `../sibling` when `sibling` suffices)
([#391](https://github.com/rvben/rumdl/issues/391))
### Changed
- **Config**: `enable = []` now means "no rules enabled" (was incorrectly
treated as "use defaults"). Omitting `enable` still enables all rules.
Use `enable = ["ALL"]` to explicitly enable all rules.
### Fixed
- **MD013**: Skip mkdocstrings autodoc blocks (`::: module.Class` with indented
YAML options) during paragraph reflow
([#396](https://github.com/rvben/rumdl/issues/396))
- **MD013**: Implement trailing-word replacement for non-strict mode — reflow
now correctly handles lines where the last word pushes past the limit
- **MD044**: Skip proper name checking when link text is itself a URL (starts
with `http://`, `https://`, or `www.`), matching markdownlint behavior
([#395](https://github.com/rvben/rumdl/issues/395))
## [0.1.17] - 2026-02-09
### Fixed
- **Config**: `enable = []` is now correctly treated as "use defaults" (all rules
enabled) instead of acting as an empty allowlist that disabled all rules
- **MD013**: Preserve MkDocs admonition markers (`!!!`, `???`, `???+`) during text
reflow instead of stripping them from continuation lines
- **MD033**: Escape regex special characters in `allowed-elements` config values —
square brackets and other metacharacters were being interpreted as regex patterns
- **MD036**: Only flag emphasis as heading-like when it forms a standalone paragraph
(surrounded by blank lines), matching markdownlint behavior
### Changed
- **MD045**: Made diagnostic-only — auto-fix removed because meaningful alt text
requires human judgment. Automated placeholders derived from filenames (e.g.,
"User Profile image") are harmful for accessibility. The rule now reports missing
alt text without offering to fix it.
- **CI**: Pass version in pre-commit dispatch payload
## [0.1.16] - 2026-02-09
### Added
- **MD013**: `semantic-line-breaks` reflow mode (preview) — breaks lines at semantic
boundaries using a cascading strategy: sentence boundaries first, then clause
punctuation (`,` `;` `:` `—`), then English break-words (`and`, `or`, `but`,
`which`, `that`, `because`, etc.), then word wrap as fallback
([#388](https://github.com/rvben/rumdl/issues/388))
### Fixed
- **MD013**: Fix panic on multi-byte characters (smart quotes, em dashes) in
semantic line break sentence boundary detection
- **MD013**: Preserve element adjacency during reflow — text directly adjacent
to shortcodes, code spans, or links (e.g., `v{{< shortcode >}}`) is no longer
split across lines
- **MD013**: Use actual line length instead of URL-stripped length for line-length
checks — lines were incorrectly passing when long URLs inflated the real length
- **MD013**: Preserve Quarto/Pandoc div markers (`::: {.class}`) during text reflow
instead of reflowing them into surrounding paragraphs
- **MD001**: Track fixed heading level in `check()` for idempotent fixes — repeated
`--fix` runs no longer produce different output for multi-level heading violations
- **MD032**: Idempotent fix for ordered non-1 list followed by unordered list
- **MD032**: Use `fix()` method in proptest and allow convergence within 3 passes
- **MD062**: Bail out when unmatched angle bracket masks closing paren, preventing
false positive warnings on valid link syntax
- **MD009**: Strip all trailing Unicode whitespace in a single pass instead of
handling each whitespace character separately
- **GitHub Action**: Support multiple space-separated file paths in `path` input
- **docs**: Fix YAML typo, indentation, and `--no-exclude` descriptions
(thanks @JonathanWillitts in [#385](https://github.com/rvben/rumdl/pull/385))
- **config**: Handle markdownlint `default` key and boolean rule semantics in
`.markdownlint.json` compatibility
([#389](https://github.com/rvben/rumdl/issues/389))
### Changed
- **reflow**: Extract block boundary helpers to deduplicate paragraph detection logic
- **MD001**: Extract `compute_heading_fix()` to unify check() and fix() code paths
- **docs**: Add Helix editor formatter configuration and note about built-in support
- **docs**: Add link and nav validation guide, fix MD051 cross-file documentation
## [0.1.15] - 2026-02-07
### Added
- **MD057: `relative-to-docs` option** - New config option for validating absolute
link paths relative to a documentation root directory
### Performance
- **2.5x faster on real-world repositories** - Comprehensive performance audit and
optimization across the core pipeline, rule implementations, and infrastructure.
Validated against ripgrep, ruff, and rust repositories (1,832 files, 276K lines)
with zero regressions.
Key optimizations:
- Switch release profile from size optimization (`opt-level = "z"`) to speed (`opt-level = 3`)
- Zero-allocation line ending normalization for LF-only files (common case)
- Replace `chars().nth()` O(n) calls with O(1) byte indexing for ASCII checks
- Promote MD034 regex patterns to `LazyLock` statics (eliminated ~160K mutex cycles per 10K-line file)
- Reduce `LineInfo` memory 79% by boxing rare struct variants (3.4MB → 720KB for 10K-line files)
- Eliminate redundant `Vec<char>` allocation in horizontal rule detection
- Cache debug environment variable lookups in list parser
- Move `ListBlock` by value instead of cloning
- Fix MD044 O(n×m) boundary checks and full-document regex scans
- Fix MD052 repeated full-document HTML comment scans
- Replace per-line `HashMap` cloning in `InlineConfig` with state-transition storage
- Deduplicate `content.lines().collect()` across 48 call sites in 27 rules
- Compute content lines once in `LintContext`, share via `ctx.raw_lines()`
- Unify code block detection into a single pass shared with `LineIndex`
- Eliminate redundant `InlineConfig` parsing (parse once in `LintContext`)
- Deduplicate line offset and front matter computation during context construction
- Add O(1) line access via pre-computed line starts in `LineIndex`
- Lazy-initialize rule registry with `LazyLock` (eliminated 5+ redundant `all_rules()` calls)
- Remove unused mmap code path from file reading
### Changed
- **`explain` command completeness** - Now uses the full rule registry instead of
a manually maintained list, ensuring all rules (including MD069-MD074) are included
## [0.1.14] - 2026-02-06
### Added
- **MD074: MkDocs nav validation** - Validate that `mkdocs.yml` navigation entries
point to existing files (requires `flavor = "mkdocs"`)
- `not-found = "warn"` (default): Report nav entries pointing to non-existent files
- `omitted-files = "ignore"` (default): Optionally report markdown files not in nav
- `absolute-links = "ignore"` (default): Optionally warn about absolute paths
- Smart handling of directory nav entries, `docs_dir` setting, and session caching
- **MD057: Absolute links configuration** - New `absolute-links` config option
for controlling how absolute link paths are handled
- **MkDocs flavor: Extension-aware support** - Recognize Python-Markdown and
pymdown-extensions syntax to prevent false positives
- Python-Markdown anchor style (`{#custom-id}` header attributes)
- Unified PyMdown markup detection with span-based architecture
### Fixed
- **Tables in list items lose indentation** ([#383](https://github.com/rvben/rumdl/issues/383))
- `rumdl fmt` now preserves indentation for tables on list continuation lines
- Stack-based list context tracking handles nested lists, mixed ordered/unordered,
and respects CommonMark code block boundaries
- **MD065: Fix idempotency for consecutive horizontal rules** - Running `--fix`
twice on consecutive `---` rules no longer produces different output
- **MD032: Fix idempotency for ordered-non-1 items with code fences** - Fix mode
no longer produces different output on repeated runs for this edge case
### Changed
- **Codebase modularization** - Major internal restructuring for maintainability
- LSP server extracted into configuration, completion, and linting submodules
- File processor extracted into module directory with 3 submodules
- Config extracted into module directory with 8 submodules
- CLI extracted into commands directory with separate command handlers
- LintContext extracted into module directory with 8 submodules
## [0.1.13] - 2026-02-05
### Added
- **LSP: Code block language completion** - Autocomplete fenced code block languages
- Type ` ``` ` and press Ctrl+Space for language suggestions
- Includes 100+ languages from GitHub Linguist
- Added documentation in [docs/lsp.md](docs/lsp.md)
- **MD041: Opt-in auto-fix** - Add missing first-line heading with `--fix`
- Disabled by default to prevent unwanted changes
- Enable with `fix = true` in MD041 config
- Uses document title or filename as heading text
- **MD040: GitHub Linguist integration** - Normalize code block languages
- Recognizes language aliases (e.g., `py` → `python`)
- Uses GitHub Linguist database for accurate normalization
- **Code Block Tools: Additional built-in tools**
- Added djlint (Jinja/HTML), beautysh (bash), tombi (TOML), oxfmt (Jinja)
- 35 total built-in tool definitions
- **Code Block Tools: Configurable missing tool handling**
- `on-missing-language-definition`: What to do when language has no tools (`skip`, `warn`, `fail`)
- `on-missing-tool-binary`: What to do when tool binary not found (`skip`, `warn`, `fail`)
- **MkDocs: PyMdown Blocks support** - Recognize PyMdown extension syntax
- Supports `/// note`, `/// warning`, `/// details` and other block types
- **npm: CLI distribution** - Install via npm/npx
- `npx rumdl check .` - Run without global install
- Platform-specific packages for macOS, Linux, Windows
### Fixed
- **Code Block Tools: Embedded markdown linting is now opt-in** ([#380](https://github.com/rvben/rumdl/issues/380))
- Linting markdown inside code blocks was unexpectedly enabled by default
- Now requires explicit `[code-block-tools.languages.md]` configuration
- **MD040: Skip disabled lines when computing preferred labels**
- Fixed incorrect suggestions when some code blocks are disabled
- **Schema: Use standard integer type for timeout field** ([#374](https://github.com/rvben/rumdl/issues/374))
- Fixed JSON Schema validation in editors
## [0.1.12] - 2026-02-04
### Added
- **Code Block Tools [preview]** - Run external linters and formatters on fenced code blocks ([#331](https://github.com/rvben/rumdl/issues/331))
- `rumdl check`: Run configured linters (ruff, shellcheck, eslint, etc.) on code blocks
- `rumdl check --fix`: Run configured formatters (ruff, prettier, shfmt, etc.) to auto-format code blocks
- 31 built-in tool definitions with support for custom tools
- Language resolution via GitHub Linguist aliases (e.g., `py` → `python`, `bash` → `shell`)
- Configurable error handling per language (`fail`, `warn`, `skip`)
- See [docs/code-block-tools.md](docs/code-block-tools.md) for configuration guide
- **MD018: Per-rule magiclink configuration** - Control whether `magiclink` syntax is recognized per rule
- Add `magiclink = true` to MD018 config to skip email-like syntax
- **MD033: Auto-fix for `<a>` and `<img>` tags** - Convert simple HTML links and images to Markdown
- `<a href="url">text</a>` → `[text](url)`
- `<img src="url" alt="text">` → ``
- Requires `fix = true` in MD033 config (disabled by default)
- **MkDocs: Support for markdown-enabled HTML blocks** - Properly handle `<div markdown>` grid cards and similar patterns
- Supports 10 HTML5 sectioning elements (div, section, article, aside, details, figure, footer, header, main, nav)
- MD030 and MD035 now skip content inside markdown-enabled HTML blocks
- Added `skip_mkdocs_html_markdown()` filter to filtered_lines API
### Fixed
- **MD013: Prevent infinite loop in MkDocs admonition reflow** - Fixed edge case causing reflow to hang on certain admonition content
- **Config: Remove deprecated MD002 from example** - `rumdl.toml.example` no longer includes deprecated rules
### Changed
- **Config: Compact inline table syntax in example** - `rumdl.toml.example` now uses more readable inline tables for code-block-tools
## [0.1.11] - 2026-02-03
### Added
- **CI: Automated SchemaStore sync workflow** - Schema updates are now automatically synced to SchemaStore on release
### Fixed
- **MD044: Check proper names in link text, image alt text, and WikiLinks** ([#369](https://github.com/rvben/rumdl/issues/369))
- Previously only checked regular text and YAML front matter
- Now also checks `[link text](url)`, ``, and `[[WikiLinks]]`
- **Schema: Remove non-standard 'uint' format** ([#368](https://github.com/rvben/rumdl/issues/368))
- Fixed JSON Schema validation errors in editors when using `flavor = "obsidian"`
- Schema now uses standard `integer` with `minimum: 0` instead of non-standard `uint` format
## [0.1.10] - 2026-02-02
### Added
- **Obsidian flavor support** - New markdown flavor for Obsidian-specific syntax
- Tags (`#my-tag`) - Not flagged as missing heading space (MD018)
- Callouts (`> [!NOTE]`) - Recognized as valid blockquotes
- Highlights (`==text==`) - Not flagged as spacing issues
- Comments (`%%comment%%`) - Content inside is skipped by all rules
- Extended checkboxes (`- [/]`, `- [-]`, `- [>]`) - Recognized as valid task items
- Dataview fields (`field:: value`) - Not flagged as consecutive spaces
- Templater syntax (`<% code %>`) - Recognized as template syntax
- Block references (`^block-id`) - Not flagged
- **WASM: Obsidian flavor support** - Use `flavor: 'obsidian'` in Linter config
- Enables Obsidian-specific syntax recognition in browser/plugin contexts
- **WASM: Rule-specific configuration** - Pass rule configs to Linter constructor
- Example: `new Linter({ 'MD013': { 'line-length': 120 } })`
### Fixed
- **MD041: Skip MkDocs anchor lines** - Lines starting with `[](){#id}` are now skipped when checking for first-line heading
## [0.1.9] - 2026-02-01
### Fixed
- **MD033: Don't remove HTML tags that can't be converted to Markdown**
- Previously, auto-fix would destructively remove tags like `<img>`, `<span>`, and `<div>`
- Now only safe fixable tags (`em`, `i`, `strong`, `b`, `code`, `br`, `hr`) are converted
- All other HTML tags are left unchanged as expected
- **Fix mode: Respect per-file-flavor configuration**
- `rumdl check --fix` now correctly uses per-file flavor settings from `[per-file-flavor]`
- Previously used global flavor instead of file-specific flavor during fix coordination
## [0.1.8] - 2026-01-31
### Added
- **Inline config: Automatic support for all rules** ([#364](https://github.com/rvben/rumdl/issues/364))
- All rules now automatically support inline configuration via `<!-- rumdl-configure-file -->` comments
- Engine-level implementation ensures consistent behavior across rules
- Added `get_effective_config` helper for inline config support
- **MD033: Opt-in auto-fix for inline HTML conversion**
- New `fix = true` option enables auto-fix (disabled by default)
- Converts simple inline HTML to Markdown equivalents
- Conservative approach: only fixes clear-cut cases
- **MD036: Opt-in auto-fix for emphasis-as-heading**
- New `fix = true` option enables auto-fix (disabled by default)
- Converts emphasis-only paragraphs to proper headings
- **Test: Comprehensive MkDocs extension regression tests**
- Added 197 tests across 20 modules for MkDocs flavor
- Covers admonitions, content tabs, mkdocstrings, keys, snippets, math, and more
- Tests edge cases, cross-flavor comparison, fix preservation, and malformed syntax
### Fixed
- **MD013: Preserve MkDocs admonition and tab content during reflow** ([#361](https://github.com/rvben/rumdl/issues/361))
- Reflow now preserves required indentation inside MkDocs containers
- Admonition content no longer incorrectly converted to fenced code blocks
- Content tabs (`=== "Tab"`) properly handled during line wrapping
- **MD013: Prevent whitespace accumulation in sentence-per-line reflow** ([#360](https://github.com/rvben/rumdl/issues/360))
- Fixed infinite loop causing "failed to converge after 100 iterations" error
- Sentence-per-line mode now produces stable, idempotent output
- **MD064: Support inline configure-file comments** ([#364](https://github.com/rvben/rumdl/issues/364))
- `allow-sentence-double-space` now works with inline HTML config comments
- Sentences ending with markup (`` `code`. ``) now correctly recognized
## [0.1.7] - 2026-01-30
### Added
- **MD060: Loose last column option** ([#356](https://github.com/rvben/rumdl/issues/356))
- `loose-last-column = true` skips padding the last column in body rows
- Keeps tables compact while maintaining alignment for other columns
- Useful for tables with variable-length description columns
- (thanks @Ravlen for the suggestion)
- **MD060: Separate header/body alignment** ([#348](https://github.com/rvben/rumdl/issues/348))
- `column-align-header` controls alignment for the header row only
- `column-align-body` controls alignment for body rows only
- Existing `column-align` remains as fallback for both
- (thanks @pygarap for the suggestion)
- **MD018: MagicLink support in MkDocs flavor** ([#355](https://github.com/rvben/rumdl/issues/355))
- When using `flavor = "mkdocs"`, MD018 skips PyMdown MagicLink-style issue references
- `#10` and `#123` at the start of lines are not flagged as malformed headings
- Non-numeric patterns like `#Summary` are still flagged correctly
- (thanks @kattni for the suggestion)
### Changed
- **Docs: Add MD073 to opt-in rules table**
- MD073 (TOC validation) now listed in the opt-in rules section of docs/RULES.md
- **Release: Add SchemaStore update reminder**
- `verify-release-ready.sh` now warns when `rumdl.schema.json` has changed
- Reminds to submit a PR to SchemaStore after releasing
- **Release: Add opt-in rules documentation check**
- `verify-release-ready.sh` verifies all opt-in rules are documented in docs/RULES.md
- Prevents forgetting to document new opt-in rules
- **Release: Add config validation check**
- `verify-release-ready.sh` now tests all rule config options for validation warnings
- Catches schema bugs where valid options show "Unknown option" warnings
### Fixed
- **MD060: Fix config validation warnings for optional alignment options**
- `column-align-header` and `column-align-body` no longer show "Unknown option" warnings
- Schema now explicitly includes all valid config keys
- **MD073: Add missing indent option to config schema**
- `indent` option was accepted but not in the validation schema
- **CLI: Remove noisy hints from completions command**
- Shell completion output is now cleaner without installation hints in the generated script
## [0.1.6] - 2026-01-29
### Added
- **CLI: Shell completions subcommand** ([#327](https://github.com/rvben/rumdl/issues/327))
- `rumdl completions <shell>` generates shell completions for bash, zsh, fish, powershell, and elvish
- Auto-detects current shell and provides installation hints
- (thanks @tpoliaw in [#333](https://github.com/rvben/rumdl/pull/333))
- **CLI: Static rules.json for tooling integration** ([#351](https://github.com/rvben/rumdl/issues/351))
- Rule metadata now available at `https://raw.githubusercontent.com/rvben/rumdl/main/rules.json`
- Enables external tools (Alfred workflows, etc.) to access rule data without installing rumdl
- Release verification ensures rules.json stays in sync
### Fixed
- **MD073: Read indent config from MD007 by default** ([#353](https://github.com/rvben/rumdl/issues/353))
- TOC indentation now respects MD007's `indent` setting automatically
- Can be overridden with explicit `[MD073] indent = N`
## [0.1.5] - 2026-01-29
### Added
- **CLI: JSON output for rule metadata** ([#351](https://github.com/rvben/rumdl/issues/351))
- `rumdl rule -o json` exports all rule data as JSON
- `rumdl rule -o json-lines` for streaming/piping (one JSON object per line)
- Filter options: `--fixable` (`-f`), `--category` (`-c`)
- `--list-categories` shows available categories with rule counts
- `--explain` includes full documentation in output
- JSON includes: code, name, aliases, summary, category, fix_availability, url
### Fixed
- **MD073: Validate TOC entry indentation matches heading levels** ([#353](https://github.com/rvben/rumdl/issues/353))
- TOC entries must now have correct indentation relative to heading depth
- Example: H2 entries need 2-space indent, H3 entries need 4-space indent
- **CLI: Correctly count fixable issues** ([#349](https://github.com/rvben/rumdl/issues/349))
- "Run `rumdl fmt` to fix N issues" message now uses capability-based counting
- Previously could show fixable count for rules that don't actually auto-fix
- **CLI: Correctly count files modified by fmt** ([#347](https://github.com/rvben/rumdl/issues/347))
- Summary now shows actual number of modified files, not just files with issues
### Changed
- **Docs: Update all documentation URLs to rumdl.dev** ([#352](https://github.com/rvben/rumdl/issues/352))
- LSP code actions now link to rumdl.dev instead of GitHub
- Rule URLs in JSON output point to rumdl.dev
## [0.1.4] - 2026-01-28
### Changed
- **MD073: Simplified to marker-based TOC detection only** ([#332](https://github.com/rvben/rumdl/issues/332))
- Removed heading-based TOC detection (~200 lines of code)
- Now only supports `<!-- toc -->...<!-- tocstop -->` markers
- Ensures consistent detect/fix behavior - only validates TOCs users explicitly mark
- Rule is now opt-in: disabled by default, enable with `[MD073] enabled = true`
### Fixed
- **Rules: Call should_skip() in linting engine for opt-in rules**
- Opt-in rules (like MD073) now correctly skip when disabled
- Previously, rules had to check skip internally which caused inconsistent behavior
- Architectural fix affects all rules with opt-in behavior
- **MD073: Fix HTML anchor handling in heading ID generation**
- Headings with `<a name="..."></a>` anchors now correctly extract the anchor ID
- Centralized fix in `header_id_utils.rs` benefits all anchor-related functionality
- **MD032: Fix should_skip() to detect ordered lists**
- `likely_has_lists()` only checked for unordered markers (`*`, `-`, `+`)
- Ordered lists (e.g., `1. item`) were incorrectly skipped
- Now uses pre-computed `list_blocks` which includes all list types
## [0.1.3] - 2026-01-28
### Added
- **MD073: Table of Contents validation rule**
- Validates TOC entries match actual document headings
- Detects missing, extra, misordered, and misnamed TOC entries
- Supports multiple anchor styles: GitHub, GitLab, Markdownlint, VSCode
- Configurable: `toc-heading`, `anchor-style`, `case-sensitive`
### Fixed
- **MD064: Recognize sentence endings after markdown inline elements** ([#345](https://github.com/rvben/rumdl/issues/345))
- `allow-sentence-double-space` now works with sentences ending in inline code, emphasis, bold, strikethrough, and other markdown elements
- Example: `` `code`. Next sentence `` no longer flags a false positive
## [0.1.2] - 2026-01-27
### Fixed
- **LSP: Only auto-fix on manual saves, not autosave** ([#340](https://github.com/rvben/rumdl/issues/340))
- Prevents unwanted fixes when VS Code autosaves in the background
- Auto-fix now only triggers on explicit Cmd+S / Ctrl+S saves
- **MD032: Add auto-fix for lazy continuation lines** ([#342](https://github.com/rvben/rumdl/issues/342))
- When `allow-lazy-continuation = false`, lazy continuation lines now get proper indentation
- Detects all inline formatting at line start (emphasis, strong, strikethrough, links, images, code)
- Handles nested lists, blockquotes, and nested blockquotes correctly
- **MD037: Fix range calculation for emphasis spacing** ([#343](https://github.com/rvben/rumdl/pull/343)) (thanks @kachick)
- Fix ranges are now stored as absolute positions, preventing double-adjustment in fix mode
### Added
- **Test: Smoke test profile for package managers** ([#341](https://github.com/rvben/rumdl/pull/341)) (thanks @kachick)
- `make test-smoke` runs stable tests excluding flaky proptest/stress tests
- Useful for Nix and other package managers requiring deterministic test runs
## [0.1.1] - 2026-01-26
### Fixed
- **MD013: Preserve MkDocs snippet delimiters during reflow** ([#338](https://github.com/rvben/rumdl/issues/338))
- `-8<-` and `--8<--` delimiters now stay on their own lines when reflowing list items
- Prevents MkDocs Snippets extension syntax from being corrupted
- **MD013: Detect same-line closing tags for script/style elements** ([#339](https://github.com/rvben/rumdl/issues/339))
- Self-closing tags like `<script src="..."></script>` no longer cause subsequent lines to be skipped
- Lines after inline script/style tags are now properly checked for line length
- **MD013: Improve reflow handling for config and MkDocs syntax** ([#335](https://github.com/rvben/rumdl/issues/335), [#337](https://github.com/rvben/rumdl/issues/337))
- Recognize `abbreviations` config option (was showing "unknown option" error)
- Preserve attrlist syntax `{: .class }` on its own line during reflow
- **MD013: Require space after period for numbered list detection** ([#336](https://github.com/rvben/rumdl/issues/336))
- Prevents "failed to converge after 100 iterations" error on certain inputs
- Version numbers like `1.2.3` no longer mistakenly detected as list items
- **MD013: Handle email autolinks correctly**
- Email addresses in angle brackets (e.g., `<user@example.com>`) no longer break HTML tag extraction
- **MD005/MD030: Use blockquote-aware indent calculation**
- List continuation detection now correctly handles blockquote prefixes
- Multi-line list item detection accounts for blockquote nesting
### Changed
- **Shared blockquote utilities for indent calculation**
- Centralized logic for calculating indentation within blockquotes
- Improves consistency across MD005, MD030, and MD032 rules
## [0.1.0] - 2026-01-23
### 🎉 First Stable Release
rumdl reaches production-ready status after validation against 99,291 markdown files
across 51 popular repositories (kubernetes, react, vscode, rust-lang, tensorflow, and more)
with **zero false positives detected**.
### Added
- **CLI: `fmt --check` flag for CI usage**
- Returns non-zero exit code if files would be modified
- Enables dry-run formatting checks in CI pipelines
- **Fuzz testing infrastructure**
- Fix idempotency fuzzer verifies fixes don't oscillate
- Run with `make fuzz` (requires nightly Rust)
### Fixed
- **MD032: Single-pass idempotency for list blank lines**
- Fixes edge case where multiple fix passes could produce different results
- **MD050: Correct byte range for fix replacement**
- Fixes auto-fix for strong emphasis style in certain edge cases
### Changed
- **README: Added "Used By" section** with notable projects using rumdl
- **README: Updated SchemaStore section** to reflect current status
## [0.0.224] - 2026-01-22
### Added
- **GitHub Action: `fail-on-error` and `output-file` inputs** ([#324](https://github.com/rvben/rumdl/issues/324), [#325](https://github.com/rvben/rumdl/issues/325))
- `fail-on-error`: Control whether the workflow fails on violations (default: `true`)
- `output-file`: Write lint results to a file for use in subsequent steps
- **LSP: `source.fixAll.rumdl` code action for fix-on-save**
- Enables VS Code's "Fix All on Save" feature for rumdl
- Configure with `editor.codeActionsOnSave: { "source.fixAll.rumdl": "explicit" }`
## [0.0.223] - 2026-01-21
### Added
- **MD060: `column-align` option for table cell text alignment** ([#317](https://github.com/rvben/rumdl/issues/317))
- Options: `left`, `center`, `right` to force alignment for all columns
- Complements existing table formatting rules
- **Documentation site** ([rumdl.dev](https://rumdl.dev))
- Rule documentation with examples and configuration guides
### Fixed
- **Embedded markdown respects per-file-ignores and inline config**
- Both `rumdl check` and `rumdl fmt` now honor `<!-- rumdl-disable -->` comments
- Per-file-ignores configuration applies to embedded markdown blocks
- **Fix mode re-lint uses filtered rules** ([#319](https://github.com/rvben/rumdl/issues/319))
- After applying fixes, re-linting respects per-file-ignores and inline config
- **MD030: Auto-fix applies custom spacing config** ([#318](https://github.com/rvben/rumdl/issues/318))
- Custom `ul_single`, `ol_single`, etc. values now applied during fix
### Changed
- **Documentation improvements**
- Fixed MD051 location in RULES.md tables (thanks @Ravlen in #315)
- Added Flavors section to README (closes #316)
- Added Discord server link to issue templates (thanks @pygarap in #321)
## [0.0.222] - 2026-01-20
### Fixed
- **MD018: Comprehensive false positive fixes** ([#314](https://github.com/rvben/rumdl/issues/314))
- Skip content inside HTML comments (e.g., Jupyter cell markers `#%%`)
- Skip indented patterns to match markdownlint behavior
- Skip YAML comments inside frontmatter
- Flag all malformed headings at line start
- **Tables inside list items** ([#314](https://github.com/rvben/rumdl/issues/314))
- Tables embedded in list items are now detected, linted, and formatted correctly
- Validate delimiter indentation for list-table detection
- **MD009: Correct 1-indexed column in ASCII path**
- Column numbers in diagnostics now match editor positions
- **Schema: Include all flavor aliases** ([#312](https://github.com/rvben/rumdl/issues/312))
- Added `gfm`, `github`, `commonmark`, `qmd`, `rmd`, `rmarkdown` to schema
- JSON schema now matches documentation
- **CLI: Prevent duplicate config and use canonical rule list**
- Fixes edge case where config could be loaded multiple times
### Performance
- **Optimize code detection and cache hash computation**
- Faster detection of code blocks and spans
- **Cache per-file globsets**
- Significant speedup for rules using file-specific patterns
## [0.0.221] - 2026-01-18
### Added
- **Embedded markdown linting in fenced code blocks**
- Lint markdown content inside ` ```markdown ` code blocks
- Provides lint warnings for markdown examples in documentation
- Auto-fix support formats embedded markdown preserving code fence indentation
### Fixed
- **MD064: Remove `max-consecutive-spaces` config, improve sentence detection**
- Removed conceptually inconsistent `max-consecutive-spaces` option
- Share sentence detection logic with text reflow (DRY)
- Add proper abbreviation detection (Dr., Prof., e.g., i.e., Mr., Mrs.)
- Add CJK sentence-ending punctuation support (。, ！, ？)
- Fix UTF-8 byte boundary handling for multi-byte characters
- **MD055/MD056: Handle tables inside blockquotes in fix mode**
- Tables within blockquotes now preserve `> ` prefix when fixed
- Pipe style and column count fixes work correctly in nested contexts
- **MD060: Improve alignment detection for CJK and delimiter styles**
- Better handling of CJK character widths in table alignment
- Improved delimiter row style detection
- **MD069: Respect inline disable comments in fix mode**
- `<!-- markdownlint-disable MD069 -->` now properly prevents fixes
- Exposed inline config checks for rule disable detection
- **MD058/MD065: Recognize blockquote continuation lines as blank**
- Lines with only `>` are now treated as blank for table separation rules
- Fixes false positives in blockquoted content
- **Text reflow: Recognize sentence boundaries with quotation marks**
- Sentences ending with closing quotes (`.\"`, `!\"`, `?\"`) now detected correctly
- Supports both straight and curly quotation marks
### Changed
- **Refactored sentence detection into shared module**
- New `sentence_utils` module provides reusable sentence boundary detection
- Used by both MD064 (consecutive spaces) and text reflow (MD013)
- Reduces code duplication by ~100 lines
## [0.0.220] - 2026-01-17
### Added
- **`RUMDL_OUTPUT_FORMAT` environment variable** ([#297](https://github.com/rvben/rumdl/issues/297))
- Override output format in CI/CD without modifying config files
- Precedence: CLI flag → env var → config → default
- Example: `RUMDL_OUTPUT_FORMAT=github rumdl check .`
### Fixed
- **MD050: Skip math blocks to avoid false positives in Quarto**
- LaTeX subscripts (`x_1`) and exponentiation (`a**b`) were incorrectly flagged as strong emphasis
- Now correctly skips content inside `$$...$$` math blocks
### Changed
- **Documentation: Clarify CommonMark 0.31.2 spec version** ([#298](https://github.com/rvben/rumdl/issues/298))
- Explicitly document that rumdl uses CommonMark 0.31.2 via pulldown-cmark
- Clarify that `standard` flavor includes GFM extensions (tables, task lists, strikethrough, autolinks)
- Document `commonmark` as alias for `standard`, `github` as alias for `gfm`
- Add `output-format` setting documentation with all 12 available formats
## [0.0.219] - 2026-01-17
### Added
- **Comprehensive markdown flavor support**
- **GFM flavor**: Security-sensitive HTML tag detection in MD033, extended autolinks including `xmpp:` protocol in MD034
- **MkDocs flavor**: mkdocstrings block support, extended markdown syntax (keys `[[Ctrl]]`, caret `^^sup^^`, mark `==highlight==`, ins `++inserted++`, tilde `~sub~`)
- **MDX flavor**: JSX attribute detection (`className`, `htmlFor`, `onClick`, etc.), JSX expression handling, ESM import/export support
- **Quarto flavor**: Pandoc citation syntax (`@ref`, `[@ref]`), shortcode detection (`{{< >}}`), div blocks and callouts (`::: {.callout-*}`), math block handling for emphasis rules
- **Comprehensive UTF-8 stress tests**
- Systematic testing of all rules against 11 scripts (Bengali, Arabic, Chinese, Japanese, Korean, Thai, Hindi, Russian, Greek, Emoji, ZWJ sequences)
- Tests for check(), fix(), and fix range character boundaries
- Catches UTF-8 panics before they reach production
### Fixed
- **MD034: UTF-8 panic with multi-byte text before emails**
- Fixed crash when email follows Bengali, Chinese, Arabic, or other multi-byte UTF-8 text
- The `xmpp:` prefix check was slicing at invalid character boundaries
- Discovered during testing against kubernetes/website
- **MD031: Blank line detection after MkDocs admonition blocks**
- Corrected blank line requirements after admonition content
- **MkDocs: Nested admonition handling**
- Implemented stack-based detection for properly nested admonitions
### Changed
- **Documentation: Split flavor docs into separate files**
- `docs/flavors.md` is now a concise overview with links
- Each flavor has its own file: `docs/flavors/standard.md`, `gfm.md`, `mkdocs.md`, `mdx.md`, `quarto.md`
- Easier to find, reference, and maintain flavor-specific documentation
## [0.0.218] - 2026-01-15
### Added
- **Style Guide Preset examples**
- Google Markdown style guide config (`examples/google-style.toml`)
- Prettier-compatible markdown config (`examples/prettier-style.toml`)
### Fixed
- **MD032: Eliminate false positives on real-world repos**
- Tested against Rust Book (478 files) and MDN Content (14,100 files) with zero false positives
- Treat HTML comments as transparent for blank line checks
- Skip blank line warning when exiting blockquotes (boundary provides separation)
- Handle prose numbered items like "Chapter 19." (sentence continuation detection)
- Fix lazy continuation for code span continuations and nested sublists
- Remove incorrect uppercase heuristic that was closing lists prematurely
- **Config: Correct rule alias mappings**
- Add `descriptive-link-text` alias for MD059
- Fix `table-cell-alignment` alias to map to MD060 (was incorrectly MD059)
- **Config: Warn on per-file-ignores patterns with commas**
- Detects when patterns contain commas without braces
- Suggests correct glob syntax: `{*.md,*.txt}` instead of `*.md,*.txt`
## [0.0.217] - 2026-01-14
### Added
- **MD064: `max-consecutive-spaces` config option** (fixes #294)
- New option to allow configurable consecutive space threshold
- `max-consecutive-spaces = 2` allows traditional two-space sentence spacing
- Default remains `1` (flags any 2+ consecutive spaces)
- **Comprehensive Markdown flavor documentation**
- New `docs/flavors.md` with detailed flavor support guide
- Coverage of MkDocs, MDX, Quarto, GitHub, and Standard flavors
### Fixed
- **MD032: Detect lazy continuation in nested lists** (fixes #295)
- Lazy continuation lines in nested lists are now correctly identified
- Uses `content_column` from `ListItemInfo` to calculate proper indent threshold
- **MD032: Detect blockquote level changes as list breaks**
- Lines with different blockquote nesting levels now properly break lists
- Prevents false positives when list items span blockquote boundaries
- **MD033: Calculate correct end_line for multiline HTML tags**
- HTML tags spanning multiple lines now report accurate end positions
- Improves LSP diagnostic highlighting for multiline HTML elements
- **Config warnings show relative paths instead of absolute** (fixes #291)
- Config warning messages now use project-relative paths
- Makes error messages cleaner and more portable
- **Core correctness and performance improvements**
- Various fixes for edge cases in line processing
- Performance optimizations in lint context creation
### Community
- Opened official Discord server for the rumdl community (thanks @pygarap for the suggestion in #281)
- Join at https://discord.gg/ADTJFSFUyn
- Channels: #general, #announcements, #help, #development
- Release notifications automatically posted via GitHub webhook
## [0.0.216] - 2026-01-13
### Added
- **MD072: Custom key order for frontmatter sorting** (fixes #290)
- New `key-order` config option to define preferred key ordering
- Keys in the order list appear first, unlisted keys sort alphabetically after
- Example: `key-order = ["title", "date", "author"]` puts title first
### Fixed
- **MD060: Preserve blockquote prefix when formatting tables**
- Tables inside blockquotes (e.g., `> | A | B |`) now correctly preserve the prefix
- Supports nested blockquotes (`>>`, `>>>`) with proper prefix preservation
- Works with all formatting styles (aligned, compact, tight)
- **MD057: Skip link patterns inside LaTeX math spans** (fixes #289)
- Link-like patterns inside `$...$` or `$$...$$` are no longer flagged as broken links
- Prevents false positives from LaTeX commands that resemble markdown links
- **MD030: Skip empty marker lines without content** (fixes #288)
- Lines with only a list marker (no content) are now skipped
- Prevents incorrect warnings about spacing on empty list items
- **MD031/MD032/MD058/MD065: Auto-fix preserves nested blockquote prefix** (fixes #268)
- When inserting blank lines in blockquotes, the blockquote prefix is now preserved
- Works correctly with nested blockquotes (`>>`, `>>>`)
- Blank blockquote lines use `>` without trailing space (per markdownlint-cli)
- **MD005/MD007: Prevent oscillation between rules**
- Ordered and unordered list items are now tracked in separate groups
- Fixes cases where MD005 and MD007 would fight over indentation
## [0.0.215] - 2026-01-12
### Fixed
- **MD057: Exclude footnote definitions from relative link checking** (fixes #286)
- Footnote definitions (`[^id]: content`) were incorrectly parsed as reference link definitions
- This caused MD057 to treat footnote content as URLs (e.g., `[^1]: [link](https://example.com)` was flagged as broken link)
- Footnotes are now correctly identified and excluded from reference definition parsing
- Also benefits MD053 (unused reference definitions) which uses the same parsing
- **MD005: Auto-fix preserves blockquote prefix** (fixes #268)
- When fixing list indentation inside blockquotes, the `>` prefix is now preserved
- Previously, auto-fix could break blockquote structure
## [0.0.214] - 2026-01-11
### Added
- **Per-file flavor configuration** (fixes #283)
- Configure different markdown flavors for specific file patterns using glob patterns
- Example: `"docs/**/*.md" = "gfm"` in `[global.file-flavors]` section
- Supports all flavor values: `commonmark`, `gfm`, `mkdocs`, `obsidian`, `mdbook`
- File patterns are matched against relative paths from project root
### Fixed
- **MD028: Skip GFM alerts when checking blank lines in blockquotes** (fixes #126)
- GFM alerts (`> [!NOTE]`, `> [!TIP]`, `> [!IMPORTANT]`, `> [!WARNING]`, `> [!CAUTION]`) require blank line separation to render correctly
- MD028 now detects GFM alerts and allows blank lines between them
- Case-insensitive detection supports both `[!NOTE]` and `[!note]`
- Regular blockquotes between GFM alerts are still flagged correctly
- **MD032: Auto-fix handles varying blockquote whitespace** (fixes #268)
- Auto-fix no longer breaks blockquote structure when lines have different spacing
- Properly handles mixed `>` and `> ` prefixes within the same blockquote
- **MD032: Detect blockquotes before skip conditions** (fixes #284, #285)
- Blockquote detection now runs first, preventing false positives inside blockquotes
- Fixes cases where multi-paragraph list items in blockquotes triggered warnings
## [0.0.213] - 2026-01-08
### Fixed
- **MD031/MD046: Detect code blocks inside lists** (fixes #276)
- MD031 with `list-items=true` now correctly flags fenced code blocks inside lists
- MD046 with `style=fenced` now correctly flags indented code blocks inside lists
- Uses pulldown-cmark for accurate CommonMark-compliant code block detection
- Properly distinguishes between code blocks and footnote/admonition/tab content
- **MD057: Check reference-style link definitions for broken links** (fixes #274)
- Reference-style link definitions like `[ref]: ./path.md` are now checked
- Previously only inline links `[text](./path.md)` were validated
- All existing validations apply: external URLs skipped, .html→.md fallback, URL decoding
- **MD007: "Do What I Mean" behavior for indent configuration** (fixes #273)
- When `indent` is explicitly set without `style`, automatically use fixed style
- Setting `indent = 4` now actually uses 4-space increments as expected
- Warning shown if both `indent` and `style = "text-aligned"` are set (contradictory)
- **MD030: Prevent false positives in math blocks** (fixes #275)
- Math blocks (`$$...$$`) are now detected and excluded from list marker spacing checks
- Fixes incorrect warnings about list spacing inside LaTeX display math
## [0.0.212] - 2026-01-07
### Added
- **CLI: Show relative file paths in output by default** (fixes #266)
- Add `--show-full-path` flag to opt into absolute paths
- All output formats (text, JSON, SARIF, GitHub, GitLab, JUnit) now consistently use paths relative to project root or CWD
### Fixed
- **LSP: Respect FormattingOptions in formatting handler** (fixes #265)
- `insertFinalNewline`: Add trailing newline if missing
- `trimFinalNewlines`: Remove extra trailing newlines at EOF
- `trimTrailingWhitespace`: Remove trailing whitespace from all lines
- Applies FormattingOptions after lint fixes to match editor preferences
- **Fix coordinator: Re-check all rules after each fix for idempotency** (fixes #271)
- `rumdl check --fix` now produces idempotent results
- After each fix, all rules are re-checked with fresh LintContext
- Handles cascading fixes (e.g., MD046 → MD040) in a single run
- **MD046: Preserve code indentation and skip MkDocs admonitions** (fixes #269, #270)
- Converting fenced to indented code blocks now preserves internal indentation
- MkDocs admonitions (`!!! note`, `??? warning`, `???+ tip`) are properly skipped
- **MD012: Fix EOF blank line removal in LSP formatting** (fixes #264)
- Warning's fix replacement now correctly removes blank lines at EOF
- **MD072: Improved frontmatter key sorting** (fixes #265)
- Use line-based reordering to preserve original YAML formatting
- Fixed config warning about "unknown option: enabled"
- Compute fix in check() to enable LSP formatting
## [0.0.211] - 2026-01-06
### Fixed
- **MD032: Handle blockquote continuation lines correctly** (fixes #260)
- Blockquote continuation lines within the same blockquote level no longer incorrectly trigger "list should be surrounded by blank lines" warnings
- Changed detection from checking if line starts with `>` to comparing blockquote nesting levels
- Lists inside blockquotes like `> - item\n> continuation` now lint correctly
- **MD071: Preserve trailing newlines when fixing frontmatter spacing** (fixes #262)
- Auto-fix no longer removes the file's trailing newline
- Prevents MD047 (files should end with newline) from triggering after MD071 fix
### Changed
- **Rules: Add aliases for MD070, MD071, MD072**
- MD070: `nested-code-fence-collision`
- MD071: `blank-line-after-frontmatter`
- MD072: `frontmatter-key-sort`
## [0.0.210] - 2026-01-05
### Added
- **MD071: Blank line after frontmatter** (new rule)
- Enforces a blank line between frontmatter closing delimiter and document content
- Supports YAML (`---`) and TOML (`+++`) frontmatter formats
- Auto-fixable
- **MD072: Frontmatter key sort** (new rule, disabled by default)
- Checks that frontmatter keys are sorted alphabetically
- Supports YAML and TOML frontmatter
- Configurable `ignore` list for keys to exclude from sorting
- Enable with `MD072.enabled = true` in configuration
- Auto-fixable
### Fixed
- **MD022: Support kramdown Inline Attribute Lists** (fixes #259)
- IAL syntax like `{: .class #id}` immediately after headings no longer triggers blank line warnings
- IAL lines are treated as part of the heading element
- Multiple consecutive IAL lines supported
- Fix function preserves IAL attachment to headings
- **MD046: Use pulldown-cmark for code block detection**
- More reliable code block detection using CommonMark-compliant parser
- Fixes edge cases with nested and consecutive code blocks
- **MD072: Fix JSON depth tracking and regex initialization**
- Use LazyLock for regex to avoid recompilation
- Fix JSON nested object depth tracking when extracting keys
- **MD072: Stop at TOML table headers when extracting keys**
- TOML table headers like `[section]` now properly terminate key extraction
## [0.0.209] - 2026-01-05
### Added
- **MD070: Nested code fence collision detection** (new rule)
- Detects when inner code fences would prematurely close outer fences
- Example: outer ` ``` ` containing inner ` ``` ` causes collision
- Suggests using longer fences (` ```` `) for outer blocks
- Configurable `languages` option to limit checking to specific info strings (default: `["", "markdown", "md"]`)
- Auto-fix increases fence length to prevent collisions
### Fixed
- **MD040: Fix false positives on list-indented code blocks** (fixes #257)
- Refactored to use pulldown-cmark for CommonMark-compliant code block detection
- List-indented closing fences no longer incorrectly flagged as missing language
- Eliminates manual fence state tracking that failed for consecutive code blocks
- **CLI: Support `respect-gitignore` in config files** (fixes #258)
- Added `respect-gitignore` option to `[global]` section in `.rumdl.toml` and `pyproject.toml`
- CLI flag `--respect-gitignore=false` syntax now works (previously only `--no-respect-gitignore`)
- **Release: Fix pre-built aarch64-linux binaries for 16K page systems** (fixes #256)
- Pre-built binaries for `aarch64-unknown-linux-gnu` and `aarch64-unknown-linux-musl` now work on Asahi Linux and Raspberry Pi 5
- Set `JEMALLOC_SYS_WITH_LG_PAGE=16` to configure jemalloc for 64K page compatibility
- Previously crashed with "Unsupported system page size" on systems with 16K or 64K pages
- **List detection: Filter out YAML/TOML lists in frontmatter**
- Frontmatter arrays like `tags: [foo, bar]` no longer incorrectly detected as markdown lists
- Fixes false positives in MD004, MD005, MD007, MD030, and other list-related rules
- **List detection: Handle nested list items and tab indentation**
- Nested items where content begins on the next line are now correctly parsed
- Tab-indented nested lists now properly detected and validated
- **MD030: Eliminate false positives on hard-wrapped citations and continuations** (fixes #253)
- Replaced manual regex-based list detection with pulldown-cmark parser for context-aware list item detection
- Hard-wrapped citation continuations like `(Doe 2003, p.\n 1234)` no longer trigger false positives
- Resolves LSP formatting infinite loop (adding trailing space → MD009 → removing space → MD030)
- Added 16 comprehensive edge case tests covering citations, blockquoted lists, URLs, ISBNs, and mathematical expressions
- **Heading detection: Prevent list items from being detected as setext headings** (fixes #254)
- Per CommonMark spec 4.3, setext heading content cannot be a list item, ATX heading, blockquote, thematic break, code fence, or HTML block
- Incomplete list items (e.g., `- Orange` followed by `-`) no longer incorrectly detected as level 2 headings
- Eliminates false positives in all heading-related rules (MD041, MD022, MD023, MD024, MD025, MD026, MD003, MD043)
- Added comprehensive validation for all CommonMark setext heading exclusions
- **Output formatters: Correct severity field in JSON, JSON Lines, and Azure Pipelines formats**
- Fixed bug where severity was always output as "error" regardless of actual severity level
- JSON/JSON Lines/Azure formatters now correctly output "error", "warning", or "info" based on actual severity
## [0.0.208] - 2026-01-02
### Changed
- **Reflow: Replace manual emphasis parsing with pulldown-cmark**
- Use pulldown-cmark for CommonMark-compliant emphasis, bold, and strikethrough detection
- Removes ~100 lines of complex manual regex-based emphasis detection
- Adds GFM strikethrough (`~~text~~`) support to sentence boundary detection
### Fixed
- **Reflow: Continue emphasis markers across sentence splits** (fixes #251)
- When splitting emphasized text at sentence boundaries, each line now properly retains its emphasis markers
- Example: `*Sentence one. Sentence two.*` now correctly splits to `*Sentence one.*` and `*Sentence two.*`
- **Reflow: Fix line_length=0 to mean "no wrapping"**
- Previously `line_length=0` would wrap after every word due to condition always being true
- Now correctly treats 0 as "unlimited line length" (no wrapping)
- **MD007: Resolve false positives on nested unordered lists in ordered lists** (fixes #247)
- Nested unordered lists inside ordered lists now calculate expected indent based on parent's content column
- Fixes ping-pong bug where MD007 fix would then trigger MD005, destroying nesting structure
- **MD009: Correct byte offset calculation for multi-byte characters**
- Column positions now correctly account for UTF-8 multi-byte characters
- Fixes incorrect warning positions when lines contain non-ASCII characters
- **MD027: Continue HTML block detection until blank line after closing tag**
- CommonMark type 6 HTML blocks (like `<table>`) now correctly extend until the first blank line
- Prevents false positives for blockquote indentation inside HTML tables
- **MD034: Resolve false positives for URLs in angle brackets and parentheses**
- URLs inside angle brackets (`<https://example.com>`) no longer trigger bare URL warnings
- URLs inside parentheses with descriptions are properly handled
- **MD037: Mask inline math spans before emphasis detection**
- Asterisks inside inline math (`$p * q$`) no longer interfere with emphasis detection
- Fixes false negatives where valid `* spaced emphasis *` was not detected
- **MD051: URL-decode CJK fragments before anchor matching**
- Japanese, Korean, and Chinese heading fragments are now properly matched
- URL-encoded fragments (e.g., `#%E6%97%A5%E6%9C%AC%E8%AA%9E`) correctly resolve to decoded anchors
- **MD057: Resolve false positives for email addresses and file references**
- Email-style patterns and file references with @ symbols no longer trigger GitHub username warnings
- **MD069: Resolve false positives for nested strikethrough**
- Properly handles strikethrough markers that appear in nested contexts
- **Config: Handle relative paths in find_project_root_from**
- Configuration discovery now works correctly with relative file paths
- **Reflow: Preserve Hugo shortcodes as atomic elements**
- Hugo shortcodes (`{{< >}}` and `{{% %}}`) are no longer broken across lines during text reflow
## [0.0.207] - 2025-12-30
### Added
- **CLI: Add `--flavor` option to override markdown flavor**
- New CLI flag allows overriding the markdown flavor for a single run
- Options: `commonmark`, `gfm` (GitHub Flavored Markdown)
- Useful for testing files against different markdown specifications
### Fixed
- **MD029: Respect CommonMark list start values and preserve user intent** (fixes #247)
- Ordered lists starting at values other than 1 (e.g., `11.`, `12.`, `13.`) are now validated correctly
- Auto-fix no longer destroys user intent by changing `11.` to `1.` in properly numbered lists
- Lists with `start` attribute > 1 will warn but not auto-fix, preserving semantic numbering
- **WASM: Convert byte-based columns to character offsets for JavaScript**
- Column positions in WASM output now use character offsets instead of byte offsets
- Fixes incorrect highlighting in editors when files contain multi-byte UTF-8 characters
### Changed
- **Docs: Add mise as an installation option**
- README now includes mise (formerly rtx) as an alternative installation method
## [0.0.206] - 2025-12-28
### Fixed
- **MD013: Apply inline config before should_skip optimization** (fixes #241)
- Inline configuration (e.g., `<!-- markdownlint-configure-file {"MD013": {"tables": false}} -->`) was not being applied when files had short lines
- The should_skip optimization now uses effective config after parsing inline overrides
- **CLI: Warn about unknown rules in --enable/--disable flags** (fixes #243)
- Using non-existent rule names now produces a warning instead of silently ignoring them
- Invalid rules in all config entry points (CLI flags, config files) are now validated
- **MD034: Fix panic with multi-byte URLs and unbalanced parentheses**
- URLs containing non-ASCII characters (e.g., Chinese Wikipedia URLs) followed by unbalanced parentheses no longer cause a panic
- Fixed by using byte indices instead of character indices for string slicing
## [0.0.205] - 2025-12-27
### Fixed
- **MD037: Skip multi-line inline code spans**
- Asterisks inside code spans that cross line boundaries no longer trigger false positives
- Uses pulldown-cmark for accurate multi-line code span detection
- **MD031: Don't require blank line after frontmatter**
- Code blocks and admonitions immediately after YAML/TOML frontmatter no longer trigger warnings
- Matches markdownlint behavior for frontmatter handling
- **MD034: Handle URLs with parentheses in path** (fixes #240)
- URLs like `https://example.com/page_(1)` no longer incorrectly flag the closing parenthesis
- Centralized URL pattern handling for consistent behavior
- **MD046: Detect mixed whitespace indented code blocks**
- Indented code blocks using tabs or mixed whitespace are now properly detected
- 4-space indented fences correctly identified as indented code blocks
- **Heading fixes: Preserve original whitespace**
- Auto-fix for heading rules now preserves the original spacing around heading text
- **CLI: Support rule aliases in --enable/--disable flags**
- Rule aliases (e.g., `blanks-around-fences` for MD031) now work with CLI flags
### Changed
- **Regex patterns: Centralize URL detection**
- Consolidated URL patterns with documentation for maintainability
- Removed unused URL_REGEX and BARE_URL_REGEX
- **Tab expansion: Consolidate into shared utility**
- CommonMark-compliant tab expansion now uses shared `visual_indent` utility
- Consistent indentation detection across all rules
## [0.0.204] - 2025-12-26
### Fixed
- **MD007: Use per-parent indent logic instead of document-wide style**
- Fixed incorrect warnings when nested lists had different indentation styles
- Each parent list item now independently tracks its children's indentation
- **MD022: Distinguish horizontal rules from frontmatter delimiters**
- HR patterns (`---`, `***`, `___`) after headings no longer incorrectly trigger blank line warnings
- Centralized HR detection in `LineInfo` for consistent behavior across rules
- **LSP: Unify cross-file link extraction with MD057**
- LSP diagnostics now report correct column positions for MD057 warnings
- Created shared `extract_cross_file_links` utility used by both CLI and LSP
- Ensures consistent position tracking between editor and command-line
## [0.0.203] - 2025-12-25
### Added
- **Config: Implement Ruff-style config inheritance model**
- User config (`~/.config/rumdl/`) now serves as base, project config extends it
- Project-level `enable`/`disable` are additive to user config by default
- Clear precedence: CLI flags > project config > user config > defaults
- **MD001: Recognize frontmatter `title` as implicit H1**
- Documents with `title:` in YAML frontmatter no longer require an H1 heading
- Aligns with common static site generator behavior (Jekyll, Hugo, etc.)
### Fixed
- **MD057/MD051: Skip wikilinks in cross-file validation** (fixes #235)
- Wikilinks (`[[page]]`) were incorrectly triggering "relative link does not exist" warnings
- Wikilinks use a different linking system (Obsidian, etc.) and should not be validated as file paths
- **MD057: Fix LSP diagnostic position for cross-file links** (fixes #234)
- LSP was pointing to link text `[...]` instead of URL `(...)`
- Now correctly highlights the URL portion in editor diagnostics
- **MD012: Remove all trailing blank lines at EOF**
- Previously only removed one trailing blank line per fix pass
- Now removes all consecutive trailing blank lines in a single fix
- **Config: Add missing aliases for MD062/MD064/MD065 rules** (fixes #232)
- Added kebab-case aliases for newer rules
- Fixed MD069 typo in alias registration
- **CLI: Fail fast when `--config` conflicts with `--no-config`/`--isolated`**
- These flags are contradictory and now produce a clear error instead of silently ignoring one
- **Cross-file validation: Fix duplicate warnings for same link**
- MD051 and MD057 both contribute links to the workspace index with different column positions
- Deduplication now ignores column differences, preventing duplicate warnings
## [0.0.202] - 2025-12-24
### Added
- **MD069 (no-duplicate-list-markers): Detect accidental duplicate list markers** (fixes #227)
- Detects patterns like `- - text` that occur from copy-paste with editor auto-list-continuation
- Supports all unordered markers: `-`, `*`, `+` and mixed combinations
- Auto-fix removes the first marker: `- - text` → `- text`
- Correctly ignores CLI flags (`- --verbose`), emphasis (`* *italic*`), and nested lists
- Skips code blocks, frontmatter, and HTML comments
- **MD032 (blanks-around-lists): Add `allow-lazy-continuation` config option**
- When enabled, allows list items to continue without blank lines in specific contexts
- Useful for documentation styles that prefer compact list formatting
### Fixed
- **LSP: Make severity a first-class field on RuleConfig**
- Severity configuration now properly propagates through LSP diagnostics
- Fixes issue where custom severity settings were not reflected in editor diagnostics
### Changed
- **Documentation improvements**
- Added `.ignore` file support documentation
- Added LSP server documentation
- Added footnote rules documentation and cache configuration
- Added opt-in rules section to RULES.md
## [0.0.201] - 2025-12-24
### Added
- **CLI: Add `--fail-on` flag to control exit code by severity**
- Configure minimum severity level that triggers non-zero exit code
- Options: `error` (default), `warning`, `info`
- `--fail-on warning` exits non-zero on warnings or errors
- `--fail-on info` exits non-zero on any diagnostic
- Useful for CI/CD pipelines with different strictness levels
- **Severity: Add Info severity level for low-priority issues**
- New severity level below Warning for informational diagnostics
- Allows rules to report suggestions without affecting exit codes by default
- Can be promoted to affect exit codes via `--fail-on info`
- **MD060 (table-column-count): Smart max-width inheritance from MD013**
- Automatically inherit `max-width` from MD013 `line-length` when not explicitly set
- Reduces configuration duplication between MD013 and MD060
- Explicit MD060 `max-width` still takes precedence
### Fixed
- **MD063 (heading-capitalization): Handle inline code at start of heading** (fixes #228)
- Headings starting with inline code (e.g., `` `code` text ``) now correctly apply sentence case to subsequent text
- Previously, the word after inline code was incorrectly lowercased
- Applies to all non-text segments (code, links, HTML) at heading start
- **MD063 (heading-capitalization): Preserve acronyms and brand names in link text**
- Link text with internal capitals (e.g., `[GitHub](url)`) is now preserved
- `preserve_cased_words` setting applies to link text, not just plain text
- Prevents incorrect lowercasing of brand names within links
- **MD063 (heading-capitalization): Handle inline HTML tags**
- Inline HTML tags like `<kbd>`, `<abbr>`, `<code>` are now recognized as separate segments
- Prevents HTML tags from interfering with capitalization detection
- Supports self-closing tags and tags with attributes
- **MD063 (heading-capitalization): Preserve caret notation for control characters**
- Caret notation like `^A`, `^C`, `^Z` (representing Ctrl+A, Ctrl+C, Ctrl+Z) is preserved
- Prevents incorrect lowercasing of control character representations
- Supports full range: `^@` through `^_` and `^A` through `^Z`
## [0.0.200] - 2025-12-23
### Added
- **CLI: Add `--no-defaults` flag to `rumdl config` command** (fixes #218)
- Show only non-default configuration values, excluding defaults
- Useful for seeing only customizations without the noise of default values
- Works with both smart output (with provenance annotations) and `--output toml/json` formats
- Mutually exclusive with `--defaults` flag
- Displays helpful message when all configurations are using defaults
- Supports filtering of global config, rule configs, and per-file-ignores
- **SARIF: Map severity field to SARIF result levels**
- Error severity maps to SARIF "error" level
- Warning severity maps to SARIF "warning" level
- Improves integration with tools that consume SARIF output
- **Testing: Add O(n²) complexity regression testing infrastructure**
- Detect potential quadratic time complexity issues in rules
- Prevent performance regressions in line analysis algorithms
### Fixed
- **MD051 (link-fragments): Support extension-less cross-file links (GitHub-style)** (fixes extension-less link validation)
- Recognize extension-less paths with fragments as cross-file links (e.g., `[link](page#section)` → `page.md#section`)
- Resolve extension-less paths by trying markdown extensions (`.md`, `.markdown`, `.mdx`, etc.)
- Handles GitHub-style links that omit file extensions, matching GitHub's rendering behavior
- Applied to both link detection and cross-file validation
- Reduces false positives in repositories using extension-less links
- **MD054 (link-image-style): Fix false positives for task lists and HTML tags** (fixes false positive detection)
- Skip validation for links/images inside task list items (e.g., `- [ ] [link](url)`)
- Skip validation for links/images inside HTML tags (e.g., `<a href="url">text</a>`)
- Reduces false positives when links are part of task lists or HTML structures
- **MD057 (link-target-exists): Skip email addresses and fix absolute path bug** (fixes email and path handling)
- Skip validation for email addresses (e.g., `mailto:user@example.com`)
- Fix bug where absolute paths were incorrectly validated
- Improves accuracy by correctly identifying email links vs file links
- **MD063 (heading-capitalization): Respect lowercase-words when inline code/link is last segment** (fixes #223)
- When the last word in a heading is inline code or a link, respect `lowercase-words` configuration
- Previously, inline code/link at the end was not checked against lowercase-words list
- Ensures consistent capitalization rules apply to all heading segments
- **MD007 (ul-indent): Fix indent=1 edge case and integrate edge case tests** (fixes edge cases)
- Fix handling of `indent = 1` configuration edge cases
- Integrate comprehensive edge case tests for better coverage
- Improves reliability for non-standard indent configurations
- **MD057 (link-target-exists): Point diagnostics to URL instead of link text** (fixes #226)
- LSP diagnostics now point to the actual non-existent URL rather than the link text
- Fixes incorrect positioning in editors where diagnostics appeared at wrong location
- Makes diagnostics semantically correct by highlighting the actual problem
- Improves both CLI and LSP diagnostic accuracy
- **File processor: Canonicalize file paths to prevent duplicate warnings**
- Normalize file paths to resolve symbolic links and relative paths
- Prevents duplicate warnings when the same file is referenced via different paths
- Improves accuracy when processing files via symlinks or relative paths
### Performance
- **MD007 (ul-indent): Cache mixed list nesting detection and optimize blank line checks**
- Cache results of mixed list nesting detection to avoid redundant calculations
- Optimize blank line checks for better performance on large files
- Reduces processing time for files with complex list structures
### Changed
- **Rules: Categorize severities by document impact**
- Reorganized rule severities into meaningful categories
- Structure-critical rules (MD001, MD002, MD003, etc.) remain Error
- Style-preference rules (MD004, MD007, MD029, etc.) now Warning
- Better alignment between severity and actual impact on document quality
- **Documentation: Add severity categorization section to RULES.md**
- Documents the reasoning behind severity assignments
- Explains the distinction between structural errors and style warnings
- Helps users understand why rules have their assigned severity levels
- **Tests: Reorganize issue-specific tests into appropriate test files**
- Moved Issue #210 tests to `tests/rules/md007_test.rs`
- Moved Issue #209 tests to `tests/rules/md007_test.rs`
- Moved Issue #197 tests to `tests/cli_integration_tests.rs`
- Removed `issueXXX`-named test files per project conventions
## [0.0.199] - 2025-12-21
### Fixed
- **MD007 (ul-indent): Smart style auto-detection for custom indent values** (fixes #210)
- When `indent` is set to a non-default value (e.g., `indent = 4`) without explicit `style`, MD007 now auto-detects the appropriate style
- Pure unordered lists automatically use `fixed` style (markdownlint compatible)
- Mixed ordered/unordered lists use `text-aligned` style to avoid oscillation with MD005
- Previously, `indent = 4` was ignored when style wasn't explicitly set, causing incorrect validation
- The auto-detection handles edge cases including multi-level mixed nesting, HTML comments, code blocks, frontmatter, and blank line list separation
- **MD063 (heading-capitalization): Fix preserve-cased-words for iOS and ignore-words for first word** (fixes #215, #216)
- `preserve-cased-words` now correctly detects words starting with lowercase followed by uppercase (e.g., "iOS", "eBay")
- `ignore-words` in sentence case now properly preserves words at the start of headings (e.g., "nvim" stays "nvim")
- Previously, "iOS" was incorrectly converted to "ios" and "nvim config" became "Nvim config"
### Changed
- **Dependencies: Migrate from deprecated serde_yaml to serde_yml**
- Replace unmaintained serde_yaml 0.9.34 with maintained fork serde_yml 0.0.12
- Eliminates deprecation warnings during compilation
- Compatible API with no behavior changes
## [0.0.198] - 2025-12-20
### Fixed
- **MD057 (link-target-exists): Support mdBook and doc site .html links**
- Links to `.html` files now correctly resolve when corresponding `.md` source exists
- Handles mdBook, Jekyll, Hugo and similar doc generators that compile `.md` to `.html`
- Applied to both single-file `check()` and cross-file validation
- Reduces false positives in documentation projects like rustc-dev-guide
## [0.0.197] - 2025-12-19
### Fixed
- **Config: Resolve per-file-ignores paths relative to project root** (fixes #208)
- Previously, per-file-ignores patterns only matched with relative paths
- Now works correctly in GitHub Actions where absolute paths are used
- Patterns like `.github/file.md` now match `/home/runner/work/repo/.github/file.md`
- **MD007 (ul-indent): Remove auto-switch to fixed indentation style** (fixes #209)
- Fixed oscillation between MD005/MD007 when formatting mixed ordered/unordered lists
- Previously, setting `indent = 3` would auto-switch to fixed style, causing conflicts
- Text-aligned style (default) now correctly handles mixed lists, matching markdownlint behavior
- `rumdl fmt` now converges in a single pass for all list configurations
### Performance
- **Core: Reduce memory allocations and improve cache reliability**
- Optimized internal data structures for better memory efficiency
## [0.0.196] - 2025-12-19
### Added
- **MD064 (no-multiple-consecutive-spaces): New rule to collapse multiple consecutive spaces** (thanks @mawkler in #202)
- Flags multiple consecutive spaces (2+) in markdown prose
- Provides auto-fix to collapse them to single space
- Skips spaces in code, tables, HTML, front matter, and markdown syntax markers
- **MD065 (blanks-around-horizontal-rules): New rule to enforce blank lines around horizontal rules**
- Ensures horizontal rules have blank lines before and after
- Provides auto-fix to add missing blank lines
- Skips Setext heading underlines
- **MD066 (valid-footnote-definition): New rule to validate footnote definitions**
- Checks footnote definitions have proper syntax
- Validates footnote labels are non-empty
- **MD067 (no-empty-footnote-ref): New rule to flag empty footnote references**
- Detects footnote references with empty labels like `[^]`
- **MD068 (no-space-in-footnote-ref): New rule to flag spaces in footnote references**
- Detects footnote references with spaces like `[^ note]`
### Fixed
- **WASM: Convert byte offsets to character offsets for JavaScript**
- Fixes crashes in Obsidian plugin when applying fixes to content with multi-byte UTF-8 characters (Norwegian letters, emoji, etc.)
- **MD027 (no-multiple-space-blockquote): Skip list continuation indentation**
- No longer reports false positives for properly indented list continuations in blockquotes
- **MD032 (blanks-around-lists): Detect thematic breaks as list terminators**
- Thematic breaks (horizontal rules) now correctly terminate lists
- **MD036 (no-emphasis-as-heading): Skip emphasis inside HTML comments**
- No longer reports false positives for emphasis markers in HTML comments
- **MD041 (first-line-heading): Skip badge images before headings**
- Badge images at the start of a file no longer prevent heading detection
- **MD046 (code-block-style): Skip code blocks inside HTML comments**
- No longer reports false positives for code blocks in HTML comments
- **MD051 (link-fragments): Recognize HTML and attribute anchors**
- Cross-file fragment validation now recognizes anchors from HTML id attributes
- **MD057 (relative-links): Check for markdown source when .html link doesn't exist**
- Links to `.html` files now check for corresponding `.md` source files
## [0.0.195] - 2025-12-17
### Added
- **MD063 (heading-capitalization): New rule for heading capitalization style**
- Enforces consistent capitalization in headings (Title Case, Sentence case, or lowercase)
- Preserves all-caps acronyms (e.g., "API", "HTTP", "AWS") in title case mode
- Configurable via `MD063.style` option
### Fixed
- **MD012 (no-multiple-blanks): Detect blank lines across skipped regions**
- Fixed detection of consecutive blank lines when frontmatter or code blocks are between them
- **MD013 (line-length): Check individual line lengths in reflow default mode**
- Each line is now checked separately when reflow-mode is enabled
- **MD014 (commands-show-output): Improve message accuracy**
- Better error messages and principled skip list for code blocks
- **MD018/MD023: Skip GitHub issue refs and hashtags in heading detection**
- Lines like `#123` or `#hashtag` no longer falsely detected as ATX headings
- **MD033 (no-inline-html): Fix false positives**
- Skip angle brackets inside link reference definition titles
- Remove incorrect indentation-based code block skip
- **MD034 (no-bare-urls): Fix multiple edge cases**
- Detect `www.example.com` as bare URL
- Handle link reference definitions with titles
- Resolve false positives in code spans
- **MD037 (no-space-in-emphasis): Skip template shortcode syntax**
- Template syntax like `{* ... *}` no longer triggers false positives
- **MD042 (no-empty-links): Only flag empty URLs**
- Links with text but empty URL are flagged; empty text is no longer an error
- **MD049 (emphasis-style): Report both opening and closing markers**
- Warnings now point to both `*` or `_` markers in emphasis pairs
- **MD056 (table-column-count): Escape pipes in inline code**
- Pipes inside inline code in tables no longer cause column count errors
- **MD057 (relative-links): Multiple path handling improvements**
- Handle URL-encoded filenames (e.g., `%20` for spaces)
- Support framework aliases (e.g., `@/` for src directory)
- Handle angle-bracket paths (e.g., `<path/to/file.md>`)
- Skip validation of absolute URL paths
- Strip query parameters from URLs before file existence check
- **MD063: Preserve all-caps acronyms in title case headings**
- Words like "API", "HTTP", "URL" maintain uppercase in title case mode
- **CLI: Fix exit code and messaging**
- Return exit code 0 when all violations are fixed
- Show accurate message when all violations are fixed
- **Heading detection: Skip multiline link syntax**
- Lines inside multiline link definitions no longer falsely detected as headings
- **Internal: Add is_valid field and iterator for CommonMark headings**
- Improved heading validation for rules that depend on heading structure
### Changed
- **MD059: Update alias to 'descriptive-link-text'** (thanks @martimlobao in #201)
- Rule can now be referenced as `descriptive-link-text` in configurations
### Documentation
- **MD050: Document intentional deviation from markdownlint**
- Clarified differences in emphasis style checking behavior
## [0.0.194] - 2025-12-12
### Added
- **LSP: workspace/didChangeConfiguration support**
- LSP server now responds to configuration change notifications
- Improved robustness and error handling in language server
- **Config: Typestate pattern for configuration validation**
- Compile-time guarantees for configuration correctness
- Invalid configurations are caught earlier in the pipeline
- **CI: GitHub Marketplace branding for GitHub Action**
- Action now has proper branding for Marketplace visibility
### Fixed
- **MD057 (relative-links): Handle template variables and extensionless links**
- Template syntax (Handlebars, Mustache, Jinja2) like `{{URL}}` now recognized and skipped
- Extensionless links like `[Page](page)` now resolve to `page.md`
- All URI schemes (file://, smb://, macappstores://, etc.) properly excluded (fixes #192)
- **Config: Remove misleading GFM/CommonMark warnings (fixes #195)**
- `gfm` and `commonmark` flavors now silently work as expected
- Parser already supports GFM extensions (tables, task lists, strikethrough)
- **Config: Discover markdownlint config via upward directory traversal**
- `.markdownlint.yaml` and similar files now found in parent directories
- Matches behavior of other linting tools
- **MD007 (ul-indent): Handle tab characters after blockquote markers**
- Tab indentation after `>` marker now calculated correctly
- Fixes false positives for lists in blockquotes with tabs
- **MD012 (no-multiple-blanks): Skip blank lines inside indented code blocks**
- Blank lines within indented code blocks no longer trigger warnings
- **MD030 (list-marker-space): Analyze blockquoted list content correctly**
- Lists inside blockquotes now properly analyzed instead of skipped
- **MD053 (link-image-reference-definitions): Recognize colons in reference labels**
- Reference definitions with colons in the label (e.g., `[foo:bar]:`) now detected
- **lint_context: Check entire line for HTML comment marking**
- Lines with inline comments (`<!-- comment --> content`) now handled correctly
- Content after `-->` on the same line is properly processed
- **lint_context: Add missing HTML5 block elements**
- Added: audio, video, source, track, canvas, svg, iframe, embed, object, menu, summary, noscript, search, template
- **lint_context: Handle nested blockquotes and JS comments in ESM blocks**
- Improved parsing for complex nested structures
- **table_utils: Exclude list items from table row detection**
- List items containing pipes no longer confused with table rows
- **lib: Handle mutex poisoning gracefully instead of panicking**
- Recovers from poisoned mutexes for better stability
### Changed
- **Internal: Replace `Mutex<Option<Arc<T>>>` with `OnceLock`**
- Cleaner initialization pattern for shared state
- **Internal: Replace tuple return with `FixResult` struct**
- More descriptive return type for fix operations
### Documentation
- **README: Clarify markdownlint config upward traversal behavior**
## [0.0.193] - 2025-12-10
### Fixed
- **MD032 (blanks-around-lists): Complete fix for false positives (fixes #188, #190)**
- Pipes in code spans (e.g., `` `foo || bar` ``) no longer trigger false table detection
- Lazy continuation lines at indent=0 handled per CommonMark spec
- Removed ad-hoc URL/link exclusion hacks in favor of proper `is_table_line()` detection
- Fixed regression in v0.0.192 that affected list boundary detection
- **MD057 (relative-links): Resolve links per-file instead of caching base path**
- Relative link resolution now correctly uses each file's directory
- Prevents cross-file path resolution errors in multi-file projects
### Changed
- **Internal: Replace naive `contains('|')` checks with `is_table_line()`**
- Structural separator detection now properly validates table structure
- Requires pipes at start/end with ≥2 total, or separator line pattern
## [0.0.192] - 2025-12-10
### Added
- **Config: `cache` option to disable caching via config file (closes #153)**
- Add `cache = false` to `[global]` section to persistently disable caching
- Eliminates need to pass `--no-cache` flag on every invocation
- Works in both `.rumdl.toml` and `pyproject.toml`
- **MD052 (reference-links-images): User-configurable `ignore` option**
- Add patterns to skip specific reference links from validation
- Useful for links handled by external tools or preprocessors
- **CI: Major version tag (`v0`) for GitHub Action**
- Users can now reference the action as `rvben/rumdl@v0` for automatic minor updates
### Fixed
- **Inline config: Cross-file rules now respect disable comments (fixes #189)**
- MD051 and MD057 now honor `<!-- rumdl-disable -->` and `<!-- rumdl-disable-next-line -->` comments
- Previously these rules ran in a separate phase that bypassed inline config filtering
- **Config: Directory exclude patterns now match files within (fixes #186)**
- Patterns like `docs/` now correctly exclude all files in that directory
- Glob patterns are automatically expanded to match contained files
- **MD032 (blanks-around-lists): No longer triggers on pipes in inline code (fixes #188)**
- Tables inside inline code blocks no longer confuse list boundary detection
- **MD005/MD037/MD041: Resolve false positives**
- MD041: Setext headers with colons no longer cause issues
- MD037: List items containing asterisks handled correctly
- MD005: Sublist parsing improved for complex nested structures
- **Text reflow: Use proper table detection**
- Replaced simplistic pipe check with robust table structure detection
- Prevents incorrect reflow of table content
### Changed
- **Internal: Replace `RUMDL_FILE_PATH` env var with `source_file` field**
- Cleaner API for accessing the current file path in rule implementations
- No user-facing changes
## [0.0.191] - 2025-12-08
### Fixed
- **CLI: Handle broken pipe gracefully when piping to head (fixes #184)**
- Output to closed pipes (e.g., `rumdl check . | head`) no longer causes error messages
- SIGPIPE is now handled properly on Unix systems
- **Config: Resolve exclude patterns relative to config file location (fixes #185)**
- Exclude patterns like `docs/*` now resolve relative to the project root (config file directory)
- Previously patterns resolved relative to the current working directory, causing them to fail when running from a different directory with `--config`
- This matches the behavior of Biome, Black, and markdownlint-cli2
- **Note:** Users running from the project root (the common case) are unaffected
- **Docs: Correct build status badge to track release workflow**
## [0.0.190] - 2025-12-06
### Added
- **GitHub Action: New `rumdl-action` for CI integration**
- Supports `version`, `path`, and `config` inputs
- Enables easy integration of rumdl into GitHub workflows
- **CLI: Add `--rules` as alias for `--enable`**
- More intuitive option name for specifying which rules to run
- Example: `rumdl check --rules MD001,MD003 file.md`
### Fixed
- **LSP: Detect client diagnostic capabilities to avoid duplicate diagnostics (fixes #182)**
- LSP server now checks if the client supports `textDocument/publishDiagnostics`
- Prevents duplicate diagnostics when client pulls diagnostics instead of receiving pushes
- **MD005 (list-indent): Group sublists by parent content column**
- Correctly handles ordered lists where marker widths vary (e.g., "1. " vs "10. ")
- Sublists under parents with different content columns are now grouped separately
- Prevents false positives when indent differences are caused by marker width variation
- **MD037/MD038: Handle multi-byte UTF-8 characters in emphasis detection**
- Emphasis detection now correctly handles characters like CJK and emoji
- Prevents false positives and incorrect column reporting for non-ASCII content
- **MD057 (existing-relative-links): Handle absolute paths as workspace-relative**
- Absolute paths in markdown (e.g., `/CONTRIBUTING.md`) are now resolved relative to workspace root
- Fixes false positives for links that reference files from the repository root
- **MD057 (existing-relative-links): Add filesystem fallback for case-sensitivity issues**
- Improves link validation on case-insensitive filesystems
- Reduces false positives on macOS and Windows
- **CLI: Canonicalize config path before changing working directory**
- Config files specified with relative paths now work correctly
- Fixes issue where config was not found after directory change
## [0.0.189] - 2025-12-05
### Fixed
- **MD061 (link-destination-exists): Default `case_sensitive` to `true` when loading from TOML**
- Configuration parsing now correctly defaults to case-sensitive path checking
- Aligns TOML config behavior with programmatic defaults
- **MD022 (blanks-around-headings): Treat HTML comments and frontmatter as transparent**
- HTML comments between content and headings no longer trigger blank line warnings
- Frontmatter is now invisible to blank line checking around headings
- **MD033 (no-inline-html): Skip code span content in multi-line span detection**
- HTML tags inside multi-line code spans are no longer flagged
- Prevents false positives for code examples containing HTML
- **MD032 (blanks-around-lists): Handle multi-line code spans in list detection**
- Code spans spanning multiple lines no longer confuse list boundary detection
- Fixes false positives for lists near complex code span usage
- **MD044 (proper-names): Skip HTML comments by default**
- Proper name checking now excludes HTML comment content
- Reduces noise from commented-out content
- **MD029 (ol-prefix): Use pulldown-cmark for list membership detection**
- More accurate detection of which items belong to which ordered list
- Handles edge cases with code blocks and other interrupting elements
### Changed
- **MD057 (table-pipe-style): Remove unused `skip_media_files` config option**
- Simplifies configuration by removing non-functional option
## [0.0.188] - 2025-12-03
### Added
- **MD062 (link-destination-whitespace): New rule for detecting whitespace in link destinations**
- Flags links with leading/trailing whitespace in destinations like `[text]( url )`
- Supports auto-fix to trim the whitespace
- Works in both CLI and LSP environments
- **Cross-file analysis infrastructure: Foundation for multi-file linting**
- New `WorkspaceIndex` for indexing headings and anchors across files
- Background index worker for LSP with debouncing and progress reporting
- Reverse dependency tracking for efficient re-linting when files change
- MD051 now validates cross-file link fragments against the workspace index
### Fixed
- **MD051 (link-fragments): Fix false positives for cross-file links without fragments**
- Links like `[text](file.md)` no longer incorrectly flagged for empty fragments
- Only validates fragments when explicitly provided (e.g., `[text](file.md#anchor)`)
## [0.0.187] - 2025-12-02
### Added
- **WASM: Implement Linter class API with configuration support**
- New `Linter` class provides object-oriented API for WebAssembly builds
- Supports configuration options for customizing lint behavior
- Enables more flexible integration in JavaScript/TypeScript projects
### Fixed
- **MD053 (link-image-reference-definitions): Fix false positives for adjacent footnotes (fixes #177)**
- Footnote references like `[^1][^2]` no longer trigger "unused reference" warnings
- Uses pulldown-cmark for proper footnote parsing instead of regex
- **MD034 (no-bare-urls): Use pulldown-cmark for proper link parsing**
- Bare URL detection now correctly identifies URLs already inside links
- Prevents false positives for URLs that are properly formatted
- **MD010 (no-hard-tabs): Skip tabs in fenced code blocks**
- Tabs inside fenced code blocks are no longer flagged
- Preserves intentional tab characters in code examples
- **MD052 (reference-links-images): Disable shortcut reference checking by default**
- Shortcut references like `[text]` without a following `[]` no longer trigger warnings
- Reduces false positives in documents using shortcut reference style
- **MD025/MD023: Fix false positives after single-line code blocks (fixes #175)**
- Comments like `# text` inside code blocks after single-line code blocks no longer flagged
- Code block parsing now correctly handles edge cases
- **MD013/MD039: Fix reflow oscillation causing unstable output (fixes #170)**
- Running `rumdl fmt` multiple times now produces stable output
- Link formatting no longer conflicts with line length reflow
### Changed
- **Tests: Align test expectations with CommonMark spec and markdownlint reference**
- Test suite updated to match authoritative behavior
- Ensures compatibility with CommonMark specification
## [0.0.186] - 2025-12-01
### Added
- **WebAssembly support: Build and publish rumdl as npm package**
- Use rumdl in browsers and Node.js via `@anthropic/rumdl` npm package
- Includes `lint()` function for linting markdown content
- Automated npm publishing via GitHub Actions
### Fixed
- **MD058 (blanks-around-tables): Fix duplicating table content during auto-fix**
- Auto-fix no longer duplicates table content when adding blank lines
- Correctly handles tables at various positions in the document
- **MD013 (line-length): Preserve nested linked images during text reflow**
- Markdown images inside links like `[](url)` now preserved during line wrapping
- Previously, the reflow logic could break such nested constructs
### Changed
- **Docs: Document rumdl-fmt pre-commit hook with version check**
- Added documentation for the `rumdl-fmt` hook in pre-commit configuration
- Includes version compatibility information
## [0.0.185] - 2025-11-28
### Added
- **Inline config: Support rule aliases in inline configuration comments**
- Use human-readable aliases like `line-length` instead of `MD013` in inline comments
- Example: `<!-- rumdl-disable line-length -->` now works alongside `<!-- rumdl-disable MD013 -->`
- All rules now document their aliases in the rule documentation
### Fixed
- **MD061 (forbidden-terms): Register rule in mod.rs**
- MD061 was added in v0.0.184 but was never registered, causing "Unknown rule" errors
- The rule now works correctly for detecting forbidden terms like TODO, FIXME, etc.
## [0.0.184] - 2025-11-28
### Added
- **MD061 (forbidden-terms): New rule to flag forbidden terms in documents**
- Configurable list of forbidden terms with case-sensitivity options
- Useful for style guides, compliance, and consistency enforcement
### Fixed
- **Code blocks: Respect CommonMark fence indentation limit (0-3 spaces)**
- CommonMark spec limits fenced code block indentation to 0-3 spaces relative to container
- Document-level fences must have at most 3 spaces of indentation
- Fences inside list items are correctly recognized (indentation is relative to list content)
- Prevents false negatives when checking content inside deeply nested structures
- **LSP: Use whole-table fixes for table rules (MD055, MD056, MD058, MD060)**
- Table rule fixes now replace the entire table instead of individual cells
- Prevents fix corruption when multiple cells need adjustment
- Improves reliability of auto-fix in IDEs
## [0.0.183] - 2025-11-27
### Fixed
- **MD056 (table-column-count): Respect MkDocs flavor for pipes in inline code (fixes #165)**
- In MkDocs/Python-Markdown flavor, pipes inside backticks are NOT cell delimiters
- Tables with inline code containing pipes (e.g., `` `x | y` ``) no longer trigger false positives
- GFM flavor behavior unchanged (pipes in code ARE delimiters per spec)
- **MD060 (table-format): Respect MkDocs flavor for pipes in inline code**
- Consistent with MD056 fix for MkDocs flavor support
- Table formatting now correctly handles inline code with pipes in MkDocs mode
- **Schema: Use kebab-case for GlobalConfig properties**
- JSON schema now uses kebab-case (e.g., `line-length`) matching config file conventions
- Improves IDE auto-completion and validation for configuration files
### Changed
- **Refactor: Consolidate table row parsing into TableUtils**
- Unified table parsing logic into shared utility module
- Reduces code duplication between MD056 and MD060
- **Refactor: Decompose MD013 into module structure**
- Split large `md013_line_length.rs` into separate files:
- `mod.rs`: Main rule implementation
- `helpers.rs`: Helper functions
- `tests.rs`: Unit tests
- Improves code organization and maintainability
## [0.0.182] - 2025-11-25
### Added
- **cache clean command**
- Added `rumdl clean` command to clear the cache directory
- Useful for forcing re-linting of all files
- **MD013 (line-length): Visual width mode for CJK and emoji**
- Added `visual-width` option that counts CJK characters as double-width
- Properly handles emoji and other wide characters for accurate line length checking
- Useful for East Asian language documentation
- **MD013 (line-length): Custom abbreviation support**
- Added `abbreviations` config option for defining custom abbreviations
- Custom abbreviations are respected during sentence-per-line reflow
- Prevents unwanted line breaks after user-defined abbreviations
- **Configuration file discovery: .config/rumdl.toml support**
- Added support for `.config/rumdl.toml` as a valid configuration location
- Follows XDG Base Directory-style pattern for project-level config
- Discovery order: `.rumdl.toml` → `.config/rumdl.toml` → `pyproject.toml`
- **MD022 docs: Per-level configuration documentation**
- Added documentation for per-level `lines-above` and `lines-below` arrays
- Shows how to specify different spacing requirements for each heading level (h1-h6)
### Fixed
- **MD042 (no-empty-links): Wiki-style links no longer flagged as empty (fixes discussion #153)**
- Wiki-style links like `[[Page Name]]`, `[[Folder/Page]]`, and `[[Page|Display Text]]` are now recognized
- Prevents false positives for Obsidian/Notion-style wiki link syntax
- Block references like `[[#^block-id]]` are also handled correctly
- **MD056 (table-column-count): Handle escaped pipes correctly in GFM tables (fixes #163)**
- GFM table parsing now correctly handles backslash-escaped pipes (`\|`)
- Inline code spans do NOT protect pipes from being cell delimiters (per GFM spec)
- Only `\|` is treated as literal pipe content
- **MD060 (table-format): Accept tables that are already aligned (fixes #164)**
- Tables with consistent column widths across all rows are now accepted as aligned
- Short separator rows (e.g., `:--` instead of `:---`) no longer trigger reformatting
- Reduces false positives for manually aligned tables
### Changed
- **Documentation: Corrected timestamps in CHANGELOG for versions 0.0.172-0.0.178**
## [0.0.181] - 2025-11-20
### Added
- **cache-dir configuration option**
- Added support for `cache-dir` in `.rumdl.toml` and `pyproject.toml` config files
- Precedence: CLI `--cache-dir` → `RUMDL_CACHE_DIR` env var → config file → default
- Follows Ruff's configuration pattern for cache directory management
- **Termux installation support**
- Added Termux User Repository install method to README (thanks @ha1ix in #157)
### Changed
- **BREAKING: Cache directory renamed from `.rumdl-cache` to `.rumdl_cache`**
- Aligns with Ruff's naming convention (`.ruff_cache`)
- Follows Rust ecosystem preference for underscores over hyphens
- Migration: Old `.rumdl-cache` directories can be safely deleted
### Fixed
- **LSP: Exclude Unfixable rules from formatting and Fix All (fixes #158)**
- Filter warnings from Unfixable rules (like MD033) before applying fixes during LSP formatting
- Prevents unintended destructive changes (e.g., HTML deletion) during document formatting
- Unfixable rules still available through Quick Fix actions for user choice
- Fix All action now correctly excludes Unfixable rules
- **Cache directory now created at project root, not CWD (fixes #159)**
- Cache directory is now anchored to the project root (determined by `.git` location)
- Prevents multiple cache directories when running from subdirectories
- Works consistently whether invoked from project root or any subdirectory
- Behavior:
- **With .git**: Cache at `.git` parent directory (true project root)
- **Without .git**: Cache at config file location (fallback)
- Follows Ruff's fix (PR #7962) for consistent cache placement
## [0.0.180] - 2025-11-19
### Added
- **MD022 (blanks-around-headings): Per-level unlimited blank handling**
- Added per-level configuration to allow unlimited blank lines around specific heading levels
- Enables fine-grained control over blank line requirements for different heading levels
- Supports both global and per-level unlimited blank line settings
### Fixed
- **Range utilities: Handle UTF-8 character boundaries safely (fixes #154)**
- Fixed panic when processing files with multi-byte UTF-8 characters (CJK, emoji)
- String slicing operations now safely handle byte indices that fall in the middle of multi-byte characters
- Added comprehensive test suite covering Korean, Chinese, Japanese characters, emoji, and edge cases
- All string slicing operations now occur at valid UTF-8 character boundaries
- **MD041 (first-line-heading): HTML comments are now correctly skipped (fixes #155)**
- HTML comments before headings are now treated as non-visible content and ignored
- Aligns with documentation and markdownlint behavior
- Multi-line HTML comments are properly handled using parser metadata
- Fixed handling of deeply nested HTML headings
## [0.0.179] - 2025-11-18
### Fixed
- **MD041 (first-line-heading): Fixed multiline HTML detection for headings**
- Replace manual HTML parsing with centralized `ctx.html_tags()` parser
- Correctly handles nested HTML elements with multiline attributes
- Fixes Issue #152: No longer incorrectly flags headings when preceded by multiline HTML
- **MD037 (no-space-in-emphasis): Improved warning message clarity**
- Truncate long emphasis text in warning messages to prevent display issues
- Long emphasis spans are now shown with ellipsis for better readability
### Performance
- **MD005 (list-indent): Eliminated O(n²) complexity by pre-computing parent relationships**
- Pre-compute parent list item relationships during initial parsing
- Reduces algorithmic complexity from O(n²) to O(n) for deeply nested lists
- Improves performance for files with complex list structures
- **Optimized link/image regex patterns to prevent catastrophic backtracking**
- Added atomic grouping and possessive quantifiers to regex patterns in `lint_context.rs`
- Prevents regex engine from exponential backtracking on pathological inputs
- Improves parsing speed for files with many links and images
### Changed
- **MD033 (no-inline-html): Refactored to use centralized HTML parser**
- Removed 109 lines of broken `find_multiline_html_tags()` function that detected 0 multiline HTML tags
- Replaced two-pass approach (single-line regex + multiline detection) with single-pass using `ctx.html_tags()`
- Now correctly detects all multiline HTML tags (previously missed 100% of them)
- Follows pattern from MD041 refactoring for consistency
- Preserves all existing filters (code blocks, kramdown, comments, etc.)
### Added
- **Comprehensive MD005 test coverage**
- Added 21 new edge case tests for list indentation detection
- Updated unicode test to verify dynamic parent relationship detection
- Ensures robustness for nested and complex list structures
## [0.0.178] - 2025-11-17
### Performance
- **Critical: Eliminated O(n²) bottleneck in list block parsing (Issue #148)**
- Replaced nested loop with forward-scanning O(n) algorithm using state tracking
- Performance improvement: 900-line files with nested lists now parse in ~20ms (previously 50+ seconds)
- User-reported case: 890-line file improved from 50.2s to ~0.2s (250x speedup)
- User-reported case: 1780-line file improved from 347s to ~0.4s (867x speedup)
- Algorithm now scales linearly O(n) instead of quadratically O(n²)
- Uses two tracking variables to detect list-breaking content and validate continuation indentation
- Extracted `reset_tracking_state()` helper to eliminate code duplication
- Replaced magic number 2 with `UNORDERED_LIST_MIN_CONTINUATION_INDENT` constant
### Changed
- **Code cleanup: Removed unused O(n²) trap methods**
- Removed `links_on_line()` and `images_on_line()` methods from `LintContext`
- These methods were never used but could cause performance issues if called in loops
### Added
- **Comprehensive regression tests for list block parsing**
- 27 new tests covering edge cases: consecutive items, list-breaking content, indentation rules
- Performance tests verify O(n) scaling for 1000+ item lists
- Tests reproduce exact Issue #148 pattern (nested lists with brackets)
## [0.0.177] - 2025-11-17
### Fixed
- **MD003 (heading-style): Fixed non-deterministic test failure**
- Eliminated flaky behavior caused by HashMap iteration order when multiple heading styles had equal counts
- Added deterministic tiebreaker that prefers ATX style over Setext styles
- Ensures consistent style detection across all test runs
- **MD044 (proper-names): Added international character support**
- Added tilde characters (ã, õ) and Nordic å to ASCII normalization
- Enables proper name matching for international cities like São Paulo, Coimbra, and Århus
- **MD013 (line-length): Fixed infinite hang in sentence-per-line reflow mode**
- Fixed abbreviation detection causing hang on words ending in letter sequences
- Corrected logic to properly detect abbreviations vs. sentence endings
- Issue #150: Resolves infinite loop that occurred with certain text patterns
- **Table style detection improvements**
- MD055/MD056/MD058: Now analyze all table rows to detect style, not just first row
- Prevents false positives when first row has different separator style than rest of table
- **Code quality fixes**
- Resolved Issues #151 and #152 with comprehensive test coverage
- Eliminated code duplication in MD005, MD042, and other rules
- Extracted shared helper methods to reduce maintenance burden
### Added
- **MD013: Comprehensive abbreviation detection tests**
- 42 test cases covering sentence-per-line reflow edge cases
- Validates correct behavior for abbreviations, acronyms, and punctuation patterns
- **Text reflow test organization**
- Separated 42 integration tests into dedicated test file (`text_reflow_test.rs`)
- Follows established codebase pattern for test organization
- Improves maintainability while keeping 1 private helper test inline
### Changed
- **MD029 (ordered-list-prefix): Improved warning messages**
- Messages now include contextual style information
- Shows both configured style and detected document style for clarity
- Examples: "configured style 'one'" vs "document style 'ordered'"
- **MD029 rule name standardization**
- Removed deprecated MD029-style suffix variant
- Unified to single consistent "MD029" identifier across codebase
### Performance
- **Memory optimization in MD005 and MD042**
- Reduced memory allocations in list consistency and link reference checks
- More efficient data structure usage
## [0.0.176] - 2025-11-14
### Added
- **Git-style intelligent configuration merging**
- User and project configs can now intelligently combine instead of simply overriding
- `disable` array uses union semantics - user can add to project disables
- `enable` array uses replace semantics - project can enforce specific rules
- When a rule appears in both `enable` and `disable`, `enable` wins (project can override user)
- Configuration hierarchy is now explicit and follows Git's pattern:
- Default (0) < UserConfig (1) < PyprojectToml (2) < ProjectConfig (3) < CLI (4)
- Renamed config sources for clarity:
- `ConfigSource::RumdlToml` → `UserConfig` (global user config)
- `ConfigSource::RumdlToml` → `ProjectConfig` (project-level config files)
- Removed `Markdownlint` source (now uses `ProjectConfig`)
- New API methods:
- `SourcedValue::merge_union()` - additive merging for disable arrays
- `SourcedValue::merge_override()` - replacement merging for enable arrays
- Example use cases:
- User disables MD013 globally → Project enables MD013 → MD013 is enabled ✓
- Project disables MD001, MD003 → User disables MD013 → All three are disabled ✓
- 17 comprehensive tests covering merge semantics, precedence, and conflict resolution
- Matches configuration pattern used by git, eslint, prettier, and other modern tools
### Performance
- **Major memory optimizations eliminating string allocations**
- Converted `LineInfo` to use byte ranges instead of owned strings (50-80% memory reduction)
- Eliminates N string allocations where N = number of lines in document
- Added zero-copy `content(&self, source: &str) -> &str` method for on-demand access
- Converted `LineIndex` to borrow `&'a str` instead of cloning (eliminates one full document copy per lint operation)
- Converted `ParsedLink`/`ParsedImage` to use `Cow<'a, str>` (60-80% reduction in heap allocations for link/image parsing)
- Zero-cost when borrowing, minimal overhead when owning - most data borrowed directly from source
- **Eliminated O(n²) complexity in multiple rules**
- Fixed quadratic bottlenecks in MD027 (Multiple spaces after blockquote symbol)
- Fixed quadratic bottlenecks in MD020 (No space inside hashes on closed atx style heading)
- Fixed quadratic bottlenecks in MD046 (Code block style)
- All three rules now use pre-computed context data with O(1) lookups
- **Optimized CLI commands**
- `rumdl config` now executes in ~18ms (eliminated duplicate rule instantiation)
- Removed duplicate instantiation of ~50 rule objects between main.rs and formatter.rs
- Cleaner architecture with single source of truth for rule instances
### Fixed
- **Config loading: User config now always loaded as base layer (#131)**
- Fixed LSP server ignoring user config when finding project-level config files
- User configuration is now always loaded first (unless `--no-config` is used)
- Project configs merge on top of user config, CLI flags have highest priority
- Configuration hierarchy now consistent between CLI and LSP:
1. User/global config (`~/.config/rumdl/rumdl.toml`) - base layer
2. Project config (discovered or explicit) - overrides user config
3. CLI flags - highest priority
- Matches pattern used by git, eslint, prettier, and other tools
- Added regression test verifying user config preserved with explicit project config
- **MD035: Frontmatter delimiter false positives (#40)**
- Fixed incorrect flagging of YAML/TOML frontmatter delimiters (`---`/`+++`) as horizontal rules
- Rule now correctly skips frontmatter in three places:
- `most_prevalent_hr_style()` - don't count frontmatter HRs for prevalence
- `check()` - don't flag frontmatter delimiters as violations
- `fix()` - don't replace frontmatter delimiters when fixing
- Uses pre-computed `LineInfo.in_front_matter` field for efficient detection
### Changed
- **Architecture: Consistent use of pre-computed context data**
- MD035 now uses pre-computed `LineInfo.in_front_matter` instead of function calls
- All rules now consistently use `ctx.line_index` for line-based operations
- Eliminates redundant function calls and O(n) scans
- More efficient with O(1) field access patterns
- **Test infrastructure improvements**
- Implemented dynamic fixture downloading for performance tests
- Test downloads now happen on-demand instead of being checked into repository
- Better handling of large test files for performance benchmarking
- **Test quality improvements**
- Corrected ESM block test assertions to match actual MDX behavior
- ESM blocks only exist at TOP of MDX files and end at first non-ESM line
- Tests now assert correct behavior rather than documenting implementation quirks
- Fixed MD033 test assertions to properly validate HTML inline rules
### Breaking Changes
**For library users only** (CLI users unaffected):
- `LineInfo.content` field is no longer public - use new `content(source: &str)` method instead
- Since `LineInfo` was primarily internal API, impact should be minimal
## [0.0.175] - 2025-11-12
### Added
- **Universal wiki-link and Obsidian block reference support**
- Added recognition of `[[wiki-links]]` across all rules
- Added support for Obsidian block references `[[note#^block-id]]`
- Prevents false positives in link/heading detection rules
- Works with all extended markdown flavors that support wiki-links
- **Type-safe configuration wrappers**
- Added `HeadingLevel` type with validation (1-6 range)
- Added type-safe wrappers for MD007, MD009, MD010 configuration
- Added type-safe configuration for MD013 and MD030
- Added type-safe configuration for MD022, MD012, MD060
- Compile-time validation prevents invalid configuration values
- Backward compatible with snake_case field names via serde aliases
- **Memory profiling for non-Linux platforms**
- Added cross-platform memory profiling support
- Uses physical memory instead of virtual memory for benchmarking
- Enables performance analysis on macOS and Windows
### Fixed
- **MD041: mdBook preprocessor directive false positives**
- MD041 now skips files containing only mdBook preprocessor directives (e.g., `{{#include file.md}}`)
- These composition/routing files are not standalone content and don't require headings
- Eliminates false positives on mdBook include-only files
- Supports all mdBook directives: `{{#include}}`, `{{#playground}}`, `{{#rustdoc_include}}`, etc.
- Handles files with directives mixed with HTML comments
- Added comprehensive test coverage for mdBook directive patterns
- **MD042: Inline code in link text false positives**
- Fixed incorrect flagging of code spans within link text
- MD042 now correctly handles patterns like `[code with \`backticks\`](url)`
- Prevents false positives when links contain inline code
- Added regression test covering this case
- **Config import: markdownlint option mapping (#137)**
- `rumdl import` now correctly maps markdownlint-specific option names to rumdl equivalents
- MD013: Maps `stern` → `strict`, warns about incompatible options (code_block_line_length, heading_line_length)
- MD054: Warns about incompatible `style`/`styles` options (rumdl uses individual boolean flags)
- Prevents "Unknown option" warnings when importing markdownlint configs
- Added helpful warnings explaining config model differences
- **MD011: Footnote reference false positives (#147)**
- Fixed incorrect flagging of valid markdown: `[link](url)[^footnote]`
- MD011 now skips footnote references (starting with `^`)
- Already had support for reference links, now extended to footnotes
- Added regression test covering this case
- **MD013: Front matter line length (#146)**
- MD013 now correctly skips YAML/TOML front matter blocks
- Prevents false positives on long lines in document metadata
- Works with both `---` (YAML) and `+++` (TOML) front matter delimiters
- Added regression test covering this case
- **Config system: kebab-case support**
- MD041: Converted to serde-based config with kebab-case support
- Added backward compatibility aliases for snake_case field names
- Now properly supports both `front-matter-title` and `front_matter_title`
- Removed undocumented config options from MD032 and MD038
- **MD004: Simplified configuration**
- Refactored to use derives instead of manual implementation
- Cleaner code while maintaining full functionality
- **Test infrastructure improvements**
- Migrated from deprecated `Command::cargo_bin` to `cargo_bin_cmd!` macro
- Fixed Homebrew installation docs (removed unnecessary `brew tap`)
### Changed
- **Consistency checking: Prevalence-based style detection**
- Converted MD003, MD004, MD046, MD048, MD049, MD050, MD055 from "first-found" to "prevalence-based" logic
- Rules now count ALL occurrences and enforce the most commonly used style
- Tie-breakers prefer industry standards (GitHub, CommonMark recommendations)
- More accurate representation of document's actual style intent
- Better user experience: rules adapt to what you mostly use, not just the first instance
- Affected rules:
- MD003 (heading-style): ATX vs Setext vs Closed ATX
- MD004 (ul-style): dash, asterisk, plus, or consistent
- MD046 (code-block-style): fenced vs indented
- MD048 (code-fence-style): backticks vs tildes
- MD049 (emphasis-style): asterisk vs underscore
- MD050 (strong-style): asterisk vs underscore
- MD055 (table-pipe-style): leading/trailing pipe consistency
- Updated documentation to reflect prevalence-based behavior
- **Dependencies**
- Upgraded schemars from 0.8 to 1.1
- Updated all dependencies to latest compatible versions
- Updated dependencies to latest patch versions
## [0.0.174] - 2025-11-10
### Fixed
- **MD046: False positives with multi-paragraph footnotes (#142)**
- Fixed incorrect flagging of indented continuation paragraphs in footnote definitions
- Added comprehensive footnote detection following CommonMark footnote extension spec
- Implements proper block continuation context tracking for footnotes
- Prevents confusion between actual indented code blocks and footnote continuations
- Example that now works correctly:
```markdown
[^1]: First paragraph of footnote.
Second paragraph (indented, but not a code block).
```
- Added 11 specification-based tests ensuring robust footnote handling
- **CLI: Differentiate `--quiet` and `--silent` flags (#141)**
- `--quiet`: Shows diagnostics but suppresses progress/summary (matches Ruff behavior)
- `--silent`: Suppresses all output including diagnostics (exit code still reflects violations)
- Previously both flags behaved identically, hiding all output
- Updated help text to clarify the distinction
- **LSP: Server now respects `--config` argument (#140)**
- Added proper support for `--config` flag in `rumdl server` command
- Config file path validation with clear error messages for missing files
- Config path properly passed through entire LSP initialization stack
- Enables custom configuration files for LSP/editor integrations
## [0.0.173] - 2025-11-08
### Added
- **MD060: Table format rule with Prettier-style auto-compact**
- New rule enforcing consistent table column alignment (aligned/compact/tight styles)
- **Auto-compact threshold**: Tables exceeding max-width automatically use compact formatting
- Configurable via `max-width` setting (0 = inherit from MD013's line-length)
- Generates informative warnings showing actual width vs threshold
- Disabled by default (opt-in feature)
- Handles edge cases: zero-width characters, escaped pipes, HTML comments
- Respects column alignments (left/center/right) in aligned mode
- **MD043: Wildcard pattern support for heading structures**
- New wildcard patterns for flexible heading structure validation
- Allows `*` placeholders in heading text for dynamic content
- Edge case handling for complex heading hierarchies
- **MD044: HTML elements configuration option**
- New `html-elements` configuration to customize proper name handling
- Allows project-specific proper name enforcement
- **RUMDL_CACHE_DIR environment variable**
- Override default cache directory location via environment variable
- Useful for CI/CD pipelines and custom cache management
### Fixed
- **Definition list support in reflow mode (#136)**
- Definition lists (Extended Markdown syntax) are now correctly preserved during text reflow
- Prevents incorrect joining of terms with definitions (e.g., `Term\n: Definition` no longer becomes `Term : Definition`)
- Supports PHP Markdown Extra, Kramdown, Pandoc, Hugo, and other extended Markdown flavors
- Works with all reflow modes including sentence-per-line
- **MD054: Error message alignment**
- Corrected error messages to match implementation behavior
- Documentation now accurately reflects link/image style validation
- **MD041: Front matter title config schema**
- Added missing `front_matter_title` options to configuration schema
- Fixes config validation errors when using front matter title feature
- **LSP: pyproject.toml validation**
- LSP now verifies `[tool.rumdl]` section exists before using pyproject.toml
- Prevents errors when pyproject.toml exists but doesn't contain rumdl config
- **CLI: Rule registration**
- Fixed MD057, MD059, MD060 rules not being registered in CLI
- All rules now properly available for use
### Changed
- **Documentation: TOML format migration**
- Converted all configuration examples from YAML to TOML
- Reflects modern configuration best practices
- Improved clarity for MD013 reflow requirement in README
## [0.0.172] - 2025-11-06
### Added
- **MD059: Link text style rule achieving 100% markdownlint compatibility**
- New rule for enforcing consistent link text formatting
- Matches markdownlint column reporting behavior
- Completes full parity with markdownlint rule set
- **pulldown-cmark BrokenLink callback integration**
- Enhanced link reference validation using parser callback
- Better detection of undefined reference-style links
- Improved accuracy for MD052 and related rules
### Fixed
- **Critical: UTF-8 character boundary panic in ordered list detection**
- Fixed crash when processing markdown with multi-byte UTF-8 characters (Japanese, Chinese, Korean, etc.)
- Root cause: Character index was incorrectly used as byte index for string slicing
- Now converts character positions to byte positions using `char_indices().nth()`
- Discovered during validation on javascript-algorithms repository (193k ⭐)
- Affected any non-English documentation with multi-byte UTF-8 near numbered lists
- **pulldown-cmark escaped bracket workaround**
- Workaround for pulldown-cmark bug where `\[` and `\!` are incorrectly parsed as links/images
- Violates CommonMark spec Example 14 for backslash escapes
- Filters 90% of false positives with byte-level escape pattern detection
- Correctly handles: `\[escaped\]`, `\`
- Known limitation: `\[text][ref]` reference-style links still produce 1 false positive
- Bug report filed for upstream fix
- **MD042: False positives on autolinks**
- Fixed empty link detection incorrectly flagging autolinks like `<https://example.com>`
- Autolinks are now correctly excluded from empty link checks
- **MD033: HTML tags inside HTML comments**
- Fixed false positives when HTML tags appear inside HTML comments
- Centralized HTML comment detection for consistent behavior across rules
- HTML comments now properly ignored: `<!-- <div>not flagged</div> -->`
- **MD051: GitHub ASCII normalization**
- Implemented proper GitHub anchor generation with ASCII normalization
- Handles edge cases like backtick-wrapped angle brackets: `` `<FILE>` `` → `#file`
- Verified against actual GitHub.com rendering behavior
- **MD052: Complete Jinja template support**
- Added Jinja template range checks to regex-based shortcut reference detection
- Fixes false positives on `[reference]` patterns in Jinja templates
- Completes Jinja support across all link-related rules (MD039, MD042, MD051, MD052)
- **MD039/MD042/MD051: Jinja template checks**
- Added Jinja template range checks to prevent false positives
- All rules using `ctx.links`/`ctx.images` now skip Jinja template markers
- Validated on real-world Jinja templates (pyo3/.towncrier.template.md)
- **Link/image text extraction: Whitespace preservation**
- Fixed loss of leading/trailing whitespace in link text and image alt text
- pulldown-cmark strips newlines from Text events, causing whitespace loss
- Now extracts directly from source bytes between brackets for perfect fidelity
- Verified: `[\nNewline\n](url)` → text correctly preserved with newlines
### Changed
- **⚠️ BREAKING: Removed MD002 (first-heading-h1) and MD006 (ul-start-left) rules**
- Removed for 100% markdownlint compatibility
- markdownlint deprecated these rules in favor of more specific alternatives
- **Migration**: Remove MD002 and MD006 from your `.rumdl.toml` config if present
- **Refactoring: Modernized codebase with Rust 1.91.0 features**
- Migrated `once_cell::sync::Lazy` to `std::sync::LazyLock` (stable in Rust 1.80)
- Improved error handling by replacing `unwrap()`/`expect()` with proper error propagation
- Replaced defensive fallbacks with `unreachable!()` where appropriate
- Eliminated unreachable panics and improved code robustness
- Removed `_fix` boolean redundancy and cleaned up dead code
- **Refactoring: HTML comment detection centralized**
- MD033 now uses centralized `html_utils::find_html_comments()` for consistency
- Eliminates duplicate HTML comment parsing logic
### Performance
- **12-15% speed improvement from pulldown-cmark migration**
- Migrated link/image parsing from regex-based approach to pulldown-cmark parser
- Benchmark results: 12-15% faster on representative markdown files
- More accurate CommonMark compliance
- Better handling of edge cases and complex markdown structures
## [0.0.171] - 2025-11-03
### Added
- **MD060: Table column formatting with markdownlint compatibility**
- Full table formatting rule with auto-fix capability
- Four formatting styles: `aligned`, `compact`, `tight`, and `any` (default)
- `any` style detects existing formatting and enforces consistency
- Supports CJK characters and emoji
- Handles alignment indicators (`:---`, `:---:`, `---:`)
- Skips ZWJ emoji sequences to prevent corruption
- Configuration: `MD060.enabled = false` (opt-in), `MD060.style = "any"`
### Fixed
- **MD053: CommonMark compliance for list continuations**
- Fixed false positives when backtick references appear in list continuation paragraphs
- Enhanced ordered list marker detection for multi-digit lists (10., 123.)
- Implemented column-based indentation model per CommonMark spec
- Distinguishes list continuation paragraphs from code-in-lists
- Validated against 1,562 markdown files with 0 false positives
- **LSP: Global config fallback**
- Fixed global config loading when no project-specific config exists
- Now properly falls back to `~/.config/rumdl/.rumdl.toml`
- Added test coverage for fallback behavior
- **LSP: Fixed test expectations for camelCase serialization**
- LSP initialization options use camelCase per LSP specification
- Corrected test assertions to match actual camelCase behavior
- **Fix mode: Type-safe exit code semantics**
- Added `FixMode` enum for clear fix/check mode distinction
- Improved code maintainability and type safety
### Changed
- **⚠️ BREAKING: MD013 `tables` default changed to `false`**
- Previously defaulted to `true`, causing conflicts with table formatting rules
- Now defaults to `false` for better user experience
- **Migration**: If you rely on MD013 checking table line lengths, add to `.rumdl.toml`:
```toml
[MD013]
tables = true
```
- This prevents false positives when using MD060 or other table formatting tools
## [0.0.170] - 2025-10-31
### Added
- **Include non-standard file extensions** (#127)
- New `--include` CLI flag to check files with non-standard extensions
- Example: `rumdl check --include "*.txt" --include "*.text"`
- Useful for documentation files with custom extensions
- Respects `.rumdl.toml` configuration: `include = ["*.txt", "*.text"]`
### Fixed
- **MD055: Preserve user formatting when fixing pipe placement** (#129)
- Changed from full table reconstruction to surgical pipe addition/removal
- User's intentional spacing and alignment are now preserved
- Follows Unix philosophy: do one thing (fix pipes) well
- Example: `| Cell 1 | Cell 2` → `| Cell 1 | Cell 2 |` (spacing preserved)
- **MD053: Allow backtick references with `::` and spaces** (#128)
- Fixed false positives for Rust-style references like `` [`std::vec::Vec`] ``
- Now correctly handles references containing `::` within backticks
- Improves accuracy for Rust documentation and technical content
### Documentation
- **MD028: Clarified rule behavior for consecutive blockquotes** (#126)
- Updated documentation to accurately reflect that MD028 flags consecutive blockquote *starts*
- Not a bug - working as designed per markdownlint specification
- Prevents confusion about expected behavior
### Changed
- **Rust toolchain updated to 1.91.0** (from 1.89.0)
- Ensures compatibility with latest Rust features and improvements
- Better performance and compilation times
## [0.0.169] - 2025-10-30
### Performance
**MASSIVE PERFORMANCE IMPROVEMENTS**: This release delivers 7-53x faster linting through systematic elimination of O(n²) bottlenecks and algorithmic optimizations. rumdl is now 16-29x faster than
markdownlint-cli2 on real-world repositories.
- **Fix catastrophic O(n²) bottleneck in FilteredLinesIter** ⭐ THE GAME CHANGER
- Eliminated `content.lines().collect()` from inside Iterator::next() method
- Single-file optimization delivered 7,600x speedup for affected rules
- MD011: 6.00s → 0.785ms (7,644x faster)
- MD012: 5.95s → 1.643ms (3,621x faster)
- Impact: Stress test (10k lines): 9,987ms → 187ms (53x faster)
- Impact: Rust Book (478 files): 2,646ms → 269ms (9.8x faster)
- Impact: PyO3 (76 files): 2,004ms → 262ms (7.6x faster)
- **Pre-compute LintContext data structures to eliminate redundant work**
- LineIndex: Eliminated 46× content cloning across all rules
- Jinja template ranges: Pre-computed once instead of O(n×m) scanning
- Table blocks: Computed once instead of 4× independent scans per document
- HTML comment ranges: Pre-computed with 50-70x speedup
- Code block line mapping: Optimized with binary search instead of linear scan
- Autodoc blocks (MkDocs): Pre-computed to avoid O(n²) scaling
- **Optimize link and image parsing with binary search**
- Replaced linear search with binary search for code span checks
- Faster link/image parsing across all reference-based rules
- **Replace regex with byte-level parsing for better performance**
- List detection: Manual byte scanning instead of regex
- Blockquote prefix: Byte-level parsing instead of regex
- Line info parsing: Eliminated redundant blockquote prefix parsing
- **Streaming parser architecture**
- Replaced AST parsing with pulldown-cmark streaming for lower memory overhead
- Better scalability for very large documents
- **MD046 optimization**
- Pre-compute list/tab contexts
- Use ctx.code_blocks instead of independent detection
### Benchmarks
**vs Previous Version (v0.0.168):**
- Rust Book (478 files): 2,646ms → 269ms (9.8x faster)
- Stress Test (10,514 lines): 9,987ms → 187ms (53.4x faster)
- PyO3 (76 files): 2,004ms → 262ms (7.6x faster)
**vs markdownlint-cli2 (industry standard):**
- Rust Book: 10.3s vs 0.35s (29.4x faster)
- PyO3: 4.4s vs 0.26s (16.9x faster)
**Current Performance:**
- 0.7ms per file (Rust Book)
- 6.3x parallelization efficiency
- Zero O(n²) algorithmic bottlenecks remaining
### Added
- **MDX and Quarto flavor support**
- New flavors: `MarkdownFlavor::MDX` and `MarkdownFlavor::Quarto`
- JSX component handling: Skip linting inside JSX elements
- ESM import/export support: Ignore JavaScript import/export statements
- Quarto code chunk support: Recognize `{r}`, `{python}` chunk delimiters
- Essential for modern React-based documentation and data science workflows
### Fixed
- **MD011: Intelligent URL vs text detection in reversed links**
- Smarter classification of link components as URL or text
- Reduces false positives in edge cases like `(Generic<T>)[link]`
- Correctly handles `(http://url)[text]` vs `(text)[#anchor]`
- Improved heuristics for ambiguous single-word patterns
- **MD013: Preserve HTML blocks in list items during reflow**
- HTML blocks inside list items are now preserved during auto-fix
- Prevents broken indentation and tag structure
- Better handling of complex nested content
### Changed
- Architecture improvements for pre-computation pattern
- LintContext now owns shared data structures
- Rules consume pre-computed data instead of duplicating work
- Consistent binary search APIs for range lookups
- Foundation for future optimizations
## [0.0.168] - 2025-10-28
### Added
- **CLI: Support for Quarto (.qmd) and RMarkdown (.rmd/.Rmd) files**
- Added support for `.qmd` (Quarto), `.rmd` and `.Rmd` (RMarkdown) file extensions
- Enables markdown linting for data science and scientific documentation workflows
- Particularly useful for Jupyter-based publishing and reproducible research documents
- **LSP: will_save_wait_until for proper auto-fix on save**
- Implemented LSP `textDocument/willSaveWaitUntil` capability
- Provides proper auto-fix on save support in compatible editors
- More reliable than `textDocument/didSave` approach
- Ensures fixes are applied before file is actually saved to disk
- **MD033: LSP Quick Fix to remove HTML tags while keeping content**
- New code action removes HTML opening and closing tags while preserving inner content
- Helps convert inline HTML to plain text when needed
- Available through editor Quick Fix menu (Ctrl+. or Cmd+.)
### Fixed
- **MD013: Skip auto-fix for list items containing HTML tags**
- Prevents broken auto-fix when HTML tags are present in list items
- HTML structure (indentation, tag hierarchy) is now preserved
- Errors are still reported but no destructive fix is applied
- Pragmatic solution until full HTML-aware reflow is implemented
- Fixes issues where tags like `</details>` were split or indentation was lost
- **MD013: Fix incorrect sentence splitting after abbreviations in sentence-per-line mode**
- No longer incorrectly splits sentences after common abbreviations (e.g., etc., i.e., Dr., Mr.)
- Improved sentence detection algorithm for better accuracy
- Prevents unwanted line breaks in the middle of sentences
- **MD013: Prevent autolinks from being parsed as HTML tags**
- Autolinks like `<https://example.com>` are no longer treated as HTML
- Fixes false positives in HTML detection that could skip auto-fix unnecessarily
- Properly distinguishes between markdown autolinks and actual HTML tags
- **MD013: Prevent content duplication in sentence-per-line reflow**
- Fixed bug where content could be duplicated during paragraph reflow
- Ensures each sentence appears exactly once in reflowed output
- Improves reliability of auto-fix in sentence-per-line mode
- **MD013: Improve sentence-per-line error messages and highlighting**
- Better error messages that clearly indicate sentence-per-line violations
- More accurate highlighting of problematic text
- Helps users understand what needs to be fixed
- **MD013: Skip template directives at paragraph start in sentence-per-line mode**
- Template directives (like `{{ variable }}`) at start of paragraphs no longer cause issues
- Prevents false positives in templated markdown files
- Improves compatibility with static site generators and template engines
- **MD013: Treat template directives as paragraph boundaries**
- Template directives now properly separate paragraphs during reflow
- Prevents template syntax from being joined with regular content
- Better handling of mixed template and markdown content
- **MD013: Join single-sentence paragraphs in sentence-per-line mode**
- Single-sentence paragraphs that span multiple lines are now properly joined
- Fixes issues where short paragraphs were incorrectly flagged
- Improves consistency of sentence-per-line formatting
- **MD013: Handle multiple spaces and multi-line paragraphs in sentence-per-line mode**
- Better handling of paragraphs with inconsistent spacing
- Multi-line paragraphs are now correctly reflowed
- Fixes edge cases in whitespace handling
- **MD052: Preserve backtick-wrapped patterns with dots in MkDocs mode**
- Backtick-wrapped references like `` `[foo.bar]` `` now preserve dots in slugs
- Matches MkDocs behavior of treating code-wrapped text literally
- Prevents false positives for code examples in documentation
- **MD052: Support Pandoc citations and inline footnotes in RMarkdown/Quarto**
- Recognizes Pandoc citation syntax: `[@citation]`, `@citation`, `[-@citation]`
- Recognizes inline footnote syntax: `^[footnote text]`
- No longer treats these as reference-style links
- Essential for academic and scientific writing in RMarkdown and Quarto
- **MD033: Only report opening HTML tags, not closing tags**
- MD033 violations now only report the opening tag of an HTML element
- Reduces noise in linting output (one violation per element vs two)
- Closing tags like `</div>` are no longer separately reported
- Makes HTML-related warnings clearer and less redundant
- **MD018: Skip CSS selectors and JS code inside HTML blocks**
- CSS selectors like `#slide-1` inside `<style>` tags no longer trigger MD018
- JavaScript code inside `<script>` tags is properly ignored
- Prevents false positives for Quarto and RMarkdown files with embedded HTML/CSS/JS
- Particularly important for interactive documents and custom styling
- **MD012: Enforce exactly 1 newline at EOF and clean LSP logging**
- Files now must end with exactly one newline character (not zero, not two)
- Aligns with POSIX standard and common editor behavior
- Improved LSP logging with less noise
- **CI: Use cargo run instead of release binary in rumdl pre-commit hook**
- Pre-commit hooks now use `cargo run` for more reliable execution
- Prevents issues with stale release binaries during development
- Better integration with cargo workflow
- **Schema: Allow root-level rule sections like [MD013] in config validation**
- Configuration schema now correctly allows `[MD013]` style sections
- Fixes false validation errors for valid TOML config
- Better compatibility with common configuration patterns
### Changed
- **Refactor: Detect HTML blocks before parsing headings**
- Reordered LintContext initialization to detect HTML blocks first
- Headings are no longer detected inside HTML blocks (like `<style>` or `<script>`)
- Architectural improvement that prevents multiple false positives
- Benefits MD018 and other heading-related rules
- More correct parsing aligned with markdown specifications
- **Refactor: Use filtered_lines() for front-matter handling in MD011 and MD012**
- Improved front-matter handling using consistent filtered_lines API
- Better code reuse and maintainability
- More reliable front-matter detection across rules
## [0.0.167] - 2025-10-24
### Added
- **Configuration validation with fuzzy-match suggestions**
- Comprehensive unknown key detection for `.rumdl.toml` and `pyproject.toml`
- Intelligent "did you mean?" suggestions using Levenshtein distance algorithm
- File path context in validation warnings for easy debugging
- Catches typos in global options, rule names, and rule options
- Example: `line-length` → suggests `line-length`, `reflw` → suggests `reflow`
- Zero-dependency implementation with configurable similarity threshold
- Helps users catch configuration mistakes before they cause confusion
- **MD053: Support for community comment-style references**
- Recognizes and ignores reference-style link syntax used as comments
- Supports widely-used patterns: `[//]: # (comment)`, `[comment]: #`, `[note]: #`, `[todo]: #`, `[fixme]: #`, `[hack]: #`
- Any reference with just `#` as URL is treated as a comment
- While not in CommonMark/GFM specs, used across 23+ markdown implementations
- Complements HTML comments with a less HTML-like syntax option
- Improves compatibility with existing markdown practices
- **MD013: `line-length = 0` to disable all line length checks**
- Setting `line-length = 0` now completely disables MD013 rule
- Provides explicit way to turn off line length validation entirely
- More intuitive than previous workarounds
- Useful when line length management is handled by other tools or not desired
- **MD051: mdbook template support**
- Added detection and slug generation for mdbook templates
- Recognizes `{{#template path/to/file.md}}` syntax
- Properly generates GitHub-compatible slugs for template-included headings
- Improves compatibility with mdbook documentation projects
- **LSP: Manual "Reflow paragraph" code action for MD013 warnings**
- New code action available for MD013 line length warnings when auto-reflow is disabled
- Allows users to manually reflow specific paragraphs without enabling global reflow in config
- Appears as "Reflow paragraph" in Quick Fix menu (not marked as preferred, so won't trigger on save)
- Intelligently detects paragraph boundaries and reflows entire paragraph, not just the flagged line
- Respects line length limit from warning message or defaults to 80 characters
- Provides a way to try paragraph reflow before committing to enabling it globally
- Gives users fine-grained control over which paragraphs to reflow
### Fixed
- **LSP: Preserve trailing newline in reflow action**
- Manual reflow code action now correctly preserves trailing newlines
- Prevents unwanted file modifications from reflow operations
- Maintains document structure integrity
- **LSP: Improve logging and resolve auto-fix issues**
- Enhanced LSP server logging for better debugging
- Resolved various auto-fix edge cases and reliability issues
- **MD051: Correct GitHub slug generation for angle brackets**
- Fixed incorrect slug generation for headings containing angle brackets
- Now properly handles special characters in anchor generation
- Improves accuracy of link validation for complex headings
### Changed
- **MD033: Remove unhelpful message suffix**
- Simplified warning messages for inline HTML detection
- Removed redundant information to reduce noise
- Cleaner, more focused error messages
- **Code cleanup: Remove dead code**
- Removed unused `LinkImageStyle` enum from MD054
- General refactoring to improve maintainability
- Fixed clippy warnings
### Documentation
- **MD033: Document mdbook use case for semantic HTML**
- Added documentation about using semantic HTML in mdbook projects
- Clarifies when and why inline HTML might be intentionally used
- Helps users understand legitimate use cases for HTML in markdown
## [0.0.166] - 2025-10-22
### Added
- **MD013: `paragraphs` field to control paragraph line length checks** (resolves #121)
- New boolean config field `paragraphs` (defaults to `true`) allows disabling line length warnings for paragraph text
- Enables sentence-per-line formatting workflows without line length validation noise
- Still checks headings, tables, code blocks, blockquotes, and HTML when `paragraphs: false`
- Useful for semantic line breaks where sentence length is determined by content, not arbitrary limits
- Example configuration:
```yaml
MD013:
paragraphs: false # Don't warn about long paragraphs
code-blocks: true # Still check code blocks
tables: true # Still check tables
reflow: true
reflow-mode: "sentence-per-line"
```
## [0.0.165] - 2025-10-21
### Fixed
- **MD040: Always preserve indentation when adding language tags** (fixes #122)
- The MD040 rule was incorrectly removing indentation from code blocks when adding language tags
- This broke list structure when code blocks were part of list items
- Root cause: The fix logic had conditional behavior that would remove indentation for "standalone" code blocks
- Now always preserves original indentation regardless of context
- Removed 50+ lines of unnecessary `is_in_nested_context()` helper logic
- Added comprehensive tests for various indentation scenarios (0, 2, 4, 6 spaces)
### Added
- **Conventional Commits validation hook**: Git commit-msg hook validates commit message format
- Enforces Conventional Commits specification for all commits
- Provides helpful error messages for invalid formats
- Ensures consistent commit history for changelog generation
- **Automated changelog generation with git-cliff**:
- Added `make changelog-draft` for previewing CHANGELOG updates
- Semi-automated workflow: generate draft, enhance with details, commit
- Conventional Commits integration for automatic categorization
### Changed
- **Pre-push hook optimization**: Use dev profile instead of full suite for faster testing
- Prevents pre-push hook from hanging on slower machines
- Maintains adequate test coverage while improving developer experience
## [0.0.164] - 2025-10-21
### Added
- **File-Level Caching (Ruff-inspired)**: Dramatic performance improvements for repeat runs
- Blake3-based content hashing for fast cache lookups
- Automatic cache invalidation on content, config, or version changes
- Cache stored in `.rumdl-cache/{version}/{hash}.json`
- CLI flags: `--no-cache` to disable, `--cache-dir` to customize location
- Enabled by default for instant subsequent runs
- **Thread-Safe Parallel Caching**: Best of both worlds - parallelization AND caching
- Implemented Arc<Mutex<LintCache>> for safe cache sharing across threads
- Mutex locked ONLY for brief cache get/set operations
- Full parallelization during expensive linting operations
- Matches Ruff's architecture for optimal performance
- **Convergence Detection**: Added hash-based detection to identify when fixes have stabilized
- Stops iteration when content hash remains unchanged
- More efficient than counting rule applications
- Returns convergence status in fix results
- **Convergence Failure Warnings**: Report when auto-fix doesn't converge (Ruff-style)
- Warns if 100 iteration limit reached without convergence
- Shows rule codes involved in potential infinite loop
- Encourages bug reports for convergence failures
- Available via `RUMDL_DEBUG_FIX_PERF` environment variable
### Changed
- **Auto-fix Iteration**: Automatic iteration until convergence (fixes #88)
- `--fix` now automatically iterates up to 100 passes until content stabilizes (same as Ruff)
- No need to manually re-run `rumdl check --fix` multiple times
- Hash-based convergence detection prevents unnecessary iterations
- Significantly improves user experience for multi-pass fix scenarios
- **Unified Linting Architecture**: Removed ~60 lines of duplicate linting logic
- Refactored `process_file_collect_warnings` to use `process_file_inner`
- Single code path for all file processing
- Cache works for ALL output formats (text, JSON, GitLab, SARIF, JUnit)
- **Parallel File Processing for Fix Mode**: 4.8x speedup on multi-file fixes
- Previously fix mode was always sequential
- Now uses parallel processing when safe (multiple independent files)
- Each file processes all its fix iterations independently
### Fixed
- **Multi-pass Fixes**: No longer require manual re-runs to apply all possible fixes
- Previously users had to run `rumdl check --fix` multiple times
- Now automatically handles dependent rule fixes in single command
- Examples: MD010 (tabs) before MD007 (list indent), MD013 (line length) before MD009 (trailing spaces)
- **Cache Correctness**: Include enabled rules in cache key (Ruff-style)
- Cache now respects `--enable`/`--disable` CLI flags
- Different rule configurations create separate cache entries
- Prevents incorrect cached results when switching rule sets
- Changed `LintWarning.rule_name` from `Option<&'static str>` to `Option<String>` for proper serialization
- **Cache Parallelization**: Cache now works correctly with parallel processing
- No mutex contention during parallel file processing
- All output formats benefit from caching (previously only JSON/GitLab/SARIF/JUnit)
### Performance
- **Single file with cache**: 943ms → 7ms (135x faster)
- **Multi-file (21 files) cold cache**: 14.4s → 4s (parallel processing)
- **Multi-file (21 files) warm cache**: 14.4s → 0.019s (757x faster!)
- **JSON format (17 files) with cache**: 13.9s → 60ms (231x faster)
## [0.0.163] - 2025-10-20
### Changed
- **MD024**: Default `siblings_only` to true for better usability
- Multiple headings with same text now only flagged if they're direct siblings
- Reduces false positives in documents with common section headings
- More intuitive default behavior matching common use cases
### Fixed
- **MD013**: Enforce line length in sentence_per_line mode (fixes #111)
- Previously, sentence_per_line mode completely ignored line_length setting
- Now warns about single sentences exceeding configured line_length
- No auto-fix for long single sentences (requires manual rephrasing)
- Still auto-fixes multi-sentence lines by splitting on sentence boundaries
- Cleaned up warning messages by removing verbose parentheticals
- Maintains semantic integrity (won't split mid-sentence) while respecting configured line_length
- **HTML Comments**: Complete fix to ignore all content inside HTML comments (fixes #119, #20)
- All rules now properly ignore content within HTML comment blocks (`<!-- ... -->`)
- Added `in_html_comment` field to `LineInfo` for comprehensive tracking
- Extended filtered lines API with `skip_html_comments()` method
- Updated MD013, MD049, and other rules to skip HTML comment content
- Prevents false positives from commented-out markdown (MD013, MD049, MD005, MD006, MD039, MD042)
- Better handling of multi-line HTML comments across all linting rules
- **MD046**: Resolve false positives from Issue #118
- Fixed incorrect flagging of valid code block syntax
- Improved code block style detection accuracy
- **MD050**: Resolve false positives from Issue #118
- Fixed incorrect strong style detection in edge cases
- Better handling of emphasis patterns
- **Tests**: Fixed sentence_per_line_detection test assertion
- Updated test to match simplified warning message from MD013
- Test was expecting verbose message after message was simplified in earlier commit
## [0.0.162] - 2025-10-16
### Added
- **Filtered Line Iterator Architecture**: New infrastructure for rule implementation
- Provides consistent interface for filtering out front matter, code blocks, and HTML blocks
- Eliminates manual context checking in individual rules
- Improves code maintainability and reduces duplication
- Enables easier implementation of new rules
### Fixed
- **MD052**: Skip code blocks in blockquotes when checking references
- Prevents false positives for reference syntax inside code blocks within blockquotes
- Properly handles nested markdown structures
- **MD034**: Skip URLs in front matter
- URLs in YAML/TOML/JSON front matter no longer flagged as bare URLs
- Improves compatibility with static site generators
- **Tests**: Fixed flaky `profiling::tests::test_concurrent_access` test
- Added `#[serial_test::serial]` attribute to prevent race conditions
- Ensures reliable test execution in CI/CD environments
- **Documentation**: Build badge now displays correctly
### Performance
- **MD005**: Optimized continuation detection from O(n²) to O(n)
- Dramatically faster processing of documents with many list items
- Eliminates redundant line scanning
- **General**: Consolidated multiple `line_info()` calls for same line
- Reduced redundant lookups across multiple rules
- Improved overall linting performance
### Changed
- **Internal Refactoring**: Eliminated manual checks across all rules
- Removed manual front matter detection from individual rules
- Removed manual code block detection from individual rules
- Removed manual HTML block detection from individual rules
- All rules now use centralized filtering infrastructure
### Documentation
- **Per-File-Ignores**: Added comprehensive documentation for per-file-ignores feature
- Detailed usage examples with glob patterns
- Integration with both `.rumdl.toml` and `pyproject.toml`
## [0.0.161] - 2025-10-15
### Added
- **MD013**: Support for backslash hard line breaks for mdformat compatibility (closes #110)
- Backslash (`\`) at end of line now recognized as hard break alongside two-space breaks
- Original hard break format (backslash or spaces) preserved during reflow operations
- Segment-based reflow correctly handles both hard break types
- Comprehensive test coverage including mdformat compatibility tests
- Enables seamless migration from mdformat to rumdl
### Fixed
- **MD029**: Recognize properly indented nested content as list continuation
- Nested list items and paragraphs within list items now correctly identified
- Improved detection of list item boundaries
- Better handling of complex list structures
- **MD013**: Preserve semantic line breaks and fix false positives in normalize mode
- Semantic line breaks (intentional breaks for readability) now preserved
- Reduced false positives when lines are intentionally kept short
- Better detection of paragraph boundaries in normalize mode
- **MD044**: Invert code-blocks logic to match MD013 and change default to false
- Parameter logic now consistent: `true` = check code blocks, `false` = skip code blocks
- Default changed to `false` (skip code blocks) for better user experience
- Aligns with MD013's code block handling for consistency across rules
## [0.0.160] - 2025-10-15
### Fixed
- **Configuration**: Fixed `rumdl init --pyproject` command to no longer create `.rumdl.toml` file
- The command now correctly only adds rumdl configuration to `pyproject.toml`
- Prevents confusion from having duplicate configuration files
- **MD044**: Corrected field name in templates and documentation
- Fixed inconsistency in proper names configuration
- Improved accuracy of documentation examples
- **Configuration System**: Added field aliases and validation warnings for all rules
- Better backwards compatibility with alternative field names
- Helpful warnings guide users to correct configuration syntax
- Improved user experience when migrating configurations
## [0.0.159] - 2025-10-14
### Added
- **JSON Schema Generation**: New `rumdl schema` subcommand for generating JSON schema from configuration
- `rumdl schema generate` - Generate/update the schema file
- `rumdl schema check` - Verify schema is up-to-date (used in CI)
- `rumdl schema print` - Print schema to stdout
- Schema automatically generated from Rust types using `schemars`
- Prepared for SchemaStore submission to enable IDE autocomplete/validation
### Fixed
- **MD051 False Positives**: Fixed incorrect handling of backtick headings with angle brackets
- Previously treated `<FILE>` inside backticks as HTML tags and stripped them
- Now correctly processes headings like `` `import <FILE> [OPTIONS]` `` → `import-file-options`
- Removed premature `strip_html_tags()` call; anchor algorithms now handle both markdown and HTML correctly
- Added regression tests for backtick headings with special characters
- Fixes false positives in README.md table of contents
### Changed
- **Code Cleanup**: Removed unused `generate_schema` binary (functionality moved to `rumdl schema` subcommand)
## [0.0.158] - 2025-10-14
### Fixed
- **CRLF Line Ending Support**: Fixed byte position calculations in multiple rules for Windows-style line endings
- Fixed MD034, MD046, MD057 byte position calculations
- Fixed MD037, MD049, MD011 byte position calculations
- Fixed MD050, MD037, MD010, MD026 byte position calculations
- Fixed code_block_utils byte position calculation in `is_in_code_span`
- All rules now correctly handle CRLF line endings in fixes and diagnostics
- **Test Stability**: Fixed flaky tests with dependency injection pattern
- Eliminated race conditions from parallel test execution
- Tests no longer modify global environment variables
- Added `serial_test` crate for unavoidable global operations
- All 1731 tests now pass reliably in parallel execution
### Changed
- **Code Architecture**: Major refactoring to improve maintainability
- Extracted `formatter` module (397 lines) - output formatting logic
- Extracted `watch` module (491 lines) - watch mode functionality
- Extracted `file_processor` module (792 lines) - file processing logic
- Extracted `stdin_processor` module (212 lines) - stdin handling
- main.rs reduced from 3268 to 1394 lines (57% reduction)
- Improved code organization and testability
- **Line Ending Handling**: Refactored line ending preservation
- Line ending detection and normalization now at I/O boundaries
- Internal code always works with consistent LF line endings
- More efficient: 1 normalization per file instead of per-rule
- Cleaner separation of concerns
- Simplified MD047, MD012, MD022 to always use LF internally
- Removed unnecessary line ending detection from rules
- Added comprehensive end-to-end CRLF tests
### Removed
- **Legacy Fix Implementation**: Removed deprecated fix wrapper functions
- Removed `apply_fixes()` wrapper
- Removed `apply_fixes_stdin_coordinated()` wrapper
- Removed `apply_fixes_stdin()` legacy implementation
- Removed `RUMDL_NO_FIX_COORDINATOR` environment variable
- Fix Coordinator is now the only fix strategy (3 weeks stable, ~75% faster)
## [0.0.157] - 2025-10-13
### Changed
- **Removed legacy fix implementation** - Removed old single-pass fix implementation and `RUMDL_NO_FIX_COORDINATOR` environment variable. Fix Coordinator is now the only fix strategy, providing ~75%
faster fixes with better coverage.
### Added
- **MD042**: Full support for MkDocs paragraph anchors (#100)
- Recognize Python-Markdown `attr_list` extension syntax: `[](){ #anchor }`
- Support for both anchor IDs (`#id`) and CSS classes (`.class`)
- Support optional colon syntax: `[](){: #anchor }`
- UTF-8 boundary validation and DoS prevention (500 char limit)
- 28 comprehensive tests covering edge cases
- Complete documentation with links to official Python-Markdown specs
- References: [attr_list](https://python-markdown.github.io/extensions/attr_list/), [mkdocs-autorefs](https://mkdocstrings.github.io/autorefs/)
- **MD042**: Smart URL detection in empty links (#104)
- When link text looks like a URL (e.g., `[https://example.com]()`), use it as the destination
- Supports <http://>, <https://>, <ftp://>, ftps:// protocols
- More intelligent fixes than placeholder URLs
- **Always respect exclude patterns by default** (#99)
- Exclude patterns now always respected, even for explicitly provided files
- Matches behavior of ESLint, Pylint, Mypy
- Added `--no-exclude` flag to disable all exclusions when needed
- LSP support for exclude patterns
- Shows warnings with actionable hints when excluding files
- **Hidden directory scanning** (#102)
- Now scans hidden directories (like `.documentation`) by default
- More thorough markdown file discovery
### Fixed
- **MD033**: Code blocks in blockquotes false positives (#105)
- Fixed incorrect flagging of HTML tags inside fenced code blocks within blockquotes
- Properly strips blockquote markers before detecting fence markers
- 25 new tests covering nested blockquotes and edge cases
- **MD034**: Empty link construct false positives (#104)
- Fixed incorrect flagging of URLs in `[url]()` and `[url][]` patterns
- Prevents text corruption during formatting
- Added patterns to properly exclude empty link constructs
- **MD042**: Improved fix quality
- Removed "useless" placeholder fixes that just create new problems
- Only provides fixes when we have enough information for valid links
- No longer auto-fixes `[]()` or `[text]()` with placeholders
### Changed
- **BREAKING**: Exclude patterns now always respected by default
- Previously: `--force-exclude` flag needed to respect excludes for explicit files
- Now: Excludes always respected by default
- Migration: Use `--no-exclude` flag if you need the old behavior
## [0.0.156] - 2025-10-08
### Fixed
- **Build**: Removed feature-gated benchmark binaries that were causing unnecessary reinstalls
- Benchmark binaries now only built when explicitly requested
- Reduces package size and installation time
## [0.0.155] - 2025-10-08
### Fixed
- **PyPI Package**: Fixed package structure by removing unused cdylib and dependencies
- Removed unnecessary C dynamic library configuration
- Cleaner Python package distribution
## [0.0.154] - 2025-10-08
### Fixed
- **MD013**: Implemented segment-based reflow to preserve hard breaks
- Properly handles double-space line breaks
- Integration tests updated for new behavior
### Performance
- **MD034**: Reuse buffers to reduce per-line allocations
- **MD005**: Eliminate LineIndex creation overhead
- **MD030**: Eliminate O(n²) complexity by caching line collection
### Documentation
- Organized LintContext optimization documentation
## [0.0.153] - 2025-10-07
### Performance
- **Major optimization**: 54 rules now use LintContext character frequency caching
- Significant performance improvement across the board
- Reduced redundant scanning of document content
- **MD051**: Optimized link fragment validation
- Faster processing of heading anchors and fragments
### Fixed
- **MD013**: Improved nested list handling in reflow mode
- Better preservation of list structure during reformatting
## [0.0.152] - 2025-10-06
### Fixed
- **MD013**: Multi-paragraph list reflow improvements and refactoring
- Better handling of complex list structures
- More reliable paragraph detection within lists
## [0.0.151] - 2025-10-05
### Fixed
- **MD007**: Fixed tab indentation and cascade behavior
- Properly handles tabs in list indentation
- Correct cascade behavior matching markdownlint
## [0.0.150] - 2025-10-04
### Fixed
- **MD007**: Multiple fixes for list indentation
- Correct blockquote list handling
- Fixed text-aligned indentation to match markdownlint cascade behavior
- Updated test expectations for cascade behavior
## [0.0.149] - 2025-10-03
### Added
- **Configuration**: JSON Schema for rumdl.toml configuration (#89)
- IDE autocomplete and validation support
- Better configuration documentation
- **Configuration**: Per-file rule ignores (#92)
- Glob pattern support for ignoring rules on specific files
- Example: `[per-file-ignores] "docs/*.md" = ["MD013"]`
## [0.0.148] - 2025-10-02
### Fixed
- **MD042**: Display improvements
- Show exact source text in error messages
- Correct display of shorthand reference links
- **MkDocs**: Strip backticks from MkDocs auto-references (#97)
- Prevents false positives on `` [`module.Class`][] `` patterns
## [0.0.147] - 2025-10-01
### Added
- **MkDocs**: Added mkdocstrings support (#94)
- Recognizes mkdocstrings YAML options
- Multiple rules migrated to use LintContext for better MkDocs handling
### Fixed
- **MD041**: Removed auto-fix capability (#93)
- Auto-fixing front-heading violations was unreliable
- Now only reports issues without attempting fixes
- **MD026**: Corrected documentation to match implementation (#95)
- Documentation now accurately reflects punctuation handling
- **Jinja2**: Added Jinja2 template support (#96)
- Prevents false positives in template syntax
- Better support for MkDocs projects using Jinja2
- **MD013**: Prevent false positives for already-reflowed content
- Smarter detection of intentional line breaks
- **MD034**: Properly excludes URLs/emails in code spans and HTML
- No more false positives on inline code URLs
- **MD054**: Fixed column indexing bug
- Correct error position reporting
- **MD033 & MD032**: Resolved false positives (#90, #91)
- More accurate HTML tag detection
- Better handling of code blocks
## [0.0.146] - 2025-09-24
### Added
- **Fix Coordinator**: New intelligent fix system as default behavior (#88)
- ~75% faster execution on large files (15.6s vs 60.7s for OpenAPI spec)
- ~90% of issues fixed in single pass (vs 2-3 passes previously required)
- Topological sort ensures optimal rule ordering based on dependencies
- Handles cyclic dependencies gracefully
- Opt-out available via RUMDL_NO_FIX_COORDINATOR=1
- Debug output available via RUMDL_DEBUG_FIX_PERF=1
### Changed
- Fix mode now uses Fix Coordinator by default for dramatic performance gains
- Fix strategy prioritizes intelligent ordering over bulk fixes
### Performance
- First pass: 87% faster than v0.0.141, 74% faster than v0.0.143
- Completes 3 full passes (35.5s) faster than v0.0.141 does single pass (115.6s)
- Reduces LintContext creations through intelligent batching
## [0.0.145] - 2025-09-23
### Fixed
- **MD032**: Refined to handle nested code blocks correctly
- Various CI test failures and compatibility improvements
## [0.0.144] - 2025-09-22
## [0.0.142] - 2025-09-20
### Fixed
- **MD013**: Refactored to emit warning-based fixes for LSP compatibility (#79)
- MD013 reflow now works correctly when using LSP formatting in editors like Helix
- Generates proper warning-based fixes with byte ranges instead of document transforms
- Preserves trailing newlines and handles multi-line list items correctly
## [0.0.141] - 2025-09-15
### Added
- **MD013**: New normalize mode for combining short lines in paragraphs (related to #76)
- Added `reflow_mode` configuration with "default" and "normalize" options
- Normalize mode combines short lines to use the full configured line length
- Enables bulk removal of manual line breaks by setting high line_length with normalize mode
- Preserves markdown structure (lists, code blocks, tables, hard breaks)
### Fixed
- **MD013**: Fixed multi-line list item handling to avoid extra spaces when combining
- **MD012**: Fixed line number reporting for EOF blank lines
- **LSP**: Return null instead of empty array when no formatting available (related to #79)
- **MD038**: Resolved false positives and added regression tests
## [0.0.140] - 2025-09-11
### Fixed
- **LSP**: Support formatting documents without textDocument/didOpen (related to #79)
- Added lazy loading with disk fallback for unopened documents
- Editors like Helix can now format files without opening them first
- Implemented DocumentEntry structure to track document source and version
- Added intelligent caching for disk-loaded documents
- Maintains full compatibility with traditional LSP clients (VS Code)
- **MD051**: Fixed false positives in large documents with multiline inline code spans
- Multiline inline code spans were incorrectly treated as code blocks
- This caused headings after line ~600 to not be detected properly
- Removed incorrect TOC detection logic that was causing issues
- **MD032**: Fixed false positive for sequential ordered list continuations
- Sequential ordered list items (1., 2., 3.) no longer incorrectly flagged
- Added proper detection for list continuations vs separate lists
- Improved handling of lists interrupted by code blocks
- **MD052**: Fixed false positive for literal brackets in backticks
- Text like `[from ...]` in inline code no longer flagged as broken reference
- Added workaround for multiline code span detection issues
- Properly distinguishes between literal text and reference links
- **Documentation**: Added comprehensive inline configuration documentation
- Created detailed guide for rumdl-disable/enable comment syntax
- Documented all supported inline configuration formats
- Added examples for disabling rules per line, block, and file
- **Fix Counting**: Corrected issue where unfixable warnings were counted as fixed
- MD013 warnings in table cells now correctly reported as unfixable
- Fix count now reflects actual fixes applied, not total warnings
## [0.0.139] - 2025-09-09
### Fixed
- **UTF-8 Handling**: Fixed panic when processing files with multi-byte UTF-8 characters (fixes #85)
- Added proper character boundary checking in code block detection
- Prevents string slicing panics with German umlauts (ä, ö, ü) and other multi-byte characters
- Added comprehensive test coverage for various international scripts
- **MD038**: Fixed false positives caused by escaped backticks (fixes #77)
- Escaped backticks (`\``) no longer create phantom code spans
- Implemented two-pass algorithm to properly handle escaped characters
- Reduced false positive count from 179 to 0 in affected documents
- **MD032**: Improved markdownlint compatibility and edge case detection (fixes #77)
- Changed default configuration to match markdownlint (no blank lines around lists in blockquotes)
- Added detection for ordered lists starting with numbers other than 1
- Better compliance with CommonMark specification for list formatting
- **Code Block Detection**: Fixed critical bug where lines after code blocks were marked as inside them
- Corrected fenced code block end position to include the newline after closing fence
- Prevented code span detection from running inside fenced code blocks
- Fixed MD051 false positives where headings after code blocks weren't detected
- Eliminated invalid overlapping code spans that caused parsing errors
## [0.0.138] - 2025-09-05
### Fixed
- **LSP**: Fixed formatting to return empty array instead of null when no edits available (fixes #79)
- Helix editor now properly receives LSP formatting responses
- Added textDocument/rangeFormatting support for better editor compatibility
- Fixed critical position calculation bug that could cause incorrect text edit ranges
## [0.0.137] - 2025-09-04
### Fixed
- **MD051**: Fixed GitHub anchor generation for headers with arrow patterns (fixes #82)
- Headers like `WAL->L0 Compaction` now correctly generate `#wal-l0-compaction` anchors
- Arrow patterns (`->`, `-->`) now convert to the correct number of hyphens based on surrounding spaces
## [0.0.136] - 2025-09-03
## [0.0.135] - 2025-09-03
## [0.0.134] - 2025-09-02
### Added
- **MD051**: HTML anchor tag support for any element with id/name attributes
- Supports `<a>`, `<span>`, `<div>` and any other HTML element with id attribute
- Case-sensitive matching for HTML anchors (case-insensitive for Markdown)
- Handles multiple id attributes (only first is used per HTML spec)
### Fixed
- **MD007**: Implemented proper indentation style configuration for markdownlint compatibility
- Added IndentStyle enum with TextAligned (default) and Fixed (markdownlint) modes
- Auto-configures style="fixed" when loading from .markdownlint.yaml files
- Resolves 5-space indentation detection issues (#77)
- **MD029**: Improved list numbering style compatibility
- Added OneOrOrdered style (markdownlint default) accepting either all-ones or sequential
- Changed default from Ordered to OneOrOrdered for better compatibility
- **MD050**: Fixed false positives for emphasis patterns inside HTML `<code>` tags
- Patterns like `__pycache__`, `__init__` no longer flagged inside code elements
- **MD052**: Fixed reference checking to skip HTML content lines
- Skip any line starting with '<' to match markdownlint behavior
- Fixed regex to properly handle nested brackets in references like `[`Union[T, None]`]`
- **MD053**: Improved duplicate reference detection
- Detect when same reference is defined multiple times
- Handle case-insensitive duplicates per CommonMark spec
- Remove overly aggressive filters that skip valid references
- **MD028**: Aligned with markdownlint behavior for blank lines in blockquotes
- Flag all blank lines between blockquotes as ambiguous
- Better distinguish blockquote separators from internal blank lines
- **MD005**: Fixed respect for MD007 configuration and nested list handling
- **MD006**: Skip validation for lists inside blockquotes where indentation is expected
## [0.0.133] - 2025-08-30
### Fixed
- **MD028/MD009**: Complete fix for rule conflict where MD028 and MD009 were "fighting each other" (fixes #66)
- MD028 now only flags truly blank lines inside blockquotes, not `>` or `>` lines
- MD009 simplified to remove special cases for empty blockquote lines
- Both rules now correctly follow CommonMark specifications
## [0.0.132] - 2025-08-30
### Added
- **LSP**: Added "Fix all rumdl issues" code action for bulk fixes when multiple fixable diagnostics are present
## [0.0.131] - 2025-08-28
### Fixed
- **MD002**: Implemented markdownlint compatibility - MD002 no longer triggers when first heading is on the first line, regardless of level (fixes #65)
- **MD034**: Added support for multi-line MkDocs snippet blocks where markers appear on separate lines (fixes #70)
## [0.0.130] - 2025-08-27
### Fixed
- **MD052**: Fixed false positives with IPv6 URLs containing brackets (e.g., `http://[::1]:8080/path[0]`)
- **MD053**: Made rule warning-only, removed automatic fixes to prevent accidental removal of intentionally kept references (fixes #69)
- **MD009/MD028**: Resolved formatting loop between trailing spaces and blank blockquote lines
- **MD002/MD041**: Fixed interaction where MD002 incorrectly flagged documents starting with level-1 heading
- **MD011**: Prevented false positives with math-like expressions (e.g., `[0,1]`) outside code blocks
- **MD034**: Improved bare URL detection to avoid false positives with bracketed paths in URLs
## [0.0.129] - 2025-08-26
### Added
- **MkDocs Extended Support**: Enhanced MkDocs compatibility with PyMdown Extensions
- Snippets syntax (`--8<--`) support (fixes #62)
- Admonitions (`!!!`, `???`, `???+`) for collapsible note blocks
- Tabs (`=== "Tab Name"`) for content organization
- Footnotes (`[^ref]`) for reference-style citations
- Cross-references and auto-doc blocks
### Fixed
- **MkDocs Validation**: Made validation more lenient to detect malformed syntax
- **Configuration Migration**: Fixed migration of multiple disabled rules from markdownlint config
## [0.0.128] - 2025-08-25
### Fixed
- **MD042/MD052**: Added support for simple identifiers in MkDocs auto-references
## [0.0.127] - 2025-08-25
### Added
- **MkDocs Support**: Added MkDocs markdown flavor (closes #63)
- New `flavor = "mkdocs"` configuration option
- MkDocs auto-references (e.g., `[class.Name][]`, `[module.function][]`) are no longer flagged as errors
- MD042 and MD052 rules now recognize MkDocs-specific patterns
- Type-safe enum implementation for better extensibility
### Fixed
- **MD005**: Dynamic indent detection to respect user's chosen pattern (fixes #64)
- Analyzes existing document to detect 2-space vs 4-space indentation
- Preserves user's indentation style instead of forcing a default
### Changed
- **Configuration**: Renamed MD013 'enable_reflow' to 'reflow' with backwards compatibility
## [0.0.126] - 2025-08-23
### Fixed
- **Build**: Fixed output filename collision warning during `cargo install` (#61)
## [0.0.125] - 2025-08-22
### Added
- **CLI**: Added `--stdin-filename` flag for better stdin processing
- Specify filename when reading from stdin for better error messages
- Enables MD057 (relative link checking) to work correctly with stdin
- Provides proper filename context in all output formats
- Improves editor integration capabilities
### Fixed
- **CLI**: Fixed `rumdl fmt -` to output original content when no issues found
- Previously incorrectly output "No issues found in stdin" message
- Now correctly outputs the original content unchanged
- **MD029**: Corrected list continuity detection and fix functionality
- Improved handling of sublists and indented content
- Better markdownlint compatibility
### Changed
- **Build**: Added mise version validation to pre-release checks
- Prevents CI failures from non-existent mise versions
## [0.0.124] - 2025-08-22
### Added
- **Formatting**: Added stdin/stdout formatting support (closes #59)
- `rumdl fmt` command for formatting markdown files (alias for `check --fix`)
- `--stdin` with `--fix` now outputs formatted content to stdout
- Support for `-` as stdin indicator (Unix convention: `rumdl fmt -`)
- Clear separation between linting (diagnostics to stderr) and formatting (content to stdout)
- Documentation updated with formatting examples
### Fixed
- **MD052**: Don't flag GitHub alerts as undefined references (closes #60)
- GitHub alert syntax (`[!NOTE]`, `[!TIP]`, `[!WARNING]`, `[!IMPORTANT]`, `[!CAUTION]`) no longer flagged
- Improved compatibility with GitHub-flavored markdown
- **MD009**: Fixed heading trailing space removal
- Headings now have ALL trailing spaces removed (they serve no purpose in headings)
- **CLI**: Fixed stdin diagnostics output in check mode
- Diagnostics now correctly output to stderr by default in check mode without `--fix`
## [0.0.123] - 2025-08-21
### Added
- **MD013**: Comprehensive markdown pattern preservation during text reflow
- Preserves reference links, footnotes, math formulas, wiki links, and more
- Centralized regex patterns for better maintainability
### Fixed
- **CLI**: Correct unfixable rules status display and fix counts (closes #56)
- Rules marked as unfixable now show `[unfixable]` in yellow instead of `[fixed]`
- Fix count now correctly excludes unfixable rules (e.g., "3 of 6" instead of "6 of 6")
- Added FixCapability enum to Rule trait for compile-time safety
- **MD013**: Preserve reference links during text reflow
- Reference-style links are now properly preserved when reflowing text
- Fixed indicator display to correctly show `[fixed]` when issues are resolved
- **Tests**: Mark kramdown definition list doctest as text to fix test failures
### Changed
- **Internal**: Centralized markdown pattern regexes and extended reflow support
- Improved code organization and reduced duplication
- Better performance through shared regex compilation
## [0.0.122] - 2025-08-19
### Added
- **Configuration Discovery**: Automatic upward directory traversal to find configuration files (closes #58)
- Searches parent directories for `.rumdl.toml`, `rumdl.toml`, or `pyproject.toml`
- Similar behavior to `git`, `ruff`, and `eslint`
- Stops at `.git` directory boundaries
- **--isolated flag**: New flag to disable all configuration discovery (Ruff-compatible)
- Alias for `--no-config` for better ecosystem compatibility
## [0.0.121] - 2025-08-19
### Fixed
- **MD051**: Resolved remaining Issue #39 edge cases for link fragment validation
- Fixed ampersand handling at boundaries: "& text" → "--text", "text &" → "text-"
- Fixed cross-file link detection to properly ignore absolute paths (e.g., `/tags#anchor`)
- Improved Liquid template handling to skip links with filters (e.g., `{{ url | relative_url }}`)
- Fixed test expectations to match actual GitHub behavior for multiple spaces and trailing punctuation
- Verified Jekyll/kramdown GFM underscore handling works correctly for technical identifiers
### Improved
- **MD051**: Enhanced anchor generation accuracy and security
- Added comprehensive security hardening (Unicode normalization, RTL/LTR override prevention)
- Improved emoji detection and boundary handling
- Better performance with optimized regex patterns and early exit checks
- Added regression tests for all Issue #39 scenarios
## [0.0.120] - 2025-08-16
### Performance
- Incremental improvements to various rule implementations
## [0.0.119] - 2025-08-15
### Fixed
- **MD051**: Fixed GitHub anchor generation algorithm to correctly handle consecutive spaces
- "Test & Heading!" now correctly generates "test--heading" instead of "test-heading"
- Improved compliance with GitHub's official anchor generation behavior
- Fixed whitespace normalization bug that was collapsing multiple spaces to single spaces
### Improved
- **Code Quality**: Removed all `#[allow(dead_code)]` violations in codebase
- Removed unused `InternalCodeBlockState` enum from document_structure.rs
- Removed unused `extract_url_from_link` function from md057_existing_relative_links.rs
- Removed unused `is_in_code_block` function from md007_ul_indent.rs
## [0.0.118] - 2025-08-14
### Performance
- Incremental improvements to various rule implementations
## [0.0.117] - 2025-08-14
### Fixed
- MD037: Fixed false positive with asterisks in inline code spans (issue #49)
- Inline code content is now properly masked before emphasis detection
- MD011: Fixed false positive with array access patterns in link titles (issue #50)
- Context detection now properly skips patterns inside code spans
- MD052: Fixed false positive with square brackets in HTML attributes (issue #51)
- HTML tag detection prevents reference checking within HTML elements
- Added centralized skip context detection for improved accuracy across rules
## [0.0.116] - 2025-08-13
### Added
- Kramdown-style custom header IDs support (#44)
- Headers can now have custom IDs using the `{#custom-id}` syntax
- Custom IDs are preserved when fixing MD051 (link fragments)
- MD026 (trailing punctuation) now ignores headers with custom IDs
- Safe character validation: accepts Unicode letters/numbers, hyphens, underscores, and colons
- Rejects problematic characters like spaces, quotes, brackets, and HTML/CSS special chars
### Fixed
- Pre-release script now correctly handles dynamic versioning in pyproject.toml
- Added Cargo.lock validation and `cargo publish --dry-run` checks to prevent release failures
## [0.0.115] - 2025-08-12
### Fixed
- Various bug fixes and improvements
## [0.0.114] - 2025-08-09
### Fixed
- Various bug fixes and improvements
## [0.0.113] - 2025-08-09
### Fixed
- Various bug fixes and improvements
## [0.0.112] - 2025-08-08
### Fixed
- Various bug fixes and improvements
## [0.0.110] - 2025-08-08
### Changed
- Various bug fixes and improvements
## [0.0.107] - 2025-08-06
### Fixed
- MD036: Remove automatic fix to prevent document corruption when bold/italic text is used as image captions, labels, or warnings (#23)
- MD011: No longer flags patterns like `()[1]` inside inline code as reversed links (#19)
- MD052: No longer flags reference patterns inside HTML comments as undefined references (#20)
## [0.0.106] - 2025-08-05
### Changed
- Moved benchmark binaries from Python package distribution
- Benchmark tools are now in `benchmark/bin/` directory
- Added `build-benchmarks` feature flag to explicitly build benchmarks
- Python package now only includes the main `rumdl` binary
- Significantly reduced installed package size
## [0.0.105] - 2025-08-05
### Fixed
- MD029: Fixed list continuation detection to properly handle variable marker widths (fixes #16)
- List items with double-digit markers (e.g., "10. ") now correctly require 4+ spaces for continuation
- List items with triple-digit markers (e.g., "100. ") now correctly require 5+ spaces for continuation
- List items with any number of digits now correctly calculate required continuation indentation
- MD027: Improved tab handling and added bounds checking for range calculations
- Installation: Improved update experience for Cursor/Windsurf editors
### Added
- `--update` flag to check for newer versions and update if available
- Version checking with update notifications
- Marketplace-aware installation for VS Code forks
- Comprehensive tests for MD029 with large number markers (triple and quadruple digits)
### Changed
- Clarified `--force` flag behavior in help text
## [0.0.104] - 2025-08-02
### Added
- File-wide inline configuration support with `disable-file`, `enable-file`, and `configure-file` comments
- Support for JSON configuration within inline comments to customize rule behavior per file
- Enhanced inline configuration to handle edge cases with multiple comments on the same line
- Support for enabling specific rules when all rules are disabled
### Fixed
- Process inline configuration comments in order of appearance on the same line
- Skip processing inline configuration comments inside code blocks
## [0.0.102] - 2025-07-24
## [0.0.101] - 2025-07-23
## [0.0.100] - 2025-07-22
### Performance Improvements
- **MD032**: Eliminated redundant DocumentStructure creation through optimization interface delegation
- Refactored check() method to delegate to check_with_structure() for shared parsing
- Added fix_with_structure() helper method for optimized fixing operations
- **List Processing**: Major refactoring of complex list block merging logic for better maintainability
- Extracted merge_adjacent_list_blocks into clean ListBlockMerger struct
- Introduced BlockSpacing enum for clear categorization of list spacing types
- Separated compatibility checking, spacing analysis, and merging logic into focused methods
- **Memory Management**: Added comprehensive performance stress tests for deeply nested lists
- Created benchmarks for up to 20 levels of nesting with measurable performance baselines
- Established performance thresholds: <3ms parsing, <4ms rule checking for extreme nesting
- Added memory stress testing to prevent performance regressions
### Code Quality
- Improved separation of concerns in list processing logic
- Enhanced code maintainability through better structured algorithms
- Added comprehensive test coverage for pathological markdown structures
## [0.0.99] - 2025-07-22
### Fixed
- MD034: Added support for `ftps://` URLs
- MD034: Fixed detection of URLs in HTML comments (now properly ignored)
- MD039: Fixed escaped character handling in link text
- MD044: Fixed clippy warnings and improved pattern matching for proper names
- MD052: Enhanced nested bracket handling in reference links and images
- Fixed flaky performance tests by increasing timeout threshold for CI environments
- Improved test stability for Unicode list indentation tests
## [0.0.98] - 2025-07-18
### Added
- Homebrew tap support for easy installation on macOS and Linux
- Created `homebrew-rumdl` tap repository
- Added automatic archive creation for macOS builds in release workflow
- Included SHA256 checksum generation for each platform
- Set up automated formula updates on new releases
- Homebrew installation instructions in README
### Changed
- Enhanced release workflow to create platform-specific tar.gz archives
- Added repository dispatch to notify homebrew-rumdl on new releases
## [0.0.97] - 2025-07-17
### Changed
- Updated exit code handling for consistency:
- Configuration errors now return exit code 2 (was 1)
- File not found errors now return exit code 2 (was 1)
- Invalid command arguments now return exit code 2 (was 1)
- Standardized exit codes across all error conditions:
- Exit code 1: Reserved for linting issues found
- Exit code 2: Tool errors (config parse errors, file not found, invalid arguments)
### Fixed
- Improved consistency in exit code handling across the entire codebase
- Updated all tests to expect correct exit codes for different error scenarios
## [0.0.96] - 2025-07-16
### Added
- MD013: Text reflow/wrapping functionality for automatic line breaking (fixes #13)
- New `enable_reflow` configuration option (disabled by default)
- Intelligently wraps long lines while preserving Markdown formatting
- Preserves bold, italic, links, code spans, and other Markdown elements
- Proper list continuation indentation that aligns with the text after markers
- Preserves hard line breaks (two trailing spaces)
- Does not wrap code blocks, tables, headings, or reference definitions
- Added `pulldown-cmark` dependency (v0.12.2) for improved Markdown parsing
- Comprehensive test coverage for text reflow functionality
### Changed
- MD013: Enhanced fix functionality with optional text reflow (opt-in feature)
### Fixed
- MD013: Fixed list indentation to properly align continuation lines with the text content
## [0.0.95] - 2025-07-15
### Added
- Implemented 3-tier exit code system following Ruff's convention:
- Exit code 0: Success (no issues found)
- Exit code 1: Linting violations found
- Exit code 2: Tool error (config error, file access error, etc.)
- Added exit_codes module for cleaner exit code management
### Changed
- Updated all error handlers to use appropriate exit codes
- CI/CD systems can now distinguish between markdown issues (exit 1) and tool failures (exit 2)
### Documentation
- Updated README with exit code documentation
- Added exit codes section to CLI reference
## [0.0.94] - 2025-07-04
### Performance Improvements
- Implemented lazy code span loading - 3.8x speedup for 94% of rules that don't use code spans
- MD013: 34.5% faster check operations through aggressive early returns
- MD038: 14% faster by leveraging lazy code span loading
- MD044: 93.5% faster with global regex caching
- MD047: 8.3% faster using pre-computed line data
- MD053: 39.7% faster by leveraging pre-parsed reference definitions
- Overall LintContext creation improved by 11.7%
### Fixed
- MD053: Fixed escaped character handling in reference definitions
## [0.0.93] - 2025-07-03
## [0.0.92] - 2025-07-02
### Fixed
- MD036: Align with markdownlint behavior - emphasis ending with punctuation (e.g., `**Note:**`) is no longer flagged
## [0.0.91] - 2025-07-02
## [0.0.90] - 2025-07-01
## [0.0.89] - 2025-07-01
### Added
- Comprehensive unit test coverage for all 54 linting rules (~742 new tests)
- Unit tests for LSP server functionality
- Unit tests for all 11 output formatters
- Unit tests for Python bindings
- Test coverage improved from 75.4% to 77.1%
### Fixed
- MD005: Fixed blockquote handling to correctly ignore intentional separations
- MD054: Fixed overlapping match detection for link/image styles
- Strong style utility module refactored to remove unused code
- MD029: Fixed nested code block detection by implementing proper CommonMark fence closing rules
- MD052: Fixed false positives for arrays and references inside inline code spans
- MD013: Fixed line length calculation to intelligently handle URLs in non-strict mode
### Improved
- Test infrastructure now includes both unit and integration tests
- Better test organization with inline unit tests in implementation files
## [0.0.88] - 2025-06-28
### Added
- 11 new output formatters for enhanced compatibility:
- `grouped` - Groups violations by file
- `pylint` - Pylint-compatible format
- `azure` - Azure Pipeline logging format
- `concise` - Minimal file:line:col format
- `github` - GitHub Actions annotation format
- `gitlab` - GitLab Code Quality report format
- `json` - Machine-readable JSON format
- `json_lines` - JSONL format (one JSON object per line)
- `junit` - JUnit XML format for CI integration
- `sarif` - SARIF format for security tools
- `text` - Default human-readable format with colors
### Changed
- **BREAKING**: Upgraded to Rust 2024 edition (requires Rust 1.87.0+)
- Improved code quality by fixing all 283 clippy warnings
### Fixed
- Config `output_format` field now properly merges from configuration files
- Pylint formatter now outputs correct `CMD` codes instead of generic `C` codes
### Optimized
- Removed unused dependencies (`glob`, `walkdir`)
- Reduced binary size with aggressive compilation flags (LTO, strip, opt-level=z)
- Improved performance through better regex compilation and caching
## [0.0.87] - 2025-06-16
## [0.0.86] - 2025-06-14
## [0.0.85] - 2025-06-11
## [0.0.84] - 2025-06-10
### Added
- Type-safe serde-based configuration system for all 24 configurable rules
- Dedicated config modules for each rule with compile-time validation
- Full IDE support with autocomplete for configuration options
- Centralized utilities for common parsing patterns
### Changed
- **BREAKING**: Internal configuration structure refactored (external API unchanged)
- Migrated all rules from manual TOML parsing to serde deserialization
- Improved performance through centralized parsing for:
- Link and URL detection
- Code span identification
- List item processing
- Block element detection
- Pre-computed line information for better performance
- ~40% reduction in configuration boilerplate code
### Fixed
- MD030: Correct handling of tab characters in list items
### Performance
- Significant performance improvements across multiple rules through:
- Centralized regex compilation and caching
- Reduced redundant parsing operations
- More efficient text processing algorithms
- Optimized pattern matching for MD044
## [0.0.83] - 2025-06-07
### Fixed
- Various bug fixes and performance improvements
## [0.0.82] - 2025-06-06
### Fixed
- Various bug fixes and stability improvements
## [0.0.81] - 2025-06-04
### Added
- Initial implementation of remaining rules for markdownlint parity
[Unreleased]: https://github.com/rvben/rumdl/compare/v0.0.200...HEAD
[0.0.200]: https://github.com/rvben/rumdl/compare/v0.0.199...v0.0.200
[0.0.163]: https://github.com/rvben/rumdl/compare/v0.0.162...v0.0.163
[0.0.162]: https://github.com/rvben/rumdl/compare/v0.0.161...v0.0.162
[0.0.161]: https://github.com/rvben/rumdl/compare/v0.0.160...v0.0.161
[0.0.160]: https://github.com/rvben/rumdl/compare/v0.0.159...v0.0.160
[0.0.159]: https://github.com/rvben/rumdl/compare/v0.0.158...v0.0.159
[0.0.158]: https://github.com/rvben/rumdl/compare/v0.0.157...v0.0.158
[0.0.157]: https://github.com/rvben/rumdl/compare/v0.0.156...v0.0.157
[0.0.156]: https://github.com/rvben/rumdl/compare/v0.0.155...v0.0.156
[0.0.155]: https://github.com/rvben/rumdl/compare/v0.0.154...v0.0.155
[0.0.154]: https://github.com/rvben/rumdl/compare/v0.0.153...v0.0.154
[0.0.153]: https://github.com/rvben/rumdl/compare/v0.0.152...v0.0.153
[0.0.152]: https://github.com/rvben/rumdl/compare/v0.0.151...v0.0.152
[0.0.151]: https://github.com/rvben/rumdl/compare/v0.0.150...v0.0.151
[0.0.150]: https://github.com/rvben/rumdl/compare/v0.0.149...v0.0.150
[0.0.149]: https://github.com/rvben/rumdl/compare/v0.0.148...v0.0.149
[0.0.148]: https://github.com/rvben/rumdl/compare/v0.0.147...v0.0.148
[0.0.147]: https://github.com/rvben/rumdl/compare/v0.0.146...v0.0.147
[0.0.146]: https://github.com/rvben/rumdl/compare/v0.0.145...v0.0.146
[0.0.145]: https://github.com/rvben/rumdl/compare/v0.0.144...v0.0.145
[0.0.144]: https://github.com/rvben/rumdl/compare/v0.0.143...v0.0.144
[0.0.142]: https://github.com/rvben/rumdl/compare/v0.0.141...v0.0.142
[0.0.140]: https://github.com/rvben/rumdl/compare/v0.0.139...v0.0.140
[0.0.138]: https://github.com/rvben/rumdl/compare/v0.0.137...v0.0.138
[0.0.137]: https://github.com/rvben/rumdl/compare/v0.0.136...v0.0.137
[0.0.136]: https://github.com/rvben/rumdl/compare/v0.0.135...v0.0.136
[0.0.135]: https://github.com/rvben/rumdl/compare/v0.0.134...v0.0.135
[0.0.134]: https://github.com/rvben/rumdl/compare/v0.0.133...v0.0.134
[0.0.133]: https://github.com/rvben/rumdl/compare/v0.0.132...v0.0.133
[0.0.132]: https://github.com/rvben/rumdl/compare/v0.0.131...v0.0.132
[0.0.131]: https://github.com/rvben/rumdl/compare/v0.0.130...v0.0.131
[0.0.130]: https://github.com/rvben/rumdl/compare/v0.0.129...v0.0.130
[0.0.129]: https://github.com/rvben/rumdl/compare/v0.0.128...v0.0.129
[0.0.128]: https://github.com/rvben/rumdl/compare/v0.0.127...v0.0.128
[0.0.127]: https://github.com/rvben/rumdl/compare/v0.0.126...v0.0.127
[0.0.126]: https://github.com/rvben/rumdl/compare/v0.0.125...v0.0.126
[0.0.125]: https://github.com/rvben/rumdl/compare/v0.0.124...v0.0.125
[0.0.124]: https://github.com/rvben/rumdl/compare/v0.0.123...v0.0.124
[0.0.123]: https://github.com/rvben/rumdl/compare/v0.0.122...v0.0.123
[0.0.122]: https://github.com/rvben/rumdl/compare/v0.0.121...v0.0.122
[0.0.121]: https://github.com/rvben/rumdl/compare/v0.0.120...v0.0.121
[0.0.120]: https://github.com/rvben/rumdl/compare/v0.0.119...v0.0.120
[0.0.119]: https://github.com/rvben/rumdl/compare/v0.0.118...v0.0.119
[0.0.118]: https://github.com/rvben/rumdl/compare/v0.0.117...v0.0.118
[0.0.117]: https://github.com/rvben/rumdl/compare/v0.0.116...v0.0.117
[0.0.116]: https://github.com/rvben/rumdl/compare/v0.0.115...v0.0.116
[0.0.115]: https://github.com/rvben/rumdl/compare/v0.0.114...v0.0.115
[0.0.114]: https://github.com/rvben/rumdl/compare/v0.0.113...v0.0.114
[0.0.113]: https://github.com/rvben/rumdl/compare/v0.0.112...v0.0.113
[0.0.112]: https://github.com/rvben/rumdl/compare/v0.0.111...v0.0.112
[0.0.110]: https://github.com/rvben/rumdl/compare/v0.0.109...v0.0.110
[0.0.107]: https://github.com/rvben/rumdl/compare/v0.0.106...v0.0.107
[0.0.106]: https://github.com/rvben/rumdl/compare/v0.0.105...v0.0.106
[0.0.105]: https://github.com/rvben/rumdl/compare/v0.0.104...v0.0.105
[0.0.104]: https://github.com/rvben/rumdl/compare/v0.0.103...v0.0.104
[0.0.102]: https://github.com/rvben/rumdl/compare/v0.0.101...v0.0.102
[0.0.101]: https://github.com/rvben/rumdl/compare/v0.0.100...v0.0.101
[0.0.100]: https://github.com/rvben/rumdl/compare/v0.0.99...v0.0.100
[0.0.99]: https://github.com/rvben/rumdl/compare/v0.0.98...v0.0.99
[0.0.98]: https://github.com/rvben/rumdl/compare/v0.0.97...v0.0.98
[0.0.97]: https://github.com/rvben/rumdl/compare/v0.0.96...v0.0.97
[0.0.96]: https://github.com/rvben/rumdl/compare/v0.0.95...v0.0.96
[0.0.95]: https://github.com/rvben/rumdl/compare/v0.0.94...v0.0.95
[0.0.94]: https://github.com/rvben/rumdl/compare/v0.0.93...v0.0.94
[0.0.93]: https://github.com/rvben/rumdl/compare/v0.0.92...v0.0.93
[0.0.92]: https://github.com/rvben/rumdl/compare/v0.0.91...v0.0.92
[0.0.91]: https://github.com/rvben/rumdl/compare/v0.0.90...v0.0.91
[0.0.90]: https://github.com/rvben/rumdl/compare/v0.0.89...v0.0.90
[0.0.89]: https://github.com/rvben/rumdl/compare/v0.0.88...v0.0.89
[0.0.88]: https://github.com/rvben/rumdl/compare/v0.0.87...v0.0.88
[0.0.87]: https://github.com/rvben/rumdl/compare/v0.0.86...v0.0.87
[0.0.86]: https://github.com/rvben/rumdl/compare/v0.0.85...v0.0.86
[0.0.85]: https://github.com/rvben/rumdl/compare/v0.0.84...v0.0.85
[0.0.84]: https://github.com/rvben/rumdl/compare/v0.0.83...v0.0.84
[0.0.83]: https://github.com/rvben/rumdl/compare/v0.0.82...v0.0.83
[0.0.82]: https://github.com/rvben/rumdl/compare/v0.0.81...v0.0.82
[0.0.81]: https://github.com/rvben/rumdl/releases/tag/v0.0.81