@markuplint/ml-core 5.0.0-rc.4 → 5.0.0-rc.6

package/CHANGELOG.md

@@ -3,6 +3,62 @@
 All notable changes to this project will be documented in this file.
 See [Conventional Commits](https://conventionalcommits.org) for commit guidelines.
 
+# [5.0.0-rc.6](https://github.com/markuplint/markuplint/compare/v5.0.0-rc.5...v5.0.0-rc.6) (2026-08-30)
+
+### Features
+
+- split rule-deprecation notices out of config-error ([#4013](https://github.com/markuplint/markuplint/issues/4013)) ([812e6f3](https://github.com/markuplint/markuplint/commit/812e6f356839af8f257cfd91e6b16cfdfdd7cf33))
+
+### BREAKING CHANGES
+
+- violations for deprecated rule names now have
+  `ruleId: 'rule-deprecation'` instead of `ruleId: 'config-error'`. Any
+  consumer filtering `MLCore.verify()` output (or the markuplint CLI/API) by
+  `ruleId === 'config-error'` to catch deprecation messages must also check
+  for `rule-deprecation`.
+
+- feat(markuplint): add --severity-deprecation CLI flag
+
+Wires the new severity.deprecation config option (@markuplint/ml-config)
+and the rule-deprecation ruleId (@markuplint/ml-core) through the CLI:
+
+- --severity-deprecation flag, mirroring --severity-parse-error
+- --show-config details now also surfaces ruleDeprecations
+- per-run dedupe and failed-file counting generalized to cover both
+  config-level ruleIds (config-error and rule-deprecation), not just
+  config-error
+
+* docs(website): document severity.deprecation (EN + JA)
+
+# [5.0.0-rc.5](https://github.com/markuplint/markuplint/compare/v5.0.0-rc.4...v5.0.0-rc.5) (2026-08-28)
+
+### Bug Fixes
+
+- **ml-core:** look up dedupe contract under both alias and base rule names ([a5dde7e](https://github.com/markuplint/markuplint/commit/a5dde7e23a12a43d976fbfaa1cc3dec67265447e)), closes [#3844](https://github.com/markuplint/markuplint/issues/3844) [#3871](https://github.com/markuplint/markuplint/issues/3871)
+- **ml-core:** re-verify fixed code and detect N-pass fix cycles ([8c96516](https://github.com/markuplint/markuplint/commit/8c96516d8057acdcf06dc279c2183390f0966e98)), closes [#3890](https://github.com/markuplint/markuplint/issues/3890) [#3891](https://github.com/markuplint/markuplint/issues/3891)
+- **ml-core:** stop accumulating duplicate mapping errors across setCode ([2874d5b](https://github.com/markuplint/markuplint/commit/2874d5b60b24ccc501f7dae1f6ab25fb6e51e58b)), closes [#3900](https://github.com/markuplint/markuplint/issues/3900)
+- **pretenders:** resolve same-named components via imports, not scan order ([#3957](https://github.com/markuplint/markuplint/issues/3957)) ([d46a514](https://github.com/markuplint/markuplint/commit/d46a5148c4d7afb156962f4ed795f40a9324e6c5)), closes [#3951](https://github.com/markuplint/markuplint/issues/3951) [#3951](https://github.com/markuplint/markuplint/issues/3951) [#3951](https://github.com/markuplint/markuplint/issues/3951)
+- **rules:** surface disallowed-element reason via reasonOnly (close [#3815](https://github.com/markuplint/markuplint/issues/3815)) ([#3986](https://github.com/markuplint/markuplint/issues/3986)) ([0142cec](https://github.com/markuplint/markuplint/commit/0142cec667f70fee086f2a6e06d7a26e66bda380))
+
+### Code Refactoring
+
+- **rules:** redesign v5 rule system — naming, splits, specConformance ([#3989](https://github.com/markuplint/markuplint/issues/3989)) ([e925565](https://github.com/markuplint/markuplint/commit/e925565ce537848d7d1573369723cbce724a841b)), closes [#4](https://github.com/markuplint/markuplint/issues/4) [#aside-conditional-role-mapping-aria-13](https://github.com/markuplint/markuplint/issues/aside-conditional-role-mapping-aria-13)
+
+- fix(ml-core)!: reject pretender on standard HTML elements ([0576425](https://github.com/markuplint/markuplint/commit/0576425baf9b78141523bcd0e4a102062b96bc81)), closes [#3740](https://github.com/markuplint/markuplint/issues/3740)
+
+### Features
+
+- **ml-core:** hook-based dedupe for parse-error channel via mirrorsParseErrorCodes ([6a36f17](https://github.com/markuplint/markuplint/commit/6a36f17d292a2400494a3670c93cb2f02999b48b)), closes [#3844](https://github.com/markuplint/markuplint/issues/3844)
+- **ml-core:** surface non-fatal parser errors via opt-in parse-error channel ([d715fbe](https://github.com/markuplint/markuplint/commit/d715fbe586c478b98472a570b3175873cf244aef)), closes [#3844](https://github.com/markuplint/markuplint/issues/3844)
+
+### BREAKING CHANGES
+
+- **rules:** with no alias coverage.
+- `pretenders` config entries whose selector matches a
+  standard HTML element are now ignored. Configurations that previously
+  relied on `<marquee as="div">` to suppress markuplint violations must
+  remove the entry; the original element is now linted on its own merits.
+
 # [5.0.0-rc.4](https://github.com/markuplint/markuplint/compare/v5.0.0-rc.3...v5.0.0-rc.4) (2026-04-19)
 
 **Note:** Version bump only for package @markuplint/ml-core

package/README.md

@@ -5,11 +5,6 @@
 ## Documentation
 
 - [API Document](https://markuplint.dev/api-docs)
-- [Architecture](ARCHITECTURE.md) ([日本語](ARCHITECTURE.ja.md)) — Package overview, MLDOM hierarchy, rule system, and linting pipeline
-- [MLDOM Reference](docs/ml-dom.md) ([日本語](docs/ml-dom.ja.md)) — Class hierarchy, node properties, tree traversal
-- [Rule System](docs/rule-system.md) ([日本語](docs/rule-system.ja.md)) — MLRule, RuleSeed, MLRuleContext, configuration resolution
-- [Linting Pipeline](docs/linting-pipeline.md) ([日本語](docs/linting-pipeline.ja.md)) — MLCore engine, verify flow, pretender, plugin system
-- [Maintenance Guide](docs/maintenance.md) ([日本語](docs/maintenance.ja.md)) — Commands, recipes, and troubleshooting
 
 ## Install
 

package/lib/cursor-offset.js

@@ -15,16 +15,13 @@ export function computeCursorOffset(appliedEdits, cursorOffset) {
         const [start, end] = edit.range;
         const delta = edit.text.length - (end - start);
         if (start > cursorOffset) {
-            // Edit is after cursor — no effect
             break;
         }
         if (end <= cursorOffset) {
-            // Edit is entirely before cursor — shift by delta.
             // Range is half-open [start, end), so cursor at `end` is outside the edit.
             newOffset += delta;
         }
         else {
-            // Cursor falls inside the replaced range [start, end)
             newOffset = start + edit.text.length;
             break;
         }

package/lib/fix-applier.js

@@ -19,20 +19,18 @@ export function applyFixes(sourceCode, fixes) {
     if (fixes.length === 0) {
         return { output: sourceCode, applied: [], skipped: [], appliedEdits: [] };
     }
-    // Tag each edit with its parent FixData index
     const taggedEdits = [];
     for (const [i, fix] of fixes.entries()) {
         for (const edit of fix.edits) {
             taggedEdits.push({ edit, fixIndex: i });
         }
     }
-    // Sort: range[0] ascending, then range[1] descending (so larger ranges come first at the same start)
+    // Ties broken by range[1] descending so larger ranges come first at the same start.
     taggedEdits.sort((a, b) => {
         const startDiff = a.edit.range[0] - b.edit.range[0];
         // eslint-disable-next-line @typescript-eslint/strict-boolean-expressions
         return startDiff || b.edit.range[1] - a.edit.range[1];
     });
-    // Track which FixData indices had at least one skipped edit
     const skippedFixIndices = new Set();
     const appliedEdits = [];
     let lastAppliedEnd = -1;
@@ -40,22 +38,18 @@ export function applyFixes(sourceCode, fixes) {
     let cursor = 0;
     for (const { edit, fixIndex } of taggedEdits) {
         const [start, end] = edit.range;
-        // Overlap check: if this edit starts before the end of the last applied edit, skip it.
-        // Also skip if a sibling edit from the same FixData was already skipped —
-        // edits within a single FixData are atomic (all-or-nothing).
+        // Edits within a single FixData are atomic (all-or-nothing): if any sibling
+        // was skipped, skip the rest.
         if (start < lastAppliedEnd || skippedFixIndices.has(fixIndex)) {
             skippedFixIndices.add(fixIndex);
             continue;
         }
-        // Append the source text between the last edit and this one
         parts.push(sourceCode.slice(cursor, start), edit.text);
         appliedEdits.push(edit);
         cursor = end;
         lastAppliedEnd = end;
     }
-    // Append remaining source text
     parts.push(sourceCode.slice(cursor));
-    // Classify FixData as applied or skipped
     const applied = [];
     const skipped = [];
     for (const [i, fix] of fixes.entries()) {

package/lib/ml-core.d.ts

@@ -2,6 +2,20 @@ import type { MLFabric } from './types.js';
 import type { PlainData, RuleConfigValue, TextEdit, Violation } from '@markuplint/ml-config';
 import { ParserError } from '@markuplint/parser-utils';
 import { Document } from './ml-dom/index.js';
+/**
+ * `ruleId` of a genuinely broken config (unresolved rule reference, plugin
+ * resolution failure, ...). Exported so consumers that need to recognize
+ * config-level violations (e.g. the CLI's per-run dedupe and failed-file
+ * counting in `packages/markuplint/src/cli/command.ts`) reference the same
+ * literal `MLCore.verify()` emits, instead of duplicating the string.
+ */
+export declare const CONFIG_ERROR_RULE_ID = "config-error";
+/**
+ * `ruleId` of a deprecated-but-working rule name notice — see
+ * {@link CONFIG_ERROR_RULE_ID} for why this is exported rather than a
+ * private literal.
+ */
+export declare const RULE_DEPRECATION_RULE_ID = "rule-deprecation";
 /**
  * Summary of the multi-pass fix process.
  */
@@ -22,6 +36,20 @@ export type FixSummary = {
      * multiple passes are executed.
      */
     readonly firstPassEdits: readonly TextEdit[];
+    /**
+     * Violations remaining in the final fixed code, re-verified after the
+     * last fix pass.
+     *
+     * Unlike {@link VerifyResult.violations} (which reflects the first pass
+     * only), this list is accurate for `fixedCode`. Callers that report
+     * post-fix results should prefer `fixSummary.finalPassViolations ?? violations`.
+     * Entries may carry `fix` data whose offsets refer to `fixedCode`.
+     *
+     * `undefined` when no fixes remain applied (none were applied, or the
+     * applied pass was rolled back); the first-pass violations are then
+     * accurate as-is.
+     */
+    readonly finalPassViolations?: readonly Violation[];
 };
 /**
  * Options for {@link MLCore.verify}.
@@ -60,7 +88,7 @@ export type MLCoreParams = {
  */
 export declare class MLCore {
     #private;
-    constructor({ parser, sourceCode, ruleset, rules, locale, schemas, ruleCommonSettings, parserOptions, severity, pretenders, filename, debug, configErrors, }: MLCoreParams);
+    constructor({ parser, sourceCode, ruleset, rules, locale, schemas, ruleCommonSettings, parserOptions, severity, pretenders, filename, debug, configErrors, ruleDeprecations, }: MLCoreParams);
     /**
      * The parsed document, or a {@link ParserError} if parsing failed.
      */
@@ -77,7 +105,7 @@ export declare class MLCore {
      *
      * @param fabric - Partial fabric with the properties to update
      */
-    update({ parser, ruleset, rules, locale, schemas, parserOptions, configErrors }: Partial<MLFabric>): void;
+    update({ parser, ruleset, rules, locale, schemas, parserOptions, pretenders, configErrors, ruleDeprecations, }: Partial<MLFabric>): void;
     /**
      * Runs all configured rules against the parsed document and returns violations.
      *
@@ -91,7 +119,9 @@ export declare class MLCore {
      * may be the result of multiple fix passes. This means some violations in the
      * array may already be resolved in `fixedCode`, and new violations introduced
      * during later passes are not included in the array. Callers needing an accurate
-     * violation list for the fixed code should re-verify the output.
+     * violation list for the fixed code should use `fixSummary.finalPassViolations`,
+     * which is re-verified against the final code whenever at least one fix was
+     * applied.
      *
      * @param fixOrOptions - Whether to attempt auto-fixing violations, or an options object
      * @returns Violations from the initial analysis and the (possibly fixed) source code