@markuplint/parser-utils 4.6.8 → 4.7.1

package/CHANGELOG.md

@@ -3,13 +3,26 @@
 All notable changes to this project will be documented in this file.
 See [Conventional Commits](https://conventionalcommits.org) for commit guidelines.
 
-## [4.6.8](https://github.com/markuplint/markuplint/compare/@markuplint/parser-utils@4.6.7...@markuplint/parser-utils@4.6.8) (2024-10-14)
+## [4.7.1](https://github.com/markuplint/markuplint/compare/@markuplint/parser-utils@4.7.0...@markuplint/parser-utils@4.7.1) (2024-10-27)
+
+
+### Performance Improvements
+
+* **parser-utils:** adjusted siblings correction timing to reduce exponential complexity ([676357c](https://github.com/markuplint/markuplint/commit/676357c438df7545f472787c9032463f9fdba515))
+
 
-**Note:** Version bump only for package @markuplint/parser-utils
 
 
 
+# [4.7.0](https://github.com/markuplint/markuplint/compare/@markuplint/parser-utils@4.6.8...@markuplint/parser-utils@4.7.0) (2024-10-15)
 
+### Features
+
+- **parser-utils:** expose `getOffsetsFromCode` function ([8ef7aec](https://github.com/markuplint/markuplint/commit/8ef7aec26d3198328c86ebeffaa0bd9c879a1f0e))
+
+## [4.6.8](https://github.com/markuplint/markuplint/compare/@markuplint/parser-utils@4.6.7...@markuplint/parser-utils@4.6.8) (2024-10-14)
+
+**Note:** Version bump only for package @markuplint/parser-utils
 
 ## [4.6.7](https://github.com/markuplint/markuplint/compare/@markuplint/parser-utils@4.6.6...@markuplint/parser-utils@4.6.7) (2024-09-23)
 

package/lib/debug.d.ts

@@ -2,3 +2,8 @@ import type { MLASTNode } from '@markuplint/ml-ast';
 import debug from 'debug';
 export declare const log: debug.Debugger;
 export declare function domLog(nodeList: readonly (MLASTNode | null)[]): void;
+export declare class PerformanceTimer {
+    #private;
+    push(name?: string): "" | undefined;
+    log(): void;
+}

package/lib/debug.js

@@ -1,6 +1,62 @@
+var __classPrivateFieldGet = (this && this.__classPrivateFieldGet) || function (receiver, state, kind, f) {
+    if (kind === "a" && !f) throw new TypeError("Private accessor was defined without a getter");
+    if (typeof state === "function" ? receiver !== state || !f : !state.has(receiver)) throw new TypeError("Cannot read private member from an object whose class did not declare it");
+    return kind === "m" ? f : kind === "a" ? f.call(receiver) : f ? f.value : state.get(receiver);
+};
+var __classPrivateFieldSet = (this && this.__classPrivateFieldSet) || function (receiver, state, value, kind, f) {
+    if (kind === "m") throw new TypeError("Private method is not writable");
+    if (kind === "a" && !f) throw new TypeError("Private accessor was defined without a setter");
+    if (typeof state === "function" ? receiver !== state || !f : !state.has(receiver)) throw new TypeError("Cannot write private member to an object whose class did not declare it");
+    return (kind === "a" ? f.call(receiver, value) : f ? f.value = value : state.set(receiver, value)), value;
+};
+var _PerformanceTimer_logs, _PerformanceTimer_counter;
 import debug from 'debug';
 import { nodeListToDebugMaps } from './debugger.js';
 export const log = debug('ml-parser');
 export function domLog(nodeList) {
     log('Parse result: %O', nodeListToDebugMaps(nodeList, true));
 }
+export class PerformanceTimer {
+    constructor() {
+        _PerformanceTimer_logs.set(this, []);
+        _PerformanceTimer_counter.set(this, -1);
+    }
+    push(name) {
+        var _a;
+        if (!log.enabled) {
+            return '';
+        }
+        __classPrivateFieldSet(this, _PerformanceTimer_counter, (_a = __classPrivateFieldGet(this, _PerformanceTimer_counter, "f"), _a++, _a), "f");
+        const now = performance.now();
+        const last = __classPrivateFieldGet(this, _PerformanceTimer_logs, "f").at(-1);
+        if (last && Number.isNaN(last[2])) {
+            last[2] = now;
+        }
+        name = name || `#${__classPrivateFieldGet(this, _PerformanceTimer_counter, "f")}`;
+        __classPrivateFieldGet(this, _PerformanceTimer_logs, "f").push([name, now, Number.NaN]);
+    }
+    log() {
+        if (!log.enabled) {
+            return;
+        }
+        this.push('end');
+        __classPrivateFieldGet(this, _PerformanceTimer_logs, "f").pop();
+        const map = new Map();
+        for (const content of __classPrivateFieldGet(this, _PerformanceTimer_logs, "f")) {
+            const diff = content[2] - content[1];
+            const name = content[0];
+            if (map.has(name)) {
+                const [total, count] = map.get(name);
+                map.set(name, [total + diff, count + 1]);
+            }
+            else {
+                map.set(name, [diff, 1]);
+            }
+        }
+        for (const [name, [total, count]] of map) {
+            const avg = total / count;
+            log.extend(name)('%dms (avg: %dms, count: %d)', total.toExponential(3), avg.toExponential(3), count);
+        }
+    }
+}
+_PerformanceTimer_logs = new WeakMap(), _PerformanceTimer_counter = new WeakMap();

package/lib/get-location.d.ts

@@ -12,3 +12,7 @@ export declare function getPosition(rawCodeFragment: string, startOffset: number
 };
 export declare function getEndLine(rawCodeFragment: string, startLine: number): number;
 export declare function getEndCol(rawCodeFragment: string, startCol: number): number;
+export declare function getOffsetsFromCode(rawCode: string, startLine: number, startCol: number, endLine: number, endCol: number): {
+    offset: number;
+    endOffset: number;
+};

package/lib/get-location.js

@@ -27,3 +27,25 @@ export function getEndCol(rawCodeFragment, startCol) {
     const lastLine = lines.pop();
     return lineCount > 1 ? lastLine.length + 1 : startCol + rawCodeFragment.length;
 }
+export function getOffsetsFromCode(rawCode, startLine, startCol, endLine, endCol) {
+    const lines = rawCode.split('\n');
+    let offset = 0;
+    let endOffset = 0;
+    for (let i = 0; i < startLine - 1; i++) {
+        const line = lines[i];
+        if (line == null) {
+            continue;
+        }
+        offset += line.length + 1;
+    }
+    offset += startCol - 1;
+    for (let i = 0; i < endLine - 1; i++) {
+        const line = lines[i];
+        if (line == null) {
+            continue;
+        }
+        endOffset += line.length + 1;
+    }
+    endOffset += endCol - 1;
+    return { offset, endOffset };
+}

package/lib/parser.js

@@ -9,19 +9,20 @@ var __classPrivateFieldSet = (this && this.__classPrivateFieldSet) || function (
     if (typeof state === "function" ? receiver !== state || !f : !state.has(receiver)) throw new TypeError("Cannot write private member to an object whose class did not declare it");
     return (kind === "a" ? f.call(receiver, value) : f ? f.value = value : state.set(receiver, value)), value;
 };
-var _Parser_instances, _Parser_booleanish, _Parser_defaultState, _Parser_endTagType, _Parser_ignoreTags, _Parser_maskChar, _Parser_tagNameCaseSensitive, _Parser_selfCloseType, _Parser_spaceChars, _Parser_rawTextElements, _Parser_authoredElementName, _Parser_originalRawCode, _Parser_rawCode, _Parser_defaultDepth, _Parser_walkMethodSequentailPrevNode, _Parser_arrayize, _Parser_concatText, _Parser_concatTextNodes, _Parser_convertIntoInvalidNode, _Parser_createOffsetSpaces, _Parser_createRemnantNode, _Parser_exposeRemnantNodes, _Parser_getEndLocation, _Parser_orphanEndTagToBogusMark, _Parser_pairing, _Parser_parseEndTag, _Parser_parseStartTag, _Parser_parseTag, _Parser_removeChild, _Parser_removeDeprecatedNode, _Parser_removeOffsetSpaces, _Parser_reset, _Parser_setRawCode, _Parser_siblingsCorrection, _Parser_trimText;
+var _Parser_instances, _Parser_booleanish, _Parser_defaultState, _Parser_endTagType, _Parser_ignoreTags, _Parser_maskChar, _Parser_tagNameCaseSensitive, _Parser_selfCloseType, _Parser_spaceChars, _Parser_rawTextElements, _Parser_authoredElementName, _Parser_originalRawCode, _Parser_rawCode, _Parser_defaultDepth, _Parser_walkMethodSequentailPrevNode, _Parser_arrayize, _Parser_concatText, _Parser_concatTextNodes, _Parser_convertIntoInvalidNode, _Parser_createOffsetSpaces, _Parser_createRemnantNode, _Parser_exposeRemnantNodes, _Parser_getEndLocation, _Parser_orphanEndTagToBogusMark, _Parser_pairing, _Parser_parseEndTag, _Parser_parseStartTag, _Parser_parseTag, _Parser_removeChild, _Parser_removeDeprecatedNode, _Parser_removeOffsetSpaces, _Parser_reset, _Parser_setRawCode, _Parser_trimText;
 import { isVoidElement as detectVoidElement } from '@markuplint/ml-spec';
 import { v4 as uuid } from 'uuid';
 import { attrTokenizer } from './attr-tokenizer.js';
 import { defaultSpaces } from './const.js';
-import { domLog } from './debug.js';
+import { domLog, PerformanceTimer } from './debug.js';
 import { detectElementType } from './detect-element-type.js';
 import { AttrState, TagState } from './enums.js';
-import { getEndCol, getEndLine, getPosition } from './get-location.js';
+import { getEndCol, getEndLine, getOffsetsFromCode, getPosition } from './get-location.js';
 import { ignoreBlock, restoreNode } from './ignore-block.js';
 import { ignoreFrontMatter } from './ignore-front-matter.js';
 import { ParserError } from './parser-error.js';
 import { sortNodes } from './sort-nodes.js';
+const timer = new PerformanceTimer();
 export class Parser {
     constructor(options, defaultState) {
         _Parser_instances.add(this);
@@ -96,28 +97,40 @@ export class Parser {
         try {
             // Initialize raw code
             __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_setRawCode).call(this, rawCode, rawCode);
+            timer.push('beforeParse');
+            const beforeParsedCode = this.beforeParse(this.rawCode, options);
             // Override raw code
-            __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_setRawCode).call(this, this.beforeParse(this.rawCode, options));
+            __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_setRawCode).call(this, beforeParsedCode);
             __classPrivateFieldSet(this, _Parser_authoredElementName, options?.authoredElementName, "f");
             let frontMatter = null;
             if (options?.ignoreFrontMatter) {
+                timer.push('ignoreFrontMatter');
                 const fm = ignoreFrontMatter(this.rawCode);
                 __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_setRawCode).call(this, fm.code);
                 frontMatter = fm.frontMatter;
             }
+            timer.push('ignoreBlock');
             const blocks = ignoreBlock(this.rawCode, __classPrivateFieldGet(this, _Parser_ignoreTags, "f"), __classPrivateFieldGet(this, _Parser_maskChar, "f"));
             __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_setRawCode).call(this, blocks.replaced);
+            timer.push('tokenize');
             const tokenized = this.tokenize(options);
             const ast = tokenized.ast;
             const isFragment = tokenized.isFragment;
             __classPrivateFieldSet(this, _Parser_defaultDepth, options?.depth ?? __classPrivateFieldGet(this, _Parser_defaultDepth, "f"), "f");
+            timer.push('traverse');
             const traversed = this.traverse(ast, null, __classPrivateFieldGet(this, _Parser_defaultDepth, "f"));
+            timer.push('afterTraverse');
             const nodeTree = this.afterTraverse([...traversed.childNodes, ...traversed.siblings]);
+            timer.push('flattenNodes');
             let nodeList = this.flattenNodes(nodeTree);
+            timer.push('afterFlattenNodes');
             nodeList = this.afterFlattenNodes(nodeList);
+            timer.push('restoreNode');
             nodeList = restoreNode(this, nodeList, blocks, false);
+            timer.push('afterParse');
             nodeList = this.afterParse(nodeList, options);
             if (frontMatter) {
+                timer.push('frontMatter');
                 const newNodeList = [...nodeList];
                 let firstText = '';
                 const firstTextNode = newNodeList.shift();
@@ -142,6 +155,7 @@ export class Parser {
                 }
                 nodeList = [fmNode, ...newNodeList];
             }
+            timer.log();
             domLog(nodeList);
             __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_reset).call(this);
             return {
@@ -174,9 +188,22 @@ export class Parser {
         }
         const childNodes = [];
         const siblings = [];
+        const existence = new Set();
         for (const originNode of originNodes) {
+            timer.push('nodeize');
             const nodes = this.nodeize(originNode, parentNode, depth);
-            const after = this.afterNodeize(nodes, parentNode, depth);
+            const filteredNodes = [];
+            for (const node of nodes) {
+                // Remove duplicated nodes
+                const id = `${node.startOffset}:${node.endOffset}:${node.nodeName}:${node.type}:${node.raw}`;
+                if (existence.has(id)) {
+                    continue;
+                }
+                existence.add(id);
+                filteredNodes.push(node);
+            }
+            timer.push('afterNodeize');
+            const after = this.afterNodeize(filteredNodes, parentNode, depth);
             childNodes.push(...after.siblings);
             siblings.push(...after.ancestors);
         }
@@ -186,7 +213,10 @@ export class Parser {
         };
     }
     afterTraverse(nodeTree) {
-        return __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_siblingsCorrection).call(this, nodeTree);
+        return Array.prototype.toSorted == null
+            ? // TODO: Use sort instead of toSorted until we end support for Node 18
+                [...nodeTree].sort(sortNodes)
+            : nodeTree.toSorted(sortNodes);
     }
     nodeize(originNode, parentNode, depth) {
         return [];
@@ -232,6 +262,7 @@ export class Parser {
         return nodeList;
     }
     visitDoctype(token) {
+        timer.push('visitDoctype');
         const node = {
             ...token,
             ...this.createToken(token),
@@ -241,6 +272,7 @@ export class Parser {
         return [node];
     }
     visitComment(token, options) {
+        timer.push('visitComment');
         const isBogus = options?.isBogus ?? !token.raw.startsWith('<!--');
         const node = {
             ...token,
@@ -252,6 +284,7 @@ export class Parser {
         return [node];
     }
     visitText(token, options) {
+        timer.push('visitText');
         const node = {
             ...token,
             ...this.createToken(token),
@@ -270,6 +303,7 @@ export class Parser {
         return [node];
     }
     visitElement(token, childNodes = [], options) {
+        timer.push('visitElement');
         const createEndTagToken = options?.createEndTagToken;
         const namelessFragment = options?.namelessFragment ?? false;
         const overwriteProps = options?.overwriteProps;
@@ -311,6 +345,7 @@ export class Parser {
         return [startTag, ...siblings];
     }
     visitPsBlock(token, childNodes = [], conditionalType = null, originBlockNode) {
+        timer.push('visitPsBlock');
         const block = {
             ...token,
             ...this.createToken(token),
@@ -335,6 +370,7 @@ export class Parser {
         return traversed.siblings;
     }
     visitSpreadAttr(token) {
+        timer.push('visitSpreadAttr');
         const raw = token.raw.trim();
         if (raw === '') {
             return null;
@@ -355,6 +391,7 @@ export class Parser {
         };
     }
     visitAttr(token, options) {
+        timer.push('visitAttr');
         const raw = token.raw;
         const quoteSet = options?.quoteSet;
         const startState = options?.startState ?? AttrState.BeforeName;
@@ -587,26 +624,7 @@ export class Parser {
         };
     }
     getOffsetsFromCode(startLine, startCol, endLine, endCol) {
-        const lines = __classPrivateFieldGet(this, _Parser_rawCode, "f").split('\n');
-        let offset = 0;
-        let endOffset = 0;
-        for (let i = 0; i < startLine - 1; i++) {
-            const line = lines[i];
-            if (line == null) {
-                continue;
-            }
-            offset += line.length + 1;
-        }
-        offset += startCol - 1;
-        for (let i = 0; i < endLine - 1; i++) {
-            const line = lines[i];
-            if (line == null) {
-                continue;
-            }
-            endOffset += line.length + 1;
-        }
-        endOffset += endCol - 1;
-        return { offset, endOffset };
+        return getOffsetsFromCode(this.rawCode, startLine, startCol, endLine, endCol);
     }
     walk(nodeList, walker, depth = 0) {
         for (const node of nodeList) {
@@ -635,7 +653,10 @@ export class Parser {
             newChildNodes.splice(currentIndex, 1, appendingChild);
         }
         Object.assign(parentNode, {
-            childNodes: __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_siblingsCorrection).call(this, newChildNodes),
+            childNodes: Array.prototype.toSorted == null
+                ? // TODO: Use sort instead of toSorted until we end support for Node 18
+                    [...newChildNodes].sort(sortNodes)
+                : newChildNodes.toSorted(sortNodes),
         });
     }
     replaceChild(parentNode, oldChildNode, ...replacementChildNodes) {
@@ -792,12 +813,9 @@ _Parser_booleanish = new WeakMap(), _Parser_defaultState = new WeakMap(), _Parse
         newNodeList.push(node);
     }
     return newNodeList;
-}, _Parser_pairing = function _Parser_pairing(startTag, endTag, appendChild = true) {
+}, _Parser_pairing = function _Parser_pairing(startTag, endTag) {
     Object.assign(startTag, { pairNode: endTag });
     Object.assign(endTag, { pairNode: startTag });
-    if (!appendChild) {
-        return;
-    }
     this.appendChild(startTag.parentNode, endTag);
 }, _Parser_parseEndTag = function _Parser_parseEndTag(token, namelessFragment) {
     const parsed = __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_parseTag).call(this, token, true, false, namelessFragment);
@@ -1080,34 +1098,6 @@ _Parser_booleanish = new WeakMap(), _Parser_defaultState = new WeakMap(), _Parse
 }, _Parser_setRawCode = function _Parser_setRawCode(rawCode, originalRawCode) {
     __classPrivateFieldSet(this, _Parser_rawCode, rawCode, "f");
     __classPrivateFieldSet(this, _Parser_originalRawCode, originalRawCode ?? __classPrivateFieldGet(this, _Parser_originalRawCode, "f"), "f");
-}, _Parser_siblingsCorrection = function _Parser_siblingsCorrection(nodes) {
-    const stack = new Set();
-    const newNodes = [];
-    const oldNodes = Array.prototype.toSorted == null
-        ? // TODO: Use sort instead of toSorted until we end support for Node 18
-            [...nodes].sort(sortNodes)
-        : nodes.toSorted(sortNodes);
-    const nameToLastOpenTag = {};
-    for (const node of oldNodes) {
-        const id = `${node.startOffset}::${node.nodeName}`;
-        if (stack.has(id)) {
-            continue;
-        }
-        stack.add(id);
-        if (node.type === 'endtag') {
-            const openTag = nameToLastOpenTag[node.nodeName];
-            if (openTag && !openTag.pairNode) {
-                __classPrivateFieldGet(this, _Parser_instances, "m", _Parser_pairing).call(this, openTag, node, false);
-                newNodes.push(node);
-                continue;
-            }
-        }
-        else if (node.type === 'starttag') {
-            nameToLastOpenTag[node.nodeName] = node;
-        }
-        newNodes.push(node);
-    }
-    return newNodes;
 }, _Parser_trimText = function _Parser_trimText(nodeList) {
     const newNodeList = [];
     let prevNode = null;

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@markuplint/parser-utils",
-	"version": "4.6.8",
+	"version": "4.7.1",
 	"description": "Utility module for markuplint parser plugin",
 	"repository": "git@github.com:markuplint/markuplint.git",
 	"author": "Yusuke Hirao <yusukehirao@me.com>",
@@ -28,9 +28,9 @@
 		"clean": "tsc --build --clean"
 	},
 	"dependencies": {
-		"@markuplint/ml-ast": "4.4.5",
-		"@markuplint/ml-spec": "4.7.1",
-		"@markuplint/types": "4.6.1",
+		"@markuplint/ml-ast": "4.4.7",
+		"@markuplint/ml-spec": "4.8.0",
+		"@markuplint/types": "4.6.3",
 		"@types/uuid": "10.0.0",
 		"debug": "4.3.7",
 		"espree": "10.2.0",
@@ -38,7 +38,7 @@
 		"uuid": "10.0.0"
 	},
 	"devDependencies": {
-		"@typescript-eslint/typescript-estree": "8.8.1"
+		"@typescript-eslint/typescript-estree": "8.11.0"
 	},
-	"gitHead": "e59d4e8b762c66c7e3fb8b0a0d9d99d5160b0afa"
+	"gitHead": "fab5b494f0bdc491aa83cb2c8722738d557fbefd"
 }