@mytechtoday/url-reference-mapper 1.3.0 → 2.0.0-alpha.1

package/dist/extractors.d.ts

@@ -0,0 +1,25 @@
+export interface ExtractedMetadata {
+    wordCount?: number;
+    readingTime?: number;
+    tags?: string;
+    summary?: string;
+    tldr?: string;
+    categories?: string;
+    author?: string;
+    featuredImages?: string[];
+    authorImage?: string;
+    authorUrl?: string;
+    quotes?: string[];
+    internalLinks?: string[];
+    externalLinks?: string[];
+    relatedPosts?: string[];
+}
+export interface ExtractionConfig {
+    readingSpeed?: number;
+    maxTags?: number;
+    maxQuotes?: number;
+    maxLinks?: number;
+    baseUrl?: string;
+}
+export declare function extractMetadata(filePath: string, config?: ExtractionConfig): Promise<ExtractedMetadata>;
+//# sourceMappingURL=extractors.d.ts.map

package/dist/extractors.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"extractors.d.ts","sourceRoot":"","sources":["../src/extractors.ts"],"names":[],"mappings":"AAMA,MAAM,WAAW,iBAAiB;IAChC,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,cAAc,CAAC,EAAE,MAAM,EAAE,CAAC;IAC1B,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,MAAM,CAAC,EAAE,MAAM,EAAE,CAAC;IAClB,aAAa,CAAC,EAAE,MAAM,EAAE,CAAC;IACzB,aAAa,CAAC,EAAE,MAAM,EAAE,CAAC;IACzB,YAAY,CAAC,EAAE,MAAM,EAAE,CAAC;CACzB;AAKD,MAAM,WAAW,gBAAgB;IAE/B,YAAY,CAAC,EAAE,MAAM,CAAC;IAEtB,OAAO,CAAC,EAAE,MAAM,CAAC;IAEjB,SAAS,CAAC,EAAE,MAAM,CAAC;IAEnB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAElB,OAAO,CAAC,EAAE,MAAM,CAAC;CAClB;AASD,wBAAsB,eAAe,CACnC,QAAQ,EAAE,MAAM,EAChB,MAAM,GAAE,gBAAqB,GAC5B,OAAO,CAAC,iBAAiB,CAAC,CAoB5B"}

package/dist/extractors.js

@@ -0,0 +1,320 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || function (mod) {
+    if (mod && mod.__esModule) return mod;
+    var result = {};
+    if (mod != null) for (var k in mod) if (k !== "default" && Object.prototype.hasOwnProperty.call(mod, k)) __createBinding(result, mod, k);
+    __setModuleDefault(result, mod);
+    return result;
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.extractMetadata = void 0;
+const fs = __importStar(require("fs"));
+const path = __importStar(require("path"));
+async function extractMetadata(filePath, config = {}) {
+    if (!fs.existsSync(filePath)) {
+        throw new Error(`File not found: ${filePath}`);
+    }
+    const content = fs.readFileSync(filePath, 'utf-8');
+    const ext = path.extname(filePath).toLowerCase();
+    switch (ext) {
+        case '.md':
+        case '.markdown':
+            return extractFromMarkdown(content, config);
+        case '.html':
+        case '.htm':
+            return extractFromHtml(content, config);
+        case '.txt':
+            return extractFromText(content, config);
+        default:
+            return extractFromText(content, config);
+    }
+}
+exports.extractMetadata = extractMetadata;
+function extractFromMarkdown(content, config) {
+    const metadata = {};
+    const frontmatterMatch = content.match(/^---\n([\s\S]*?)\n---/);
+    if (frontmatterMatch) {
+        const frontmatter = frontmatterMatch[1];
+        const authorMatch = frontmatter.match(/author:\s*(.+)/i);
+        if (authorMatch)
+            metadata.author = authorMatch[1].trim();
+        const categoriesMatch = frontmatter.match(/categories?:\s*(.+)/i);
+        if (categoriesMatch)
+            metadata.categories = categoriesMatch[1].trim();
+        const tagsMatch = frontmatter.match(/tags?:\s*(.+)/i);
+        if (tagsMatch)
+            metadata.tags = tagsMatch[1].trim();
+    }
+    const textContent = content.replace(/^---\n[\s\S]*?\n---\n/, '');
+    const wordCount = countWords(textContent);
+    metadata.wordCount = wordCount;
+    metadata.readingTime = calculateReadingTime(wordCount, config.readingSpeed || 225);
+    const quotes = extractQuotes(textContent, config.maxQuotes || 10);
+    if (quotes.length > 0)
+        metadata.quotes = quotes;
+    const { internalLinks, externalLinks } = extractLinks(textContent, config.baseUrl, config.maxLinks || 10);
+    if (internalLinks.length > 0)
+        metadata.internalLinks = internalLinks;
+    if (externalLinks.length > 0)
+        metadata.externalLinks = externalLinks;
+    const images = extractMarkdownImages(textContent);
+    if (images.length > 0)
+        metadata.featuredImages = images;
+    const summary = extractSummary(textContent);
+    if (summary)
+        metadata.summary = summary;
+    const tldr = extractTldr(textContent);
+    if (tldr)
+        metadata.tldr = tldr;
+    return metadata;
+}
+function extractFromHtml(content, config) {
+    const metadata = {};
+    const authorMatch = content.match(/<meta\s+name=["']author["']\s+content=["']([^"']+)["']/i);
+    if (authorMatch)
+        metadata.author = authorMatch[1];
+    const descriptionMatch = content.match(/<meta\s+name=["']description["']\s+content=["']([^"']+)["']/i);
+    if (descriptionMatch)
+        metadata.summary = descriptionMatch[1];
+    const keywordsMatch = content.match(/<meta\s+name=["']keywords["']\s+content=["']([^"']+)["']/i);
+    if (keywordsMatch)
+        metadata.tags = keywordsMatch[1];
+    const textContent = content.replace(/<[^>]+>/g, ' ').replace(/\s+/g, ' ').trim();
+    const wordCount = countWords(textContent);
+    metadata.wordCount = wordCount;
+    metadata.readingTime = calculateReadingTime(wordCount, config.readingSpeed || 225);
+    const { internalLinks, externalLinks } = extractHtmlLinks(content, config.baseUrl, config.maxLinks || 10);
+    if (internalLinks.length > 0)
+        metadata.internalLinks = internalLinks;
+    if (externalLinks.length > 0)
+        metadata.externalLinks = externalLinks;
+    const images = extractHtmlImages(content);
+    if (images.length > 0)
+        metadata.featuredImages = images;
+    const quotes = extractHtmlQuotes(content, config.maxQuotes || 10);
+    if (quotes.length > 0)
+        metadata.quotes = quotes;
+    const tldr = extractTldr(textContent);
+    if (tldr)
+        metadata.tldr = tldr;
+    return metadata;
+}
+function extractFromText(content, config) {
+    const metadata = {};
+    const wordCount = countWords(content);
+    metadata.wordCount = wordCount;
+    metadata.readingTime = calculateReadingTime(wordCount, config.readingSpeed || 225);
+    const quotes = extractQuotes(content, config.maxQuotes || 10);
+    if (quotes.length > 0)
+        metadata.quotes = quotes;
+    const { internalLinks, externalLinks } = extractTextLinks(content, config.baseUrl, config.maxLinks || 10);
+    if (internalLinks.length > 0)
+        metadata.internalLinks = internalLinks;
+    if (externalLinks.length > 0)
+        metadata.externalLinks = externalLinks;
+    const summary = extractSummary(content);
+    if (summary)
+        metadata.summary = summary;
+    const tldr = extractTldr(content);
+    if (tldr)
+        metadata.tldr = tldr;
+    return metadata;
+}
+function countWords(text) {
+    const words = text.trim().split(/\s+/).filter(word => word.length > 0);
+    return words.length;
+}
+function calculateReadingTime(wordCount, wordsPerMinute) {
+    return Math.ceil(wordCount / wordsPerMinute);
+}
+function extractQuotes(content, maxQuotes) {
+    const quotes = [];
+    const lines = content.split('\n');
+    let currentQuote = '';
+    for (const line of lines) {
+        if (line.trim().startsWith('>')) {
+            currentQuote += line.replace(/^>\s*/, '') + ' ';
+        }
+        else if (currentQuote) {
+            quotes.push(currentQuote.trim());
+            currentQuote = '';
+            if (quotes.length >= maxQuotes)
+                break;
+        }
+    }
+    if (currentQuote && quotes.length < maxQuotes) {
+        quotes.push(currentQuote.trim());
+    }
+    return quotes.slice(0, maxQuotes);
+}
+function extractHtmlQuotes(content, maxQuotes) {
+    const quotes = [];
+    const blockquoteRegex = /<blockquote[^>]*>([\s\S]*?)<\/blockquote>/gi;
+    let match;
+    while ((match = blockquoteRegex.exec(content)) !== null && quotes.length < maxQuotes) {
+        const quoteText = match[1]
+            .replace(/<[^>]+>/g, ' ')
+            .replace(/\s+/g, ' ')
+            .trim();
+        if (quoteText.length > 0) {
+            quotes.push(quoteText);
+        }
+    }
+    return quotes.slice(0, maxQuotes);
+}
+function extractLinks(content, baseUrl, maxLinks) {
+    const internalLinks = [];
+    const externalLinks = [];
+    const linkRegex = /\[([^\]]+)\]\(([^)]+)\)/g;
+    let match;
+    while ((match = linkRegex.exec(content)) !== null) {
+        const url = match[2];
+        if (url.startsWith('http://') || url.startsWith('https://')) {
+            if (baseUrl && url.startsWith(baseUrl)) {
+                if (internalLinks.length < maxLinks)
+                    internalLinks.push(url);
+            }
+            else {
+                if (externalLinks.length < maxLinks)
+                    externalLinks.push(url);
+            }
+        }
+        if (internalLinks.length >= maxLinks && externalLinks.length >= maxLinks)
+            break;
+    }
+    return { internalLinks, externalLinks };
+}
+function extractHtmlLinks(content, baseUrl, maxLinks) {
+    const internalLinks = [];
+    const externalLinks = [];
+    const linkRegex = /<a\s+[^>]*href=["']([^"']+)["']/gi;
+    let match;
+    while ((match = linkRegex.exec(content)) !== null) {
+        const url = match[1];
+        if (url.startsWith('http://') || url.startsWith('https://')) {
+            if (baseUrl && url.startsWith(baseUrl)) {
+                if (internalLinks.length < maxLinks)
+                    internalLinks.push(url);
+            }
+            else {
+                if (externalLinks.length < maxLinks)
+                    externalLinks.push(url);
+            }
+        }
+        if (internalLinks.length >= maxLinks && externalLinks.length >= maxLinks)
+            break;
+    }
+    return { internalLinks, externalLinks };
+}
+function extractTextLinks(content, baseUrl, maxLinks) {
+    const internalLinks = [];
+    const externalLinks = [];
+    const urlRegex = /https?:\/\/[^\s<>"]+/g;
+    let match;
+    while ((match = urlRegex.exec(content)) !== null) {
+        const url = match[0];
+        if (baseUrl && url.startsWith(baseUrl)) {
+            if (internalLinks.length < maxLinks)
+                internalLinks.push(url);
+        }
+        else {
+            if (externalLinks.length < maxLinks)
+                externalLinks.push(url);
+        }
+        if (internalLinks.length >= maxLinks && externalLinks.length >= maxLinks)
+            break;
+    }
+    return { internalLinks, externalLinks };
+}
+function extractMarkdownImages(content) {
+    const images = [];
+    const imageRegex = /!\[([^\]]*)\]\(([^)]+)\)/g;
+    let match;
+    while ((match = imageRegex.exec(content)) !== null) {
+        const url = match[2];
+        if (url.startsWith('http://') || url.startsWith('https://')) {
+            images.push(url);
+        }
+    }
+    return images;
+}
+function extractHtmlImages(content) {
+    const images = [];
+    const imageRegex = /<img\s+[^>]*src=["']([^"']+)["']/gi;
+    let match;
+    while ((match = imageRegex.exec(content)) !== null) {
+        const url = match[1];
+        if (url.startsWith('http://') || url.startsWith('https://')) {
+            images.push(url);
+        }
+    }
+    return images;
+}
+function extractSummary(content) {
+    const textContent = content.replace(/^---\n[\s\S]*?\n---\n/, '');
+    const paragraphs = textContent.split(/\n\n+/).map(p => p.trim()).filter(p => p.length > 0);
+    for (const para of paragraphs) {
+        if (!para.startsWith('#') && para.length > 50 && para.length < 500) {
+            const cleaned = para
+                .replace(/\[([^\]]+)\]\([^)]+\)/g, '$1')
+                .replace(/[*_`]/g, '')
+                .replace(/^>\s*/gm, '')
+                .trim();
+            const firstSentence = cleaned.split(/[.!?]\s/)[0];
+            if (firstSentence && firstSentence.length > 30) {
+                return firstSentence + '.';
+            }
+        }
+    }
+    return undefined;
+}
+function extractTldr(content) {
+    const textContent = content.replace(/^---\n[\s\S]*?\n---\n/, '');
+    const paragraphs = textContent
+        .split(/\n\n+/)
+        .map(p => p.trim())
+        .filter(p => p.length > 0 && !p.startsWith('#'));
+    if (paragraphs.length === 0)
+        return undefined;
+    let tldr = '';
+    for (const para of paragraphs) {
+        const cleaned = para
+            .replace(/\[([^\]]+)\]\([^)]+\)/g, '$1')
+            .replace(/[*_`]/g, '')
+            .replace(/^>\s*/gm, '')
+            .replace(/!\[([^\]]*)\]\([^)]+\)/g, '')
+            .trim();
+        if (cleaned.length > 0) {
+            if (tldr.length === 0) {
+                tldr = cleaned;
+            }
+            else {
+                tldr += ' ' + cleaned;
+            }
+            if (tldr.length >= 200) {
+                break;
+            }
+        }
+    }
+    if (tldr.length > 900) {
+        tldr = tldr.substring(0, 897) + '...';
+    }
+    return tldr.length >= 200 ? tldr : undefined;
+}
+//# sourceMappingURL=extractors.js.map

package/dist/extractors.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"extractors.js","sourceRoot":"","sources":["../src/extractors.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;AAAA,uCAAyB;AACzB,2CAA6B;AA6CtB,KAAK,UAAU,eAAe,CACnC,QAAgB,EAChB,SAA2B,EAAE;IAE7B,IAAI,CAAC,EAAE,CAAC,UAAU,CAAC,QAAQ,CAAC,EAAE,CAAC;QAC7B,MAAM,IAAI,KAAK,CAAC,mBAAmB,QAAQ,EAAE,CAAC,CAAC;IACjD,CAAC;IAED,MAAM,OAAO,GAAG,EAAE,CAAC,YAAY,CAAC,QAAQ,EAAE,OAAO,CAAC,CAAC;IACnD,MAAM,GAAG,GAAG,IAAI,CAAC,OAAO,CAAC,QAAQ,CAAC,CAAC,WAAW,EAAE,CAAC;IAEjD,QAAQ,GAAG,EAAE,CAAC;QACZ,KAAK,KAAK,CAAC;QACX,KAAK,WAAW;YACd,OAAO,mBAAmB,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;QAC9C,KAAK,OAAO,CAAC;QACb,KAAK,MAAM;YACT,OAAO,eAAe,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;QAC1C,KAAK,MAAM;YACT,OAAO,eAAe,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;QAC1C;YACE,OAAO,eAAe,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;IAC5C,CAAC;AACH,CAAC;AAvBD,0CAuBC;AAKD,SAAS,mBAAmB,CAAC,OAAe,EAAE,MAAwB;IACpE,MAAM,QAAQ,GAAsB,EAAE,CAAC;IAGvC,MAAM,gBAAgB,GAAG,OAAO,CAAC,KAAK,CAAC,uBAAuB,CAAC,CAAC;IAChE,IAAI,gBAAgB,EAAE,CAAC;QACrB,MAAM,WAAW,GAAG,gBAAgB,CAAC,CAAC,CAAC,CAAC;QAGxC,MAAM,WAAW,GAAG,WAAW,CAAC,KAAK,CAAC,iBAAiB,CAAC,CAAC;QACzD,IAAI,WAAW;YAAE,QAAQ,CAAC,MAAM,GAAG,WAAW,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAGzD,MAAM,eAAe,GAAG,WAAW,CAAC,KAAK,CAAC,sBAAsB,CAAC,CAAC;QAClE,IAAI,eAAe;YAAE,QAAQ,CAAC,UAAU,GAAG,eAAe,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;QAGrE,MAAM,SAAS,GAAG,WAAW,CAAC,KAAK,CAAC,gBAAgB,CAAC,CAAC;QACtD,IAAI,SAAS;YAAE,QAAQ,CAAC,IAAI,GAAG,SAAS,CAAC,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;IACrD,CAAC;IAGD,MAAM,WAAW,GAAG,OAAO,CAAC,OAAO,CAAC,uBAAuB,EAAE,EAAE,CAAC,CAAC;IAGjE,MAAM,SAAS,GAAG,UAAU,CAAC,WAAW,CAAC,CAAC;IAC1C,QAAQ,CAAC,SAAS,GAAG,SAAS,CAAC;IAC/B,QAAQ,CAAC,WAAW,GAAG,oBAAoB,CAAC,SAAS,EAAE,MAAM,CAAC,YAAY,IAAI,GAAG,CAAC,CAAC;IAGnF,MAAM,MAAM,GAAG,aAAa,CAAC,WAAW,EAAE,MAAM,CAAC,SAAS,IAAI,EAAE,CAAC,CAAC;IAClE,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,MAAM,GAAG,MAAM,CAAC;IAGhD,MAAM,EAAE,aAAa,EAAE,aAAa,EAAE,GAAG,YAAY,CAAC,WAAW,EAAE,MAAM,CAAC,OAAO,EAAE,MAAM,CAAC,QAAQ,IAAI,EAAE,CAAC,CAAC;IAC1G,IAAI,aAAa,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,aAAa,GAAG,aAAa,CAAC;IACrE,IAAI,aAAa,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,aAAa,GAAG,aAAa,CAAC;IAGrE,MAAM,MAAM,GAAG,qBAAqB,CAAC,WAAW,CAAC,CAAC;IAClD,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,cAAc,GAAG,MAAM,CAAC;IAGxD,MAAM,OAAO,GAAG,cAAc,CAAC,WAAW,CAAC,CAAC;IAC5C,IAAI,OAAO;QAAE,QAAQ,CAAC,OAAO,GAAG,OAAO,CAAC;IAGxC,MAAM,IAAI,GAAG,WAAW,CAAC,WAAW,CAAC,CAAC;IACtC,IAAI,IAAI;QAAE,QAAQ,CAAC,IAAI,GAAG,IAAI,CAAC;IAE/B,OAAO,QAAQ,CAAC;AAClB,CAAC;AAKD,SAAS,eAAe,CAAC,OAAe,EAAE,MAAwB;IAChE,MAAM,QAAQ,GAAsB,EAAE,CAAC;IAGvC,MAAM,WAAW,GAAG,OAAO,CAAC,KAAK,CAAC,yDAAyD,CAAC,CAAC;IAC7F,IAAI,WAAW;QAAE,QAAQ,CAAC,MAAM,GAAG,WAAW,CAAC,CAAC,CAAC,CAAC;IAElD,MAAM,gBAAgB,GAAG,OAAO,CAAC,KAAK,CAAC,8DAA8D,CAAC,CAAC;IACvG,IAAI,gBAAgB;QAAE,QAAQ,CAAC,OAAO,GAAG,gBAAgB,CAAC,CAAC,CAAC,CAAC;IAE7D,MAAM,aAAa,GAAG,OAAO,CAAC,KAAK,CAAC,2DAA2D,CAAC,CAAC;IACjG,IAAI,aAAa;QAAE,QAAQ,CAAC,IAAI,GAAG,aAAa,CAAC,CAAC,CAAC,CAAC;IAGpD,MAAM,WAAW,GAAG,OAAO,CAAC,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC,CAAC,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC,CAAC,IAAI,EAAE,CAAC;IAGjF,MAAM,SAAS,GAAG,UAAU,CAAC,WAAW,CAAC,CAAC;IAC1C,QAAQ,CAAC,SAAS,GAAG,SAAS,CAAC;IAC/B,QAAQ,CAAC,WAAW,GAAG,oBAAoB,CAAC,SAAS,EAAE,MAAM,CAAC,YAAY,IAAI,GAAG,CAAC,CAAC;IAGnF,MAAM,EAAE,aAAa,EAAE,aAAa,EAAE,GAAG,gBAAgB,CAAC,OAAO,EAAE,MAAM,CAAC,OAAO,EAAE,MAAM,CAAC,QAAQ,IAAI,EAAE,CAAC,CAAC;IAC1G,IAAI,aAAa,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,aAAa,GAAG,aAAa,CAAC;IACrE,IAAI,aAAa,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,aAAa,GAAG,aAAa,CAAC;IAGrE,MAAM,MAAM,GAAG,iBAAiB,CAAC,OAAO,CAAC,CAAC;IAC1C,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,cAAc,GAAG,MAAM,CAAC;IAGxD,MAAM,MAAM,GAAG,iBAAiB,CAAC,OAAO,EAAE,MAAM,CAAC,SAAS,IAAI,EAAE,CAAC,CAAC;IAClE,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,MAAM,GAAG,MAAM,CAAC;IAGhD,MAAM,IAAI,GAAG,WAAW,CAAC,WAAW,CAAC,CAAC;IACtC,IAAI,IAAI;QAAE,QAAQ,CAAC,IAAI,GAAG,IAAI,CAAC;IAE/B,OAAO,QAAQ,CAAC;AAClB,CAAC;AAKD,SAAS,eAAe,CAAC,OAAe,EAAE,MAAwB;IAChE,MAAM,QAAQ,GAAsB,EAAE,CAAC;IAGvC,MAAM,SAAS,GAAG,UAAU,CAAC,OAAO,CAAC,CAAC;IACtC,QAAQ,CAAC,SAAS,GAAG,SAAS,CAAC;IAC/B,QAAQ,CAAC,WAAW,GAAG,oBAAoB,CAAC,SAAS,EAAE,MAAM,CAAC,YAAY,IAAI,GAAG,CAAC,CAAC;IAGnF,MAAM,MAAM,GAAG,aAAa,CAAC,OAAO,EAAE,MAAM,CAAC,SAAS,IAAI,EAAE,CAAC,CAAC;IAC9D,IAAI,MAAM,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,MAAM,GAAG,MAAM,CAAC;IAGhD,MAAM,EAAE,aAAa,EAAE,aAAa,EAAE,GAAG,gBAAgB,CAAC,OAAO,EAAE,MAAM,CAAC,OAAO,EAAE,MAAM,CAAC,QAAQ,IAAI,EAAE,CAAC,CAAC;IAC1G,IAAI,aAAa,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,aAAa,GAAG,aAAa,CAAC;IACrE,IAAI,aAAa,CAAC,MAAM,GAAG,CAAC;QAAE,QAAQ,CAAC,aAAa,GAAG,aAAa,CAAC;IAGrE,MAAM,OAAO,GAAG,cAAc,CAAC,OAAO,CAAC,CAAC;IACxC,IAAI,OAAO;QAAE,QAAQ,CAAC,OAAO,GAAG,OAAO,CAAC;IAExC,MAAM,IAAI,GAAG,WAAW,CAAC,OAAO,CAAC,CAAC;IAClC,IAAI,IAAI;QAAE,QAAQ,CAAC,IAAI,GAAG,IAAI,CAAC;IAE/B,OAAO,QAAQ,CAAC;AAClB,CAAC;AAKD,SAAS,UAAU,CAAC,IAAY;IAE9B,MAAM,KAAK,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,IAAI,CAAC,EAAE,CAAC,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IACvE,OAAO,KAAK,CAAC,MAAM,CAAC;AACtB,CAAC;AAKD,SAAS,oBAAoB,CAAC,SAAiB,EAAE,cAAsB;IACrE,OAAO,IAAI,CAAC,IAAI,CAAC,SAAS,GAAG,cAAc,CAAC,CAAC;AAC/C,CAAC;AAKD,SAAS,aAAa,CAAC,OAAe,EAAE,SAAiB;IACvD,MAAM,MAAM,GAAa,EAAE,CAAC;IAC5B,MAAM,KAAK,GAAG,OAAO,CAAC,KAAK,CAAC,IAAI,CAAC,CAAC;IAElC,IAAI,YAAY,GAAG,EAAE,CAAC;IACtB,KAAK,MAAM,IAAI,IAAI,KAAK,EAAE,CAAC;QACzB,IAAI,IAAI,CAAC,IAAI,EAAE,CAAC,UAAU,CAAC,GAAG,CAAC,EAAE,CAAC;YAChC,YAAY,IAAI,IAAI,CAAC,OAAO,CAAC,OAAO,EAAE,EAAE,CAAC,GAAG,GAAG,CAAC;QAClD,CAAC;aAAM,IAAI,YAAY,EAAE,CAAC;YACxB,MAAM,CAAC,IAAI,CAAC,YAAY,CAAC,IAAI,EAAE,CAAC,CAAC;YACjC,YAAY,GAAG,EAAE,CAAC;YAClB,IAAI,MAAM,CAAC,MAAM,IAAI,SAAS;gBAAE,MAAM;QACxC,CAAC;IACH,CAAC;IAED,IAAI,YAAY,IAAI,MAAM,CAAC,MAAM,GAAG,SAAS,EAAE,CAAC;QAC9C,MAAM,CAAC,IAAI,CAAC,YAAY,CAAC,IAAI,EAAE,CAAC,CAAC;IACnC,CAAC;IAED,OAAO,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,SAAS,CAAC,CAAC;AACpC,CAAC;AAKD,SAAS,iBAAiB,CAAC,OAAe,EAAE,SAAiB;IAC3D,MAAM,MAAM,GAAa,EAAE,CAAC;IAG5B,MAAM,eAAe,GAAG,6CAA6C,CAAC;IACtE,IAAI,KAAK,CAAC;IAEV,OAAO,CAAC,KAAK,GAAG,eAAe,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,IAAI,MAAM,CAAC,MAAM,GAAG,SAAS,EAAE,CAAC;QAErF,MAAM,SAAS,GAAG,KAAK,CAAC,CAAC,CAAC;aACvB,OAAO,CAAC,UAAU,EAAE,GAAG,CAAC;aACxB,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC;aACpB,IAAI,EAAE,CAAC;QAEV,IAAI,SAAS,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YACzB,MAAM,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;QACzB,CAAC;IACH,CAAC;IAED,OAAO,MAAM,CAAC,KAAK,CAAC,CAAC,EAAE,SAAS,CAAC,CAAC;AACpC,CAAC;AAKD,SAAS,YAAY,CACnB,OAAe,EACf,OAA2B,EAC3B,QAAgB;IAEhB,MAAM,aAAa,GAAa,EAAE,CAAC;IACnC,MAAM,aAAa,GAAa,EAAE,CAAC;IAGnC,MAAM,SAAS,GAAG,0BAA0B,CAAC;IAC7C,IAAI,KAAK,CAAC;IAEV,OAAO,CAAC,KAAK,GAAG,SAAS,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAClD,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QAErB,IAAI,GAAG,CAAC,UAAU,CAAC,SAAS,CAAC,IAAI,GAAG,CAAC,UAAU,CAAC,UAAU,CAAC,EAAE,CAAC;YAC5D,IAAI,OAAO,IAAI,GAAG,CAAC,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;gBACvC,IAAI,aAAa,CAAC,MAAM,GAAG,QAAQ;oBAAE,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YAC/D,CAAC;iBAAM,CAAC;gBACN,IAAI,aAAa,CAAC,MAAM,GAAG,QAAQ;oBAAE,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YAC/D,CAAC;QACH,CAAC;QAED,IAAI,aAAa,CAAC,MAAM,IAAI,QAAQ,IAAI,aAAa,CAAC,MAAM,IAAI,QAAQ;YAAE,MAAM;IAClF,CAAC;IAED,OAAO,EAAE,aAAa,EAAE,aAAa,EAAE,CAAC;AAC1C,CAAC;AAKD,SAAS,gBAAgB,CACvB,OAAe,EACf,OAA2B,EAC3B,QAAgB;IAEhB,MAAM,aAAa,GAAa,EAAE,CAAC;IACnC,MAAM,aAAa,GAAa,EAAE,CAAC;IAGnC,MAAM,SAAS,GAAG,mCAAmC,CAAC;IACtD,IAAI,KAAK,CAAC;IAEV,OAAO,CAAC,KAAK,GAAG,SAAS,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAClD,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QAErB,IAAI,GAAG,CAAC,UAAU,CAAC,SAAS,CAAC,IAAI,GAAG,CAAC,UAAU,CAAC,UAAU,CAAC,EAAE,CAAC;YAC5D,IAAI,OAAO,IAAI,GAAG,CAAC,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;gBACvC,IAAI,aAAa,CAAC,MAAM,GAAG,QAAQ;oBAAE,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YAC/D,CAAC;iBAAM,CAAC;gBACN,IAAI,aAAa,CAAC,MAAM,GAAG,QAAQ;oBAAE,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;YAC/D,CAAC;QACH,CAAC;QAED,IAAI,aAAa,CAAC,MAAM,IAAI,QAAQ,IAAI,aAAa,CAAC,MAAM,IAAI,QAAQ;YAAE,MAAM;IAClF,CAAC;IAED,OAAO,EAAE,aAAa,EAAE,aAAa,EAAE,CAAC;AAC1C,CAAC;AAKD,SAAS,gBAAgB,CACvB,OAAe,EACf,OAA2B,EAC3B,QAAgB;IAEhB,MAAM,aAAa,GAAa,EAAE,CAAC;IACnC,MAAM,aAAa,GAAa,EAAE,CAAC;IAGnC,MAAM,QAAQ,GAAG,uBAAuB,CAAC;IACzC,IAAI,KAAK,CAAC;IAEV,OAAO,CAAC,KAAK,GAAG,QAAQ,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACjD,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QAErB,IAAI,OAAO,IAAI,GAAG,CAAC,UAAU,CAAC,OAAO,CAAC,EAAE,CAAC;YACvC,IAAI,aAAa,CAAC,MAAM,GAAG,QAAQ;gBAAE,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAC/D,CAAC;aAAM,CAAC;YACN,IAAI,aAAa,CAAC,MAAM,GAAG,QAAQ;gBAAE,aAAa,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAC/D,CAAC;QAED,IAAI,aAAa,CAAC,MAAM,IAAI,QAAQ,IAAI,aAAa,CAAC,MAAM,IAAI,QAAQ;YAAE,MAAM;IAClF,CAAC;IAED,OAAO,EAAE,aAAa,EAAE,aAAa,EAAE,CAAC;AAC1C,CAAC;AAKD,SAAS,qBAAqB,CAAC,OAAe;IAC5C,MAAM,MAAM,GAAa,EAAE,CAAC;IAG5B,MAAM,UAAU,GAAG,2BAA2B,CAAC;IAC/C,IAAI,KAAK,CAAC;IAEV,OAAO,CAAC,KAAK,GAAG,UAAU,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACnD,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QACrB,IAAI,GAAG,CAAC,UAAU,CAAC,SAAS,CAAC,IAAI,GAAG,CAAC,UAAU,CAAC,UAAU,CAAC,EAAE,CAAC;YAC5D,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACnB,CAAC;IACH,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC;AAKD,SAAS,iBAAiB,CAAC,OAAe;IACxC,MAAM,MAAM,GAAa,EAAE,CAAC;IAG5B,MAAM,UAAU,GAAG,oCAAoC,CAAC;IACxD,IAAI,KAAK,CAAC;IAEV,OAAO,CAAC,KAAK,GAAG,UAAU,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACnD,MAAM,GAAG,GAAG,KAAK,CAAC,CAAC,CAAC,CAAC;QACrB,IAAI,GAAG,CAAC,UAAU,CAAC,SAAS,CAAC,IAAI,GAAG,CAAC,UAAU,CAAC,UAAU,CAAC,EAAE,CAAC;YAC5D,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QACnB,CAAC;IACH,CAAC;IAED,OAAO,MAAM,CAAC;AAChB,CAAC;AAKD,SAAS,cAAc,CAAC,OAAe;IAErC,MAAM,WAAW,GAAG,OAAO,CAAC,OAAO,CAAC,uBAAuB,EAAE,EAAE,CAAC,CAAC;IAGjE,MAAM,UAAU,GAAG,WAAW,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC;IAG3F,KAAK,MAAM,IAAI,IAAI,UAAU,EAAE,CAAC;QAC9B,IAAI,CAAC,IAAI,CAAC,UAAU,CAAC,GAAG,CAAC,IAAI,IAAI,CAAC,MAAM,GAAG,EAAE,IAAI,IAAI,CAAC,MAAM,GAAG,GAAG,EAAE,CAAC;YAEnE,MAAM,OAAO,GAAG,IAAI;iBACjB,OAAO,CAAC,wBAAwB,EAAE,IAAI,CAAC;iBACvC,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC;iBACrB,OAAO,CAAC,SAAS,EAAE,EAAE,CAAC;iBACtB,IAAI,EAAE,CAAC;YAGV,MAAM,aAAa,GAAG,OAAO,CAAC,KAAK,CAAC,SAAS,CAAC,CAAC,CAAC,CAAC,CAAC;YAClD,IAAI,aAAa,IAAI,aAAa,CAAC,MAAM,GAAG,EAAE,EAAE,CAAC;gBAC/C,OAAO,aAAa,GAAG,GAAG,CAAC;YAC7B,CAAC;QACH,CAAC;IACH,CAAC;IAED,OAAO,SAAS,CAAC;AACnB,CAAC;AAMD,SAAS,WAAW,CAAC,OAAe;IAElC,MAAM,WAAW,GAAG,OAAO,CAAC,OAAO,CAAC,uBAAuB,EAAE,EAAE,CAAC,CAAC;IAGjE,MAAM,UAAU,GAAG,WAAW;SAC3B,KAAK,CAAC,OAAO,CAAC;SACd,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC;SAClB,MAAM,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,MAAM,GAAG,CAAC,IAAI,CAAC,CAAC,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC;IAEnD,IAAI,UAAU,CAAC,MAAM,KAAK,CAAC;QAAE,OAAO,SAAS,CAAC;IAG9C,IAAI,IAAI,GAAG,EAAE,CAAC;IACd,KAAK,MAAM,IAAI,IAAI,UAAU,EAAE,CAAC;QAE9B,MAAM,OAAO,GAAG,IAAI;aACjB,OAAO,CAAC,wBAAwB,EAAE,IAAI,CAAC;aACvC,OAAO,CAAC,QAAQ,EAAE,EAAE,CAAC;aACrB,OAAO,CAAC,SAAS,EAAE,EAAE,CAAC;aACtB,OAAO,CAAC,yBAAyB,EAAE,EAAE,CAAC;aACtC,IAAI,EAAE,CAAC;QAEV,IAAI,OAAO,CAAC,MAAM,GAAG,CAAC,EAAE,CAAC;YACvB,IAAI,IAAI,CAAC,MAAM,KAAK,CAAC,EAAE,CAAC;gBACtB,IAAI,GAAG,OAAO,CAAC;YACjB,CAAC;iBAAM,CAAC;gBACN,IAAI,IAAI,GAAG,GAAG,OAAO,CAAC;YACxB,CAAC;YAGD,IAAI,IAAI,CAAC,MAAM,IAAI,GAAG,EAAE,CAAC;gBACvB,MAAM;YACR,CAAC;QACH,CAAC;IACH,CAAC;IAGD,IAAI,IAAI,CAAC,MAAM,GAAG,GAAG,EAAE,CAAC;QACtB,IAAI,GAAG,IAAI,CAAC,SAAS,CAAC,CAAC,EAAE,GAAG,CAAC,GAAG,KAAK,CAAC;IACxC,CAAC;IAGD,OAAO,IAAI,CAAC,MAAM,IAAI,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,SAAS,CAAC;AAC/C,CAAC"}

package/dist/parsers/html-parser.d.ts

@@ -0,0 +1,30 @@
+export interface HtmlParseResult {
+    text: string;
+    meta: {
+        title?: string;
+        description?: string;
+        keywords?: string[];
+        author?: string;
+        ogTitle?: string;
+        ogDescription?: string;
+        ogImage?: string;
+        ogUrl?: string;
+        twitterCard?: string;
+        twitterTitle?: string;
+        twitterDescription?: string;
+        twitterImage?: string;
+        [key: string]: any;
+    };
+    links: Array<{
+        url: string;
+        text?: string;
+        title?: string;
+    }>;
+    images: Array<{
+        src: string;
+        alt?: string;
+        title?: string;
+    }>;
+}
+export declare function parseHtml(html: string): HtmlParseResult;
+//# sourceMappingURL=html-parser.d.ts.map

package/dist/parsers/html-parser.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"html-parser.d.ts","sourceRoot":"","sources":["../../src/parsers/html-parser.ts"],"names":[],"mappings":"AAaA,MAAM,WAAW,eAAe;IAE9B,IAAI,EAAE,MAAM,CAAC;IAEb,IAAI,EAAE;QACJ,KAAK,CAAC,EAAE,MAAM,CAAC;QACf,WAAW,CAAC,EAAE,MAAM,CAAC;QACrB,QAAQ,CAAC,EAAE,MAAM,EAAE,CAAC;QACpB,MAAM,CAAC,EAAE,MAAM,CAAC;QAChB,OAAO,CAAC,EAAE,MAAM,CAAC;QACjB,aAAa,CAAC,EAAE,MAAM,CAAC;QACvB,OAAO,CAAC,EAAE,MAAM,CAAC;QACjB,KAAK,CAAC,EAAE,MAAM,CAAC;QACf,WAAW,CAAC,EAAE,MAAM,CAAC;QACrB,YAAY,CAAC,EAAE,MAAM,CAAC;QACtB,kBAAkB,CAAC,EAAE,MAAM,CAAC;QAC5B,YAAY,CAAC,EAAE,MAAM,CAAC;QACtB,CAAC,GAAG,EAAE,MAAM,GAAG,GAAG,CAAC;KACpB,CAAC;IAEF,KAAK,EAAE,KAAK,CAAC;QACX,GAAG,EAAE,MAAM,CAAC;QACZ,IAAI,CAAC,EAAE,MAAM,CAAC;QACd,KAAK,CAAC,EAAE,MAAM,CAAC;KAChB,CAAC,CAAC;IAEH,MAAM,EAAE,KAAK,CAAC;QACZ,GAAG,EAAE,MAAM,CAAC;QACZ,GAAG,CAAC,EAAE,MAAM,CAAC;QACb,KAAK,CAAC,EAAE,MAAM,CAAC;KAChB,CAAC,CAAC;CACJ;AAQD,wBAAgB,SAAS,CAAC,IAAI,EAAE,MAAM,GAAG,eAAe,CAsCvD"}

package/dist/parsers/html-parser.js

@@ -0,0 +1,116 @@
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || function (mod) {
+    if (mod && mod.__esModule) return mod;
+    var result = {};
+    if (mod != null) for (var k in mod) if (k !== "default" && Object.prototype.hasOwnProperty.call(mod, k)) __createBinding(result, mod, k);
+    __setModuleDefault(result, mod);
+    return result;
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.parseHtml = void 0;
+const cheerio = __importStar(require("cheerio"));
+function parseHtml(html) {
+    try {
+        const $ = cheerio.load(html, {
+            xml: {
+                xmlMode: false,
+            },
+        });
+        const text = extractPlainText($);
+        const meta = extractMetaTags($);
+        const links = extractLinks($);
+        const images = extractImages($);
+        return {
+            text,
+            meta,
+            links,
+            images,
+        };
+    }
+    catch (error) {
+        console.warn('HTML parsing error:', error);
+        return {
+            text: html.replace(/<[^>]*>/g, ''),
+            meta: {},
+            links: [],
+            images: [],
+        };
+    }
+}
+exports.parseHtml = parseHtml;
+function extractPlainText($) {
+    $('script, style, noscript').remove();
+    const text = $('body').text() || $.text();
+    return text
+        .replace(/\s+/g, ' ')
+        .replace(/\n{3,}/g, '\n\n')
+        .trim();
+}
+function extractMetaTags($) {
+    const meta = {};
+    meta.title = $('title').text() || $('meta[property="og:title"]').attr('content');
+    meta.description = $('meta[name="description"]').attr('content');
+    meta.keywords = $('meta[name="keywords"]').attr('content')?.split(',').map(k => k.trim());
+    meta.author = $('meta[name="author"]').attr('content');
+    meta.ogTitle = $('meta[property="og:title"]').attr('content');
+    meta.ogDescription = $('meta[property="og:description"]').attr('content');
+    meta.ogImage = $('meta[property="og:image"]').attr('content');
+    meta.ogUrl = $('meta[property="og:url"]').attr('content');
+    meta.twitterCard = $('meta[name="twitter:card"]').attr('content');
+    meta.twitterTitle = $('meta[name="twitter:title"]').attr('content');
+    meta.twitterDescription = $('meta[name="twitter:description"]').attr('content');
+    meta.twitterImage = $('meta[name="twitter:image"]').attr('content');
+    Object.keys(meta).forEach(key => {
+        if (meta[key] === undefined) {
+            delete meta[key];
+        }
+    });
+    return meta;
+}
+function extractLinks($) {
+    const links = [];
+    $('a[href]').each((_, element) => {
+        const $el = $(element);
+        const href = $el.attr('href');
+        if (href) {
+            links.push({
+                url: href,
+                text: $el.text().trim() || undefined,
+                title: $el.attr('title') || undefined,
+            });
+        }
+    });
+    return links;
+}
+function extractImages($) {
+    const images = [];
+    $('img[src]').each((_, element) => {
+        const $el = $(element);
+        const src = $el.attr('src');
+        if (src) {
+            images.push({
+                src,
+                alt: $el.attr('alt') || undefined,
+                title: $el.attr('title') || undefined,
+            });
+        }
+    });
+    return images;
+}
+//# sourceMappingURL=html-parser.js.map

package/dist/parsers/html-parser.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"html-parser.js","sourceRoot":"","sources":["../../src/parsers/html-parser.ts"],"names":[],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;AAWA,iDAAmC;AAyCnC,SAAgB,SAAS,CAAC,IAAY;IACpC,IAAI,CAAC;QACH,MAAM,CAAC,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,EAAE;YAC3B,GAAG,EAAE;gBACH,OAAO,EAAE,KAAK;aACf;SACF,CAAC,CAAC;QAGH,MAAM,IAAI,GAAG,gBAAgB,CAAC,CAAC,CAAC,CAAC;QAGjC,MAAM,IAAI,GAAG,eAAe,CAAC,CAAC,CAAC,CAAC;QAGhC,MAAM,KAAK,GAAG,YAAY,CAAC,CAAC,CAAC,CAAC;QAG9B,MAAM,MAAM,GAAG,aAAa,CAAC,CAAC,CAAC,CAAC;QAEhC,OAAO;YACL,IAAI;YACJ,IAAI;YACJ,KAAK;YACL,MAAM;SACP,CAAC;IACJ,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QAEf,OAAO,CAAC,IAAI,CAAC,qBAAqB,EAAE,KAAK,CAAC,CAAC;QAG3C,OAAO;YACL,IAAI,EAAE,IAAI,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC;YAClC,IAAI,EAAE,EAAE;YACR,KAAK,EAAE,EAAE;YACT,MAAM,EAAE,EAAE;SACX,CAAC;IACJ,CAAC;AACH,CAAC;AAtCD,8BAsCC;AAKD,SAAS,gBAAgB,CAAC,CAAqB;IAE7C,CAAC,CAAC,yBAAyB,CAAC,CAAC,MAAM,EAAE,CAAC;IAGtC,MAAM,IAAI,GAAG,CAAC,CAAC,MAAM,CAAC,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC,IAAI,EAAE,CAAC;IAG1C,OAAO,IAAI;SACR,OAAO,CAAC,MAAM,EAAE,GAAG,CAAC;SACpB,OAAO,CAAC,SAAS,EAAE,MAAM,CAAC;SAC1B,IAAI,EAAE,CAAC;AACZ,CAAC;AAKD,SAAS,eAAe,CAAC,CAAqB;IAC5C,MAAM,IAAI,GAA4B,EAAE,CAAC;IAGzC,IAAI,CAAC,KAAK,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC,2BAA2B,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IACjF,IAAI,CAAC,WAAW,GAAG,CAAC,CAAC,0BAA0B,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IACjE,IAAI,CAAC,QAAQ,GAAG,CAAC,CAAC,uBAAuB,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,EAAE,KAAK,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,CAAC,CAAC,CAAC,IAAI,EAAE,CAAC,CAAC;IAC1F,IAAI,CAAC,MAAM,GAAG,CAAC,CAAC,qBAAqB,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAGvD,IAAI,CAAC,OAAO,GAAG,CAAC,CAAC,2BAA2B,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAC9D,IAAI,CAAC,aAAa,GAAG,CAAC,CAAC,iCAAiC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAC1E,IAAI,CAAC,OAAO,GAAG,CAAC,CAAC,2BAA2B,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAC9D,IAAI,CAAC,KAAK,GAAG,CAAC,CAAC,yBAAyB,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAG1D,IAAI,CAAC,WAAW,GAAG,CAAC,CAAC,2BAA2B,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAClE,IAAI,CAAC,YAAY,GAAG,CAAC,CAAC,4BAA4B,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IACpE,IAAI,CAAC,kBAAkB,GAAG,CAAC,CAAC,kCAAkC,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAChF,IAAI,CAAC,YAAY,GAAG,CAAC,CAAC,4BAA4B,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;IAGpE,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,OAAO,CAAC,GAAG,CAAC,EAAE;QAC9B,IAAI,IAAI,CAAC,GAAG,CAAC,KAAK,SAAS,EAAE,CAAC;YAC5B,OAAO,IAAI,CAAC,GAAG,CAAC,CAAC;QACnB,CAAC;IACH,CAAC,CAAC,CAAC;IAEH,OAAO,IAAI,CAAC;AACd,CAAC;AAKD,SAAS,YAAY,CAAC,CAAqB;IACzC,MAAM,KAAK,GAA6B,EAAE,CAAC;IAE3C,CAAC,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,EAAE;QAC/B,MAAM,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC;QACvB,MAAM,IAAI,GAAG,GAAG,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAE9B,IAAI,IAAI,EAAE,CAAC;YACT,KAAK,CAAC,IAAI,CAAC;gBACT,GAAG,EAAE,IAAI;gBACT,IAAI,EAAE,GAAG,CAAC,IAAI,EAAE,CAAC,IAAI,EAAE,IAAI,SAAS;gBACpC,KAAK,EAAE,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,SAAS;aACtC,CAAC,CAAC;QACL,CAAC;IACH,CAAC,CAAC,CAAC;IAEH,OAAO,KAAK,CAAC;AACf,CAAC;AAKD,SAAS,aAAa,CAAC,CAAqB;IAC1C,MAAM,MAAM,GAA8B,EAAE,CAAC;IAE7C,CAAC,CAAC,UAAU,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,EAAE;QAChC,MAAM,GAAG,GAAG,CAAC,CAAC,OAAO,CAAC,CAAC;QACvB,MAAM,GAAG,GAAG,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAE5B,IAAI,GAAG,EAAE,CAAC;YACR,MAAM,CAAC,IAAI,CAAC;gBACV,GAAG;gBACH,GAAG,EAAE,GAAG,CAAC,IAAI,CAAC,KAAK,CAAC,IAAI,SAAS;gBACjC,KAAK,EAAE,GAAG,CAAC,IAAI,CAAC,OAAO,CAAC,IAAI,SAAS;aACtC,CAAC,CAAC;QACL,CAAC;IACH,CAAC,CAAC,CAAC;IAEH,OAAO,MAAM,CAAC;AAChB,CAAC"}

package/dist/parsers/markdown-parser.d.ts

@@ -0,0 +1,13 @@
+export interface MarkdownParseResult {
+    text: string;
+    frontmatter: Record<string, any> | null;
+    links: Array<{
+        url: string;
+        title?: string;
+        text?: string;
+    }>;
+    rawFrontmatter?: string;
+}
+export declare function parseMarkdown(content: string): Promise<MarkdownParseResult>;
+export declare function isValidMarkdown(content: string): boolean;
+//# sourceMappingURL=markdown-parser.d.ts.map

package/dist/parsers/markdown-parser.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"markdown-parser.d.ts","sourceRoot":"","sources":["../../src/parsers/markdown-parser.ts"],"names":[],"mappings":"AAkBA,MAAM,WAAW,mBAAmB;IAElC,IAAI,EAAE,MAAM,CAAC;IAEb,WAAW,EAAE,MAAM,CAAC,MAAM,EAAE,GAAG,CAAC,GAAG,IAAI,CAAC;IAExC,KAAK,EAAE,KAAK,CAAC;QACX,GAAG,EAAE,MAAM,CAAC;QACZ,KAAK,CAAC,EAAE,MAAM,CAAC;QACf,IAAI,CAAC,EAAE,MAAM,CAAC;KACf,CAAC,CAAC;IAEH,cAAc,CAAC,EAAE,MAAM,CAAC;CACzB;AAQD,wBAAsB,aAAa,CAAC,OAAO,EAAE,MAAM,GAAG,OAAO,CAAC,mBAAmB,CAAC,CA4CjF;AAqED,wBAAgB,eAAe,CAAC,OAAO,EAAE,MAAM,GAAG,OAAO,CAQxD"}

package/dist/parsers/markdown-parser.js

@@ -0,0 +1,84 @@
+"use strict";
+var __importDefault = (this && this.__importDefault) || function (mod) {
+    return (mod && mod.__esModule) ? mod : { "default": mod };
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.isValidMarkdown = exports.parseMarkdown = void 0;
+const unified_1 = require("unified");
+const remark_parse_1 = __importDefault(require("remark-parse"));
+const remark_frontmatter_1 = __importDefault(require("remark-frontmatter"));
+const remark_stringify_1 = __importDefault(require("remark-stringify"));
+const gray_matter_1 = __importDefault(require("gray-matter"));
+const unist_util_visit_1 = require("unist-util-visit");
+async function parseMarkdown(content) {
+    try {
+        const { data: frontmatter, content: markdownContent, matter: rawFrontmatter } = (0, gray_matter_1.default)(content);
+        const processor = (0, unified_1.unified)()
+            .use(remark_parse_1.default)
+            .use(remark_frontmatter_1.default, ['yaml', 'toml'])
+            .use(remark_stringify_1.default);
+        const ast = processor.parse(markdownContent);
+        const links = [];
+        (0, unist_util_visit_1.visit)(ast, 'link', (node) => {
+            links.push({
+                url: node.url,
+                title: node.title || undefined,
+                text: extractTextFromNode(node),
+            });
+        });
+        const text = extractPlainText(markdownContent);
+        return {
+            text,
+            frontmatter: Object.keys(frontmatter).length > 0 ? frontmatter : null,
+            links,
+            rawFrontmatter: rawFrontmatter || undefined,
+        };
+    }
+    catch (error) {
+        console.warn('Markdown parsing error:', error);
+        return {
+            text: content,
+            frontmatter: null,
+            links: [],
+        };
+    }
+}
+exports.parseMarkdown = parseMarkdown;
+function extractPlainText(markdown) {
+    let text = markdown;
+    text = text.replace(/```[\s\S]*?```/g, '');
+    text = text.replace(/`[^`]+`/g, '');
+    text = text.replace(/^#{1,6}\s+/gm, '');
+    text = text.replace(/(\*\*|__)(.*?)\1/g, '$2');
+    text = text.replace(/(\*|_)(.*?)\1/g, '$2');
+    text = text.replace(/\[([^\]]+)\]\([^)]+\)/g, '$1');
+    text = text.replace(/!\[([^\]]*)\]\([^)]+\)/g, '');
+    text = text.replace(/^>\s+/gm, '');
+    text = text.replace(/^(-{3,}|\*{3,}|_{3,})$/gm, '');
+    text = text.replace(/^[\s]*[-*+]\s+/gm, '');
+    text = text.replace(/^[\s]*\d+\.\s+/gm, '');
+    text = text.replace(/\n{3,}/g, '\n\n');
+    text = text.trim();
+    return text;
+}
+function extractTextFromNode(node) {
+    if (node.type === 'text') {
+        return node.value;
+    }
+    if (node.children) {
+        return node.children.map(extractTextFromNode).join('');
+    }
+    return '';
+}
+function isValidMarkdown(content) {
+    try {
+        const processor = (0, unified_1.unified)().use(remark_parse_1.default);
+        processor.parse(content);
+        return true;
+    }
+    catch {
+        return false;
+    }
+}
+exports.isValidMarkdown = isValidMarkdown;
+//# sourceMappingURL=markdown-parser.js.map

package/dist/parsers/markdown-parser.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"markdown-parser.js","sourceRoot":"","sources":["../../src/parsers/markdown-parser.ts"],"names":[],"mappings":";;;;;;AAUA,qCAAkC;AAClC,gEAAuC;AACvC,4EAAmD;AACnD,wEAA+C;AAC/C,8DAAiC;AACjC,uDAAyC;AAwBlC,KAAK,UAAU,aAAa,CAAC,OAAe;IACjD,IAAI,CAAC;QAEH,MAAM,EAAE,IAAI,EAAE,WAAW,EAAE,OAAO,EAAE,eAAe,EAAE,MAAM,EAAE,cAAc,EAAE,GAAG,IAAA,qBAAM,EAAC,OAAO,CAAC,CAAC;QAGhG,MAAM,SAAS,GAAG,IAAA,iBAAO,GAAE;aACxB,GAAG,CAAC,sBAAW,CAAC;aAChB,GAAG,CAAC,4BAAiB,EAAE,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;aACxC,GAAG,CAAC,0BAAe,CAAC,CAAC;QAExB,MAAM,GAAG,GAAG,SAAS,CAAC,KAAK,CAAC,eAAe,CAAS,CAAC;QAGrD,MAAM,KAAK,GAA0D,EAAE,CAAC;QAExE,IAAA,wBAAK,EAAC,GAAG,EAAE,MAAM,EAAE,CAAC,IAAU,EAAE,EAAE;YAChC,KAAK,CAAC,IAAI,CAAC;gBACT,GAAG,EAAE,IAAI,CAAC,GAAG;gBACb,KAAK,EAAE,IAAI,CAAC,KAAK,IAAI,SAAS;gBAC9B,IAAI,EAAE,mBAAmB,CAAC,IAAI,CAAC;aAChC,CAAC,CAAC;QACL,CAAC,CAAC,CAAC;QAGH,MAAM,IAAI,GAAG,gBAAgB,CAAC,eAAe,CAAC,CAAC;QAE/C,OAAO;YACL,IAAI;YACJ,WAAW,EAAE,MAAM,CAAC,IAAI,CAAC,WAAW,CAAC,CAAC,MAAM,GAAG,CAAC,CAAC,CAAC,CAAC,WAAW,CAAC,CAAC,CAAC,IAAI;YACrE,KAAK;YACL,cAAc,EAAE,cAAc,IAAI,SAAS;SAC5C,CAAC;IACJ,CAAC;IAAC,OAAO,KAAK,EAAE,CAAC;QAEf,OAAO,CAAC,IAAI,CAAC,yBAAyB,EAAE,KAAK,CAAC,CAAC;QAG/C,OAAO;YACL,IAAI,EAAE,OAAO;YACb,WAAW,EAAE,IAAI;YACjB,KAAK,EAAE,EAAE;SACV,CAAC;IACJ,CAAC;AACH,CAAC;AA5CD,sCA4CC;AAQD,SAAS,gBAAgB,CAAC,QAAgB;IACxC,IAAI,IAAI,GAAG,QAAQ,CAAC;IAGpB,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,iBAAiB,EAAE,EAAE,CAAC,CAAC;IAC3C,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,UAAU,EAAE,EAAE,CAAC,CAAC;IAGpC,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,cAAc,EAAE,EAAE,CAAC,CAAC;IAGxC,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,mBAAmB,EAAE,IAAI,CAAC,CAAC;IAC/C,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,gBAAgB,EAAE,IAAI,CAAC,CAAC;IAG5C,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,wBAAwB,EAAE,IAAI,CAAC,CAAC;IAGpD,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,yBAAyB,EAAE,EAAE,CAAC,CAAC;IAGnD,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,SAAS,EAAE,EAAE,CAAC,CAAC;IAGnC,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,0BAA0B,EAAE,EAAE,CAAC,CAAC;IAGpD,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,kBAAkB,EAAE,EAAE,CAAC,CAAC;IAC5C,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,kBAAkB,EAAE,EAAE,CAAC,CAAC;IAG5C,IAAI,GAAG,IAAI,CAAC,OAAO,CAAC,SAAS,EAAE,MAAM,CAAC,CAAC;IACvC,IAAI,GAAG,IAAI,CAAC,IAAI,EAAE,CAAC;IAEnB,OAAO,IAAI,CAAC;AACd,CAAC;AAQD,SAAS,mBAAmB,CAAC,IAAS;IACpC,IAAI,IAAI,CAAC,IAAI,KAAK,MAAM,EAAE,CAAC;QACzB,OAAO,IAAI,CAAC,KAAK,CAAC;IACpB,CAAC;IAED,IAAI,IAAI,CAAC,QAAQ,EAAE,CAAC;QAClB,OAAO,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,mBAAmB,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IACzD,CAAC;IAED,OAAO,EAAE,CAAC;AACZ,CAAC;AAQD,SAAgB,eAAe,CAAC,OAAe;IAC7C,IAAI,CAAC;QACH,MAAM,SAAS,GAAG,IAAA,iBAAO,GAAE,CAAC,GAAG,CAAC,sBAAW,CAAC,CAAC;QAC7C,SAAS,CAAC,KAAK,CAAC,OAAO,CAAC,CAAC;QACzB,OAAO,IAAI,CAAC;IACd,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AARD,0CAQC"}

package/dist/parsers/text-parser.d.ts

@@ -0,0 +1,18 @@
+export interface TextParseResult {
+    text: string;
+    urls: string[];
+    lineCount: number;
+    charCount: number;
+}
+export declare function parseText(content: string): TextParseResult;
+export declare function normalizeLineEndings(text: string): string;
+export declare function extractUrls(text: string): string[];
+export declare function isPlainText(content: string): boolean;
+export declare function getTextStats(text: string): {
+    lines: number;
+    words: number;
+    chars: number;
+    urls: number;
+};
+export declare function truncateText(text: string, maxLength: number, suffix?: string): string;
+//# sourceMappingURL=text-parser.d.ts.map

package/dist/parsers/text-parser.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"text-parser.d.ts","sourceRoot":"","sources":["../../src/parsers/text-parser.ts"],"names":[],"mappings":"AAWA,MAAM,WAAW,eAAe;IAE9B,IAAI,EAAE,MAAM,CAAC;IAEb,IAAI,EAAE,MAAM,EAAE,CAAC;IAEf,SAAS,EAAE,MAAM,CAAC;IAElB,SAAS,EAAE,MAAM,CAAC;CACnB;AAcD,wBAAgB,SAAS,CAAC,OAAO,EAAE,MAAM,GAAG,eAAe,CAgC1D;AASD,wBAAgB,oBAAoB,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,CAIzD;AAQD,wBAAgB,WAAW,CAAC,IAAI,EAAE,MAAM,GAAG,MAAM,EAAE,CAUlD;AAQD,wBAAgB,WAAW,CAAC,OAAO,EAAE,MAAM,GAAG,OAAO,CAkBpD;AAQD,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,GAAG;IAC1C,KAAK,EAAE,MAAM,CAAC;IACd,KAAK,EAAE,MAAM,CAAC;IACd,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;CACd,CAYA;AAUD,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,EAAE,MAAM,GAAE,MAAc,GAAG,MAAM,CAM5F"}