@markitdownjs/shared 0.1.0

package/dist/__tests__/types.test.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=types.test.d.ts.map

package/dist/__tests__/types.test.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.test.d.ts","sourceRoot":"","sources":["../../src/__tests__/types.test.ts"],"names":[],"mappings":""}

package/dist/__tests__/types.test.js

@@ -0,0 +1,115 @@
+import { describe, it, expect } from 'vitest';
+import { createNode, isNodeOfType, walkAst, getNodeText, countTokens, findNodesOfType, detectMimeType, isSupportedExtension, generateId, readInputData, } from '../index.js';
+describe('Document AST', () => {
+    describe('createNode', () => {
+        it('should create a text node', () => {
+            const node = createNode({ type: 'text', value: 'hello' });
+            expect(node.type).toBe('text');
+            expect(node.value).toBe('hello');
+        });
+        it('should create a heading node', () => {
+            const node = createNode({
+                type: 'heading',
+                level: 1,
+                children: [createNode({ type: 'text', value: 'Title' })],
+            });
+            expect(node.type).toBe('heading');
+            expect(node.level).toBe(1);
+        });
+    });
+    describe('isNodeOfType', () => {
+        it('should correctly identify node types', () => {
+            const text = createNode({ type: 'text', value: 'test' });
+            expect(isNodeOfType(text, 'text')).toBe(true);
+            expect(isNodeOfType(text, 'heading')).toBe(false);
+        });
+    });
+    describe('walkAst', () => {
+        it('should visit all nodes', () => {
+            const doc = createNode({
+                type: 'document',
+                children: [
+                    createNode({
+                        type: 'paragraph',
+                        children: [createNode({ type: 'text', value: 'hello' })],
+                    }),
+                ],
+            });
+            const visited = [];
+            walkAst(doc, (node) => { visited.push(node.type); });
+            expect(visited).toContain('document');
+            expect(visited).toContain('paragraph');
+            expect(visited).toContain('text');
+        });
+    });
+    describe('getNodeText', () => {
+        it('should extract text from text node', () => {
+            const node = createNode({ type: 'text', value: 'hello world' });
+            expect(getNodeText(node)).toBe('hello world');
+        });
+        it('should extract text from nested nodes', () => {
+            const doc = createNode({
+                type: 'document',
+                children: [
+                    createNode({
+                        type: 'paragraph',
+                        children: [
+                            createNode({ type: 'text', value: 'hello ' }),
+                            createNode({ type: 'text', value: 'world' }),
+                        ],
+                    }),
+                ],
+            });
+            expect(getNodeText(doc)).toBe('hello world');
+        });
+    });
+    describe('countTokens', () => {
+        it('should count words', () => {
+            const node = createNode({ type: 'text', value: 'hello world foo bar' });
+            expect(countTokens(node)).toBe(4);
+        });
+    });
+    describe('findNodesOfType', () => {
+        it('should find all headings', () => {
+            const doc = createNode({
+                type: 'document',
+                children: [
+                    createNode({ type: 'heading', level: 1, children: [] }),
+                    createNode({ type: 'paragraph', children: [] }),
+                    createNode({ type: 'heading', level: 2, children: [] }),
+                ],
+            });
+            const headings = findNodesOfType(doc, 'heading');
+            expect(headings).toHaveLength(2);
+        });
+    });
+});
+describe('MIME types', () => {
+    it('should detect PDF mime type', () => {
+        expect(detectMimeType('document.pdf')).toBe('application/pdf');
+    });
+    it('should detect DOCX mime type', () => {
+        expect(detectMimeType('report.docx')).toContain('wordprocessingml');
+    });
+    it('should support .html extension', () => {
+        expect(isSupportedExtension('.html')).toBe(true);
+    });
+});
+describe('Utilities', () => {
+    it('should generate unique IDs', () => {
+        const id1 = generateId();
+        const id2 = generateId();
+        expect(id1).not.toBe(id2);
+    });
+    it('should read string input', async () => {
+        const data = await readInputData('hello');
+        expect(data).toBeInstanceOf(Uint8Array);
+        expect(new TextDecoder().decode(data)).toBe('hello');
+    });
+    it('should read Uint8Array input', async () => {
+        const input = new Uint8Array([1, 2, 3]);
+        const data = await readInputData(input);
+        expect(data).toBe(input);
+    });
+});
+//# sourceMappingURL=types.test.js.map

package/dist/__tests__/types.test.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.test.js","sourceRoot":"","sources":["../../src/__tests__/types.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,EAAE,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;AAC9C,OAAO,EACL,UAAU,EACV,YAAY,EACZ,OAAO,EACP,WAAW,EACX,WAAW,EACX,eAAe,EACf,cAAc,EACd,oBAAoB,EACpB,UAAU,EACV,aAAa,GACd,MAAM,aAAa,CAAC;AAGrB,QAAQ,CAAC,cAAc,EAAE,GAAG,EAAE;IAC5B,QAAQ,CAAC,YAAY,EAAE,GAAG,EAAE;QAC1B,EAAE,CAAC,2BAA2B,EAAE,GAAG,EAAE;YACnC,MAAM,IAAI,GAAG,UAAU,CAAW,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,CAAC,CAAC;YACpE,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;YAC/B,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;QACnC,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,8BAA8B,EAAE,GAAG,EAAE;YACtC,MAAM,IAAI,GAAG,UAAU,CAAc;gBACnC,IAAI,EAAE,SAAS;gBACf,KAAK,EAAE,CAAC;gBACR,QAAQ,EAAE,CAAC,UAAU,CAAW,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,CAAC,CAAC;aACnE,CAAC,CAAC;YACH,MAAM,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,SAAS,CAAC,CAAC;YAClC,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QAC7B,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,cAAc,EAAE,GAAG,EAAE;QAC5B,EAAE,CAAC,sCAAsC,EAAE,GAAG,EAAE;YAC9C,MAAM,IAAI,GAAG,UAAU,CAAW,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,MAAM,EAAE,CAAC,CAAC;YACnE,MAAM,CAAC,YAAY,CAAC,IAAI,EAAE,MAAM,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;YAC9C,MAAM,CAAC,YAAY,CAAC,IAAI,EAAE,SAAS,CAAC,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QACpD,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,SAAS,EAAE,GAAG,EAAE;QACvB,EAAE,CAAC,wBAAwB,EAAE,GAAG,EAAE;YAChC,MAAM,GAAG,GAAG,UAAU,CAAe;gBACnC,IAAI,EAAE,UAAU;gBAChB,QAAQ,EAAE;oBACR,UAAU,CAAgB;wBACxB,IAAI,EAAE,WAAW;wBACjB,QAAQ,EAAE,CAAC,UAAU,CAAW,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,CAAC,CAAC;qBACnE,CAAC;iBACH;aACF,CAAC,CAAC;YACH,MAAM,OAAO,GAAa,EAAE,CAAC;YAC7B,OAAO,CAAC,GAAG,EAAE,CAAC,IAAI,EAAE,EAAE,GAAG,OAAO,CAAC,IAAI,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;YACrD,MAAM,CAAC,OAAO,CAAC,CAAC,SAAS,CAAC,UAAU,CAAC,CAAC;YACtC,MAAM,CAAC,OAAO,CAAC,CAAC,SAAS,CAAC,WAAW,CAAC,CAAC;YACvC,MAAM,CAAC,OAAO,CAAC,CAAC,SAAS,CAAC,MAAM,CAAC,CAAC;QACpC,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,aAAa,EAAE,GAAG,EAAE;QAC3B,EAAE,CAAC,oCAAoC,EAAE,GAAG,EAAE;YAC5C,MAAM,IAAI,GAAG,UAAU,CAAW,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,aAAa,EAAE,CAAC,CAAC;YAC1E,MAAM,CAAC,WAAW,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,aAAa,CAAC,CAAC;QAChD,CAAC,CAAC,CAAC;QAEH,EAAE,CAAC,uCAAuC,EAAE,GAAG,EAAE;YAC/C,MAAM,GAAG,GAAG,UAAU,CAAe;gBACnC,IAAI,EAAE,UAAU;gBAChB,QAAQ,EAAE;oBACR,UAAU,CAAgB;wBACxB,IAAI,EAAE,WAAW;wBACjB,QAAQ,EAAE;4BACR,UAAU,CAAW,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,QAAQ,EAAE,CAAC;4BACvD,UAAU,CAAW,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,OAAO,EAAE,CAAC;yBACvD;qBACF,CAAC;iBACH;aACF,CAAC,CAAC;YACH,MAAM,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC,CAAC,IAAI,CAAC,aAAa,CAAC,CAAC;QAC/C,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,aAAa,EAAE,GAAG,EAAE;QAC3B,EAAE,CAAC,oBAAoB,EAAE,GAAG,EAAE;YAC5B,MAAM,IAAI,GAAG,UAAU,CAAW,EAAE,IAAI,EAAE,MAAM,EAAE,KAAK,EAAE,qBAAqB,EAAE,CAAC,CAAC;YAClF,MAAM,CAAC,WAAW,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QACpC,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;IAEH,QAAQ,CAAC,iBAAiB,EAAE,GAAG,EAAE;QAC/B,EAAE,CAAC,0BAA0B,EAAE,GAAG,EAAE;YAClC,MAAM,GAAG,GAAG,UAAU,CAAe;gBACnC,IAAI,EAAE,UAAU;gBAChB,QAAQ,EAAE;oBACR,UAAU,CAAc,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,QAAQ,EAAE,EAAE,EAAE,CAAC;oBACpE,UAAU,CAAgB,EAAE,IAAI,EAAE,WAAW,EAAE,QAAQ,EAAE,EAAE,EAAE,CAAC;oBAC9D,UAAU,CAAc,EAAE,IAAI,EAAE,SAAS,EAAE,KAAK,EAAE,CAAC,EAAE,QAAQ,EAAE,EAAE,EAAE,CAAC;iBACrE;aACF,CAAC,CAAC;YACH,MAAM,QAAQ,GAAG,eAAe,CAAC,GAAG,EAAE,SAAS,CAAC,CAAC;YACjD,MAAM,CAAC,QAAQ,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QACnC,CAAC,CAAC,CAAC;IACL,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,YAAY,EAAE,GAAG,EAAE;IAC1B,EAAE,CAAC,6BAA6B,EAAE,GAAG,EAAE;QACrC,MAAM,CAAC,cAAc,CAAC,cAAc,CAAC,CAAC,CAAC,IAAI,CAAC,iBAAiB,CAAC,CAAC;IACjE,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,8BAA8B,EAAE,GAAG,EAAE;QACtC,MAAM,CAAC,cAAc,CAAC,aAAa,CAAC,CAAC,CAAC,SAAS,CAAC,kBAAkB,CAAC,CAAC;IACtE,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,gCAAgC,EAAE,GAAG,EAAE;QACxC,MAAM,CAAC,oBAAoB,CAAC,OAAO,CAAC,CAAC,CAAC,IAAI,CAAC,IAAI,CAAC,CAAC;IACnD,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC;AAEH,QAAQ,CAAC,WAAW,EAAE,GAAG,EAAE;IACzB,EAAE,CAAC,4BAA4B,EAAE,GAAG,EAAE;QACpC,MAAM,GAAG,GAAG,UAAU,EAAE,CAAC;QACzB,MAAM,GAAG,GAAG,UAAU,EAAE,CAAC;QACzB,MAAM,CAAC,GAAG,CAAC,CAAC,GAAG,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IAC5B,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,0BAA0B,EAAE,KAAK,IAAI,EAAE;QACxC,MAAM,IAAI,GAAG,MAAM,aAAa,CAAC,OAAO,CAAC,CAAC;QAC1C,MAAM,CAAC,IAAI,CAAC,CAAC,cAAc,CAAC,UAAU,CAAC,CAAC;QACxC,MAAM,CAAC,IAAI,WAAW,EAAE,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IACvD,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,8BAA8B,EAAE,KAAK,IAAI,EAAE;QAC5C,MAAM,KAAK,GAAG,IAAI,UAAU,CAAC,CAAC,CAAC,EAAE,CAAC,EAAE,CAAC,CAAC,CAAC,CAAC;QACxC,MAAM,IAAI,GAAG,MAAM,aAAa,CAAC,KAAK,CAAC,CAAC;QACxC,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;IAC3B,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}

package/dist/document-ast.d.ts

@@ -0,0 +1,155 @@
+export type NodeType = "document" | "heading" | "paragraph" | "table" | "table-row" | "table-cell" | "list" | "list-item" | "image" | "link" | "code" | "quote" | "blockquote" | "thematic-break" | "html" | "text" | "emphasis" | "strong" | "inline-code" | "definition" | "footnote" | "page-break" | "section" | "strikethrough" | "subscript" | "superscript" | "math" | "horizontal-rule" | "raw";
+export interface DocumentNode {
+    type: NodeType;
+    children?: AnyNode[];
+    position?: NodePosition;
+    [key: string]: unknown;
+}
+export interface NodePosition {
+    start: {
+        line: number;
+        column: number;
+        offset: number;
+    };
+    end: {
+        line: number;
+        column: number;
+        offset: number;
+    };
+    sourceFile?: string;
+    pageIndex?: number;
+}
+export interface HeadingNode extends DocumentNode {
+    type: "heading";
+    level: 1 | 2 | 3 | 4 | 5 | 6;
+    children: AnyNode[];
+    id?: string;
+}
+export interface ParagraphNode extends DocumentNode {
+    type: "paragraph";
+    children: AnyNode[];
+}
+export interface TableNode extends DocumentNode {
+    type: "table";
+    children: TableRowNode[];
+    caption?: string;
+}
+export interface TableRowNode extends DocumentNode {
+    type: "table-row";
+    children: TableCellNode[];
+    isHeader?: boolean;
+}
+export interface TableCellNode extends DocumentNode {
+    type: "table-cell";
+    children: AnyNode[];
+    colspan?: number;
+    rowspan?: number;
+}
+export interface ListNode extends DocumentNode {
+    type: "list";
+    ordered: boolean;
+    start?: number;
+    children: ListItemNode[];
+}
+export interface ListItemNode extends DocumentNode {
+    type: "list-item";
+    children: AnyNode[];
+}
+export interface ImageNode extends DocumentNode {
+    type: "image";
+    src: string;
+    alt?: string;
+    title?: string;
+    data?: Uint8Array;
+    mimeType?: string;
+    width?: number;
+    height?: number;
+}
+export interface LinkNode extends DocumentNode {
+    type: "link";
+    href: string;
+    title?: string;
+    children: AnyNode[];
+}
+export interface CodeNode extends DocumentNode {
+    type: "code";
+    language?: string;
+    value: string;
+}
+export interface QuoteNode extends DocumentNode {
+    type: "quote";
+    children: AnyNode[];
+}
+export interface BlockquoteNode extends DocumentNode {
+    type: "blockquote";
+    children: AnyNode[];
+}
+export interface TextNode extends DocumentNode {
+    type: "text";
+    value: string;
+}
+export interface EmphasisNode extends DocumentNode {
+    type: "emphasis";
+    children: AnyNode[];
+}
+export interface StrongNode extends DocumentNode {
+    type: "strong";
+    children: AnyNode[];
+}
+export interface InlineCodeNode extends DocumentNode {
+    type: "inline-code";
+    value: string;
+}
+export interface ThematicBreakNode extends DocumentNode {
+    type: "thematic-break";
+}
+export interface HtmlNode extends DocumentNode {
+    type: "html";
+    value: string;
+}
+export interface DefinitionNode extends DocumentNode {
+    type: "definition";
+    label: string;
+    url: string;
+    title?: string;
+}
+export interface FootnoteNode extends DocumentNode {
+    type: "footnote";
+    identifier: string;
+    children: AnyNode[];
+}
+export interface PageBreakNode extends DocumentNode {
+    type: "page-break";
+    pageNumber?: number;
+}
+export interface SectionNode extends DocumentNode {
+    type: "section";
+    title?: string;
+    level?: number;
+    children: AnyNode[];
+}
+export interface StrikethroughNode extends DocumentNode {
+    type: "strikethrough";
+    children: AnyNode[];
+}
+export interface MathNode extends DocumentNode {
+    type: "math";
+    value: string;
+    inline?: boolean;
+}
+export interface HorizontalRuleNode extends DocumentNode {
+    type: "horizontal-rule";
+}
+export interface RawNode extends DocumentNode {
+    type: "raw";
+    format: string;
+    value: string;
+}
+export type AnyNode = DocumentNode | HeadingNode | ParagraphNode | TableNode | TableRowNode | TableCellNode | ListNode | ListItemNode | ImageNode | LinkNode | CodeNode | QuoteNode | BlockquoteNode | TextNode | EmphasisNode | StrongNode | InlineCodeNode | ThematicBreakNode | HtmlNode | DefinitionNode | FootnoteNode | PageBreakNode | SectionNode | StrikethroughNode | MathNode | HorizontalRuleNode | RawNode;
+export declare function createNode<T extends AnyNode>(node: T): T;
+export declare function isNodeOfType<T extends AnyNode>(node: AnyNode, type: T["type"]): node is T;
+export declare function walkAst(node: AnyNode, visitor: (node: AnyNode, parent: AnyNode | null) => boolean | void): void;
+export declare function getNodeText(node: AnyNode): string;
+export declare function countTokens(node: AnyNode): number;
+export declare function findNodesOfType<T extends AnyNode>(root: AnyNode, type: T["type"]): T[];
+//# sourceMappingURL=document-ast.d.ts.map

package/dist/document-ast.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"document-ast.d.ts","sourceRoot":"","sources":["../src/document-ast.ts"],"names":[],"mappings":"AAAA,MAAM,MAAM,QAAQ,GAChB,UAAU,GACV,SAAS,GACT,WAAW,GACX,OAAO,GACP,WAAW,GACX,YAAY,GACZ,MAAM,GACN,WAAW,GACX,OAAO,GACP,MAAM,GACN,MAAM,GACN,OAAO,GACP,YAAY,GACZ,gBAAgB,GAChB,MAAM,GACN,MAAM,GACN,UAAU,GACV,QAAQ,GACR,aAAa,GACb,YAAY,GACZ,UAAU,GACV,YAAY,GACZ,SAAS,GACT,eAAe,GACf,WAAW,GACX,aAAa,GACb,MAAM,GACN,iBAAiB,GACjB,KAAK,CAAC;AAEV,MAAM,WAAW,YAAY;IAC3B,IAAI,EAAE,QAAQ,CAAC;IACf,QAAQ,CAAC,EAAE,OAAO,EAAE,CAAC;IACrB,QAAQ,CAAC,EAAE,YAAY,CAAC;IACxB,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC;CACxB;AAED,MAAM,WAAW,YAAY;IAC3B,KAAK,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAA;KAAE,CAAC;IACxD,GAAG,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAC;QAAC,MAAM,EAAE,MAAM,CAAA;KAAE,CAAC;IACtD,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,SAAS,CAAC,EAAE,MAAM,CAAC;CACpB;AAED,MAAM,WAAW,WAAY,SAAQ,YAAY;IAC/C,IAAI,EAAE,SAAS,CAAC;IAChB,KAAK,EAAE,CAAC,GAAG,CAAC,GAAG,CAAC,GAAG,CAAC,GAAG,CAAC,GAAG,CAAC,CAAC;IAC7B,QAAQ,EAAE,OAAO,EAAE,CAAC;IACpB,EAAE,CAAC,EAAE,MAAM,CAAC;CACb;AAED,MAAM,WAAW,aAAc,SAAQ,YAAY;IACjD,IAAI,EAAE,WAAW,CAAC;IAClB,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,SAAU,SAAQ,YAAY;IAC7C,IAAI,EAAE,OAAO,CAAC;IACd,QAAQ,EAAE,YAAY,EAAE,CAAC;IACzB,OAAO,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,MAAM,WAAW,YAAa,SAAQ,YAAY;IAChD,IAAI,EAAE,WAAW,CAAC;IAClB,QAAQ,EAAE,aAAa,EAAE,CAAC;IAC1B,QAAQ,CAAC,EAAE,OAAO,CAAC;CACpB;AAED,MAAM,WAAW,aAAc,SAAQ,YAAY;IACjD,IAAI,EAAE,YAAY,CAAC;IACnB,QAAQ,EAAE,OAAO,EAAE,CAAC;IACpB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,OAAO,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,MAAM,WAAW,QAAS,SAAQ,YAAY;IAC5C,IAAI,EAAE,MAAM,CAAC;IACb,OAAO,EAAE,OAAO,CAAC;IACjB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,YAAY,EAAE,CAAC;CAC1B;AAED,MAAM,WAAW,YAAa,SAAQ,YAAY;IAChD,IAAI,EAAE,WAAW,CAAC;IAClB,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,SAAU,SAAQ,YAAY;IAC7C,IAAI,EAAE,OAAO,CAAC;IACd,GAAG,EAAE,MAAM,CAAC;IACZ,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,IAAI,CAAC,EAAE,UAAU,CAAC;IAClB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,QAAS,SAAQ,YAAY;IAC5C,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,MAAM,CAAC;IACb,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,QAAS,SAAQ,YAAY;IAC5C,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,KAAK,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,SAAU,SAAQ,YAAY;IAC7C,IAAI,EAAE,OAAO,CAAC;IACd,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,cAAe,SAAQ,YAAY;IAClD,IAAI,EAAE,YAAY,CAAC;IACnB,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,QAAS,SAAQ,YAAY;IAC5C,IAAI,EAAE,MAAM,CAAC;IACb,KAAK,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,YAAa,SAAQ,YAAY;IAChD,IAAI,EAAE,UAAU,CAAC;IACjB,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,UAAW,SAAQ,YAAY;IAC9C,IAAI,EAAE,QAAQ,CAAC;IACf,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,cAAe,SAAQ,YAAY;IAClD,IAAI,EAAE,aAAa,CAAC;IACpB,KAAK,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,iBAAkB,SAAQ,YAAY;IACrD,IAAI,EAAE,gBAAgB,CAAC;CACxB;AAED,MAAM,WAAW,QAAS,SAAQ,YAAY;IAC5C,IAAI,EAAE,MAAM,CAAC;IACb,KAAK,EAAE,MAAM,CAAC;CACf;AAED,MAAM,WAAW,cAAe,SAAQ,YAAY;IAClD,IAAI,EAAE,YAAY,CAAC;IACnB,KAAK,EAAE,MAAM,CAAC;IACd,GAAG,EAAE,MAAM,CAAC;IACZ,KAAK,CAAC,EAAE,MAAM,CAAC;CAChB;AAED,MAAM,WAAW,YAAa,SAAQ,YAAY;IAChD,IAAI,EAAE,UAAU,CAAC;IACjB,UAAU,EAAE,MAAM,CAAC;IACnB,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,aAAc,SAAQ,YAAY;IACjD,IAAI,EAAE,YAAY,CAAC;IACnB,UAAU,CAAC,EAAE,MAAM,CAAC;CACrB;AAED,MAAM,WAAW,WAAY,SAAQ,YAAY;IAC/C,IAAI,EAAE,SAAS,CAAC;IAChB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,iBAAkB,SAAQ,YAAY;IACrD,IAAI,EAAE,eAAe,CAAC;IACtB,QAAQ,EAAE,OAAO,EAAE,CAAC;CACrB;AAED,MAAM,WAAW,QAAS,SAAQ,YAAY;IAC5C,IAAI,EAAE,MAAM,CAAC;IACb,KAAK,EAAE,MAAM,CAAC;IACd,MAAM,CAAC,EAAE,OAAO,CAAC;CAClB;AAED,MAAM,WAAW,kBAAmB,SAAQ,YAAY;IACtD,IAAI,EAAE,iBAAiB,CAAC;CACzB;AAED,MAAM,WAAW,OAAQ,SAAQ,YAAY;IAC3C,IAAI,EAAE,KAAK,CAAC;IACZ,MAAM,EAAE,MAAM,CAAC;IACf,KAAK,EAAE,MAAM,CAAC;CACf;AAED,MAAM,MAAM,OAAO,GACf,YAAY,GACZ,WAAW,GACX,aAAa,GACb,SAAS,GACT,YAAY,GACZ,aAAa,GACb,QAAQ,GACR,YAAY,GACZ,SAAS,GACT,QAAQ,GACR,QAAQ,GACR,SAAS,GACT,cAAc,GACd,QAAQ,GACR,YAAY,GACZ,UAAU,GACV,cAAc,GACd,iBAAiB,GACjB,QAAQ,GACR,cAAc,GACd,YAAY,GACZ,aAAa,GACb,WAAW,GACX,iBAAiB,GACjB,QAAQ,GACR,kBAAkB,GAClB,OAAO,CAAC;AAEZ,wBAAgB,UAAU,CAAC,CAAC,SAAS,OAAO,EAAE,IAAI,EAAE,CAAC,GAAG,CAAC,CAExD;AAED,wBAAgB,YAAY,CAAC,CAAC,SAAS,OAAO,EAC5C,IAAI,EAAE,OAAO,EACb,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,GACd,IAAI,IAAI,CAAC,CAEX;AAED,wBAAgB,OAAO,CACrB,IAAI,EAAE,OAAO,EACb,OAAO,EAAE,CAAC,IAAI,EAAE,OAAO,EAAE,MAAM,EAAE,OAAO,GAAG,IAAI,KAAK,OAAO,GAAG,IAAI,GACjE,IAAI,CAWN;AAED,wBAAgB,WAAW,CAAC,IAAI,EAAE,OAAO,GAAG,MAAM,CAWjD;AAED,wBAAgB,WAAW,CAAC,IAAI,EAAE,OAAO,GAAG,MAAM,CAGjD;AAED,wBAAgB,eAAe,CAAC,CAAC,SAAS,OAAO,EAC/C,IAAI,EAAE,OAAO,EACb,IAAI,EAAE,CAAC,CAAC,MAAM,CAAC,GACd,CAAC,EAAE,CAML"}

package/dist/document-ast.js

@@ -0,0 +1,50 @@
+export function createNode(node) {
+    return node;
+}
+export function isNodeOfType(node, type) {
+    return node.type === type;
+}
+export function walkAst(node, visitor) {
+    const walk = (n, parent) => {
+        const result = visitor(n, parent);
+        if (result === false)
+            return false;
+        if ("children" in n && Array.isArray(n.children)) {
+            for (const child of n.children) {
+                walk(child, n);
+            }
+        }
+    };
+    walk(node, null);
+}
+export function getNodeText(node) {
+    if (node.type === "text")
+        return node.value;
+    if (node.type === "code")
+        return node.value;
+    if (node.type === "inline-code")
+        return node.value;
+    if (node.type === "math")
+        return node.value;
+    if (node.type === "html")
+        return node.value;
+    if (node.type === "raw")
+        return node.value;
+    if ("children" in node && Array.isArray(node.children)) {
+        return node.children.map((c) => getNodeText(c)).join("");
+    }
+    return "";
+}
+export function countTokens(node) {
+    const text = getNodeText(node);
+    return text.split(/\s+/).filter(Boolean).length;
+}
+export function findNodesOfType(root, type) {
+    const results = [];
+    walkAst(root, (node) => {
+        if (node.type === type)
+            results.push(node);
+    });
+    return results;
+}
+//# sourceMappingURL=document-ast.js.map

package/dist/document-ast.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"document-ast.js","sourceRoot":"","sources":["../src/document-ast.ts"],"names":[],"mappings":"AAkOA,MAAM,UAAU,UAAU,CAAoB,IAAO;IACnD,OAAO,IAAI,CAAC;AACd,CAAC;AAED,MAAM,UAAU,YAAY,CAC1B,IAAa,EACb,IAAe;IAEf,OAAO,IAAI,CAAC,IAAI,KAAK,IAAI,CAAC;AAC5B,CAAC;AAED,MAAM,UAAU,OAAO,CACrB,IAAa,EACb,OAAkE;IAElE,MAAM,IAAI,GAAG,CAAC,CAAU,EAAE,MAAsB,EAAkB,EAAE;QAClE,MAAM,MAAM,GAAG,OAAO,CAAC,CAAC,EAAE,MAAM,CAAC,CAAC;QAClC,IAAI,MAAM,KAAK,KAAK;YAAE,OAAO,KAAK,CAAC;QACnC,IAAI,UAAU,IAAI,CAAC,IAAI,KAAK,CAAC,OAAO,CAAC,CAAC,CAAC,QAAQ,CAAC,EAAE,CAAC;YACjD,KAAK,MAAM,KAAK,IAAI,CAAC,CAAC,QAAQ,EAAE,CAAC;gBAC/B,IAAI,CAAC,KAAK,EAAE,CAAC,CAAC,CAAC;YACjB,CAAC;QACH,CAAC;IACH,CAAC,CAAC;IACF,IAAI,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC;AACnB,CAAC;AAED,MAAM,UAAU,WAAW,CAAC,IAAa;IACvC,IAAI,IAAI,CAAC,IAAI,KAAK,MAAM;QAAE,OAAQ,IAAiB,CAAC,KAAK,CAAC;IAC1D,IAAI,IAAI,CAAC,IAAI,KAAK,MAAM;QAAE,OAAQ,IAAiB,CAAC,KAAK,CAAC;IAC1D,IAAI,IAAI,CAAC,IAAI,KAAK,aAAa;QAAE,OAAQ,IAAuB,CAAC,KAAK,CAAC;IACvE,IAAI,IAAI,CAAC,IAAI,KAAK,MAAM;QAAE,OAAQ,IAAiB,CAAC,KAAK,CAAC;IAC1D,IAAI,IAAI,CAAC,IAAI,KAAK,MAAM;QAAE,OAAQ,IAAiB,CAAC,KAAK,CAAC;IAC1D,IAAI,IAAI,CAAC,IAAI,KAAK,KAAK;QAAE,OAAQ,IAAgB,CAAC,KAAK,CAAC;IACxD,IAAI,UAAU,IAAI,IAAI,IAAI,KAAK,CAAC,OAAO,CAAC,IAAI,CAAC,QAAQ,CAAC,EAAE,CAAC;QACvD,OAAO,IAAI,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,WAAW,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;IAC3D,CAAC;IACD,OAAO,EAAE,CAAC;AACZ,CAAC;AAED,MAAM,UAAU,WAAW,CAAC,IAAa;IACvC,MAAM,IAAI,GAAG,WAAW,CAAC,IAAI,CAAC,CAAC;IAC/B,OAAO,IAAI,CAAC,KAAK,CAAC,KAAK,CAAC,CAAC,MAAM,CAAC,OAAO,CAAC,CAAC,MAAM,CAAC;AAClD,CAAC;AAED,MAAM,UAAU,eAAe,CAC7B,IAAa,EACb,IAAe;IAEf,MAAM,OAAO,GAAQ,EAAE,CAAC;IACxB,OAAO,CAAC,IAAI,EAAE,CAAC,IAAI,EAAE,EAAE;QACrB,IAAI,IAAI,CAAC,IAAI,KAAK,IAAI;YAAE,OAAO,CAAC,IAAI,CAAC,IAAS,CAAC,CAAC;IAClD,CAAC,CAAC,CAAC;IACH,OAAO,OAAO,CAAC;AACjB,CAAC"}

package/dist/dom.d.ts

@@ -0,0 +1,4 @@
+export declare function parseHTML(html: string): Promise<Document>;
+export declare function parseXML(xml: string): Promise<Document>;
+export declare function serializeXML(doc: Document): string;
+//# sourceMappingURL=dom.d.ts.map

package/dist/dom.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"dom.d.ts","sourceRoot":"","sources":["../src/dom.ts"],"names":[],"mappings":"AAyBA,wBAAsB,SAAS,CAAC,IAAI,EAAE,MAAM,GAAG,OAAO,CAAC,QAAQ,CAAC,CAG/D;AAED,wBAAsB,QAAQ,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAC,QAAQ,CAAC,CAG7D;AAED,wBAAgB,YAAY,CAAC,GAAG,EAAE,QAAQ,GAAG,MAAM,CAKlD"}

package/dist/dom.js

@@ -0,0 +1,35 @@
+let domParserInstance;
+async function getDOMParser() {
+    if (domParserInstance)
+        return domParserInstance;
+    if (typeof globalThis.DOMParser !== "undefined") {
+        domParserInstance = globalThis.DOMParser;
+        return domParserInstance;
+    }
+    try {
+        const mod = await import("linkedom");
+        if (mod.DOMParser) {
+            domParserInstance = mod.DOMParser;
+            return domParserInstance;
+        }
+    }
+    catch {
+        // linkedom not available
+    }
+    throw new Error("DOMParser is not available. Install 'linkedom' for Node.js support: npm install linkedom");
+}
+export async function parseHTML(html) {
+    const DP = await getDOMParser();
+    return new DP().parseFromString(html, "text/html");
+}
+export async function parseXML(xml) {
+    const DP = await getDOMParser();
+    return new DP().parseFromString(xml, "application/xml");
+}
+export function serializeXML(doc) {
+    if (typeof XMLSerializer !== "undefined") {
+        return new XMLSerializer().serializeToString(doc);
+    }
+    return doc.documentElement?.outerHTML ?? "";
+}
+//# sourceMappingURL=dom.js.map

package/dist/dom.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"dom.js","sourceRoot":"","sources":["../src/dom.ts"],"names":[],"mappings":"AAAA,IAAI,iBAA0E,CAAC;AAE/E,KAAK,UAAU,YAAY;IACzB,IAAI,iBAAiB;QAAE,OAAO,iBAAiB,CAAC;IAEhD,IAAI,OAAO,UAAU,CAAC,SAAS,KAAK,WAAW,EAAE,CAAC;QAChD,iBAAiB,GAAG,UAAU,CAAC,SAAS,CAAC;QACzC,OAAO,iBAAiB,CAAC;IAC3B,CAAC;IAED,IAAI,CAAC;QACH,MAAM,GAAG,GAAG,MAAM,MAAM,CAAC,UAAU,CAAC,CAAC;QACrC,IAAI,GAAG,CAAC,SAAS,EAAE,CAAC;YAClB,iBAAiB,GAAG,GAAG,CAAC,SAAmE,CAAC;YAC5F,OAAO,iBAAiB,CAAC;QAC3B,CAAC;IACH,CAAC;IAAC,MAAM,CAAC;QACP,yBAAyB;IAC3B,CAAC;IAED,MAAM,IAAI,KAAK,CACb,0FAA0F,CAC3F,CAAC;AACJ,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,SAAS,CAAC,IAAY;IAC1C,MAAM,EAAE,GAAG,MAAM,YAAY,EAAE,CAAC;IAChC,OAAO,IAAI,EAAE,EAAE,CAAC,eAAe,CAAC,IAAI,EAAE,WAAW,CAAC,CAAC;AACrD,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,QAAQ,CAAC,GAAW;IACxC,MAAM,EAAE,GAAG,MAAM,YAAY,EAAE,CAAC;IAChC,OAAO,IAAI,EAAE,EAAE,CAAC,eAAe,CAAC,GAAG,EAAE,iBAAiB,CAAC,CAAC;AAC1D,CAAC;AAED,MAAM,UAAU,YAAY,CAAC,GAAa;IACxC,IAAI,OAAO,aAAa,KAAK,WAAW,EAAE,CAAC;QACzC,OAAO,IAAI,aAAa,EAAE,CAAC,iBAAiB,CAAC,GAAG,CAAC,CAAC;IACpD,CAAC;IACD,OAAO,GAAG,CAAC,eAAe,EAAE,SAAS,IAAI,EAAE,CAAC;AAC9C,CAAC"}

package/dist/errors.d.ts

@@ -0,0 +1,25 @@
+export declare class MarkItDownError extends Error {
+    readonly code: string;
+    readonly cause?: Error | undefined;
+    constructor(message: string, code: string, cause?: Error | undefined);
+}
+export declare class UnsupportedFormatError extends MarkItDownError {
+    constructor(format: string);
+}
+export declare class ConversionError extends MarkItDownError {
+    readonly converterId?: string | undefined;
+    constructor(message: string, converterId?: string | undefined, cause?: Error);
+}
+export declare class FileReadError extends MarkItDownError {
+    constructor(message: string, cause?: Error);
+}
+export declare class ParseError extends MarkItDownError {
+    constructor(message: string, cause?: Error);
+}
+export declare class PluginError extends MarkItDownError {
+    constructor(message: string, cause?: Error);
+}
+export declare class CancellationError extends MarkItDownError {
+    constructor();
+}
+//# sourceMappingURL=errors.d.ts.map

package/dist/errors.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"errors.d.ts","sourceRoot":"","sources":["../src/errors.ts"],"names":[],"mappings":"AAAA,qBAAa,eAAgB,SAAQ,KAAK;aAGtB,IAAI,EAAE,MAAM;aACZ,KAAK,CAAC,EAAE,KAAK;gBAF7B,OAAO,EAAE,MAAM,EACC,IAAI,EAAE,MAAM,EACZ,KAAK,CAAC,EAAE,KAAK,YAAA;CAKhC;AAED,qBAAa,sBAAuB,SAAQ,eAAe;gBAC7C,MAAM,EAAE,MAAM;CAI3B;AAED,qBAAa,eAAgB,SAAQ,eAAe;aAGhC,WAAW,CAAC,EAAE,MAAM;gBADpC,OAAO,EAAE,MAAM,EACC,WAAW,CAAC,EAAE,MAAM,YAAA,EACpC,KAAK,CAAC,EAAE,KAAK;CAKhB;AAED,qBAAa,aAAc,SAAQ,eAAe;gBACpC,OAAO,EAAE,MAAM,EAAE,KAAK,CAAC,EAAE,KAAK;CAI3C;AAED,qBAAa,UAAW,SAAQ,eAAe;gBACjC,OAAO,EAAE,MAAM,EAAE,KAAK,CAAC,EAAE,KAAK;CAI3C;AAED,qBAAa,WAAY,SAAQ,eAAe;gBAClC,OAAO,EAAE,MAAM,EAAE,KAAK,CAAC,EAAE,KAAK;CAI3C;AAED,qBAAa,iBAAkB,SAAQ,eAAe;;CAKrD"}

package/dist/errors.js

@@ -0,0 +1,49 @@
+export class MarkItDownError extends Error {
+    code;
+    cause;
+    constructor(message, code, cause) {
+        super(message);
+        this.code = code;
+        this.cause = cause;
+        this.name = "MarkItDownError";
+    }
+}
+export class UnsupportedFormatError extends MarkItDownError {
+    constructor(format) {
+        super(`Unsupported format: ${format}`, "UNSUPPORTED_FORMAT");
+        this.name = "UnsupportedFormatError";
+    }
+}
+export class ConversionError extends MarkItDownError {
+    converterId;
+    constructor(message, converterId, cause) {
+        super(message, "CONVERSION_ERROR", cause);
+        this.converterId = converterId;
+        this.name = "ConversionError";
+    }
+}
+export class FileReadError extends MarkItDownError {
+    constructor(message, cause) {
+        super(message, "FILE_READ_ERROR", cause);
+        this.name = "FileReadError";
+    }
+}
+export class ParseError extends MarkItDownError {
+    constructor(message, cause) {
+        super(message, "PARSE_ERROR", cause);
+        this.name = "ParseError";
+    }
+}
+export class PluginError extends MarkItDownError {
+    constructor(message, cause) {
+        super(message, "PLUGIN_ERROR", cause);
+        this.name = "PluginError";
+    }
+}
+export class CancellationError extends MarkItDownError {
+    constructor() {
+        super("Conversion was cancelled", "CANCELLED");
+        this.name = "CancellationError";
+    }
+}
+//# sourceMappingURL=errors.js.map

package/dist/errors.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"errors.js","sourceRoot":"","sources":["../src/errors.ts"],"names":[],"mappings":"AAAA,MAAM,OAAO,eAAgB,SAAQ,KAAK;IAGtB;IACA;IAHlB,YACE,OAAe,EACC,IAAY,EACZ,KAAa;QAE7B,KAAK,CAAC,OAAO,CAAC,CAAC;QAHC,SAAI,GAAJ,IAAI,CAAQ;QACZ,UAAK,GAAL,KAAK,CAAQ;QAG7B,IAAI,CAAC,IAAI,GAAG,iBAAiB,CAAC;IAChC,CAAC;CACF;AAED,MAAM,OAAO,sBAAuB,SAAQ,eAAe;IACzD,YAAY,MAAc;QACxB,KAAK,CAAC,uBAAuB,MAAM,EAAE,EAAE,oBAAoB,CAAC,CAAC;QAC7D,IAAI,CAAC,IAAI,GAAG,wBAAwB,CAAC;IACvC,CAAC;CACF;AAED,MAAM,OAAO,eAAgB,SAAQ,eAAe;IAGhC;IAFlB,YACE,OAAe,EACC,WAAoB,EACpC,KAAa;QAEb,KAAK,CAAC,OAAO,EAAE,kBAAkB,EAAE,KAAK,CAAC,CAAC;QAH1B,gBAAW,GAAX,WAAW,CAAS;QAIpC,IAAI,CAAC,IAAI,GAAG,iBAAiB,CAAC;IAChC,CAAC;CACF;AAED,MAAM,OAAO,aAAc,SAAQ,eAAe;IAChD,YAAY,OAAe,EAAE,KAAa;QACxC,KAAK,CAAC,OAAO,EAAE,iBAAiB,EAAE,KAAK,CAAC,CAAC;QACzC,IAAI,CAAC,IAAI,GAAG,eAAe,CAAC;IAC9B,CAAC;CACF;AAED,MAAM,OAAO,UAAW,SAAQ,eAAe;IAC7C,YAAY,OAAe,EAAE,KAAa;QACxC,KAAK,CAAC,OAAO,EAAE,aAAa,EAAE,KAAK,CAAC,CAAC;QACrC,IAAI,CAAC,IAAI,GAAG,YAAY,CAAC;IAC3B,CAAC;CACF;AAED,MAAM,OAAO,WAAY,SAAQ,eAAe;IAC9C,YAAY,OAAe,EAAE,KAAa;QACxC,KAAK,CAAC,OAAO,EAAE,cAAc,EAAE,KAAK,CAAC,CAAC;QACtC,IAAI,CAAC,IAAI,GAAG,aAAa,CAAC;IAC5B,CAAC;CACF;AAED,MAAM,OAAO,iBAAkB,SAAQ,eAAe;IACpD;QACE,KAAK,CAAC,0BAA0B,EAAE,WAAW,CAAC,CAAC;QAC/C,IAAI,CAAC,IAAI,GAAG,mBAAmB,CAAC;IAClC,CAAC;CACF"}

package/dist/index.d.ts

@@ -0,0 +1,8 @@
+export type { MimeType, FileExtension, ConverterId, ConversionInput, ConversionOptions, OutputFormat, ConversionResult, DocumentMetadata, AssetInfo, TableData, ImageInfo, HeadingInfo, ConversionStats, ProgressCallback, ProgressInfo, Converter, Plugin, PluginContext, ConversionMiddleware, ConverterRegistry, Logger, ChunkMetadata, DocumentChunk, ChunkingOptions, } from "./types.js";
+export type { NodeType, NodePosition, DocumentNode, HeadingNode, ParagraphNode, TableNode, TableRowNode, TableCellNode, ListNode, ListItemNode, ImageNode, LinkNode, CodeNode, QuoteNode, BlockquoteNode, TextNode, EmphasisNode, StrongNode, InlineCodeNode, ThematicBreakNode, HtmlNode, DefinitionNode, FootnoteNode, PageBreakNode, SectionNode, StrikethroughNode, MathNode, HorizontalRuleNode, RawNode, AnyNode, } from "./document-ast.js";
+export { createNode, isNodeOfType, walkAst, getNodeText, countTokens, findNodesOfType, } from "./document-ast.js";
+export { extensionToMime, mimeToExtension, detectMimeType, isSupportedExtension, isSupportedMimeType, getSupportedExtensions, getSupportedMimeTypes, } from "./mime-types.js";
+export { MarkItDownError, UnsupportedFormatError, ConversionError, FileReadError, ParseError, PluginError, CancellationError, } from "./errors.js";
+export { uint8ArrayToDataUrl, blobToUint8Array, readInputData, detectMimeTypeFromData, truncateText, generateId, mergeOptions, AbortError, checkSignal, } from "./utils.js";
+export { parseHTML, parseXML, serializeXML } from "./dom.js";
+//# sourceMappingURL=index.d.ts.map

package/dist/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,YAAY,EACV,QAAQ,EACR,aAAa,EACb,WAAW,EACX,eAAe,EACf,iBAAiB,EACjB,YAAY,EACZ,gBAAgB,EAChB,gBAAgB,EAChB,SAAS,EACT,SAAS,EACT,SAAS,EACT,WAAW,EACX,eAAe,EACf,gBAAgB,EAChB,YAAY,EACZ,SAAS,EACT,MAAM,EACN,aAAa,EACb,oBAAoB,EACpB,iBAAiB,EACjB,MAAM,EACN,aAAa,EACb,aAAa,EACb,eAAe,GAChB,MAAM,YAAY,CAAC;AAEpB,YAAY,EACV,QAAQ,EACR,YAAY,EACZ,YAAY,EACZ,WAAW,EACX,aAAa,EACb,SAAS,EACT,YAAY,EACZ,aAAa,EACb,QAAQ,EACR,YAAY,EACZ,SAAS,EACT,QAAQ,EACR,QAAQ,EACR,SAAS,EACT,cAAc,EACd,QAAQ,EACR,YAAY,EACZ,UAAU,EACV,cAAc,EACd,iBAAiB,EACjB,QAAQ,EACR,cAAc,EACd,YAAY,EACZ,aAAa,EACb,WAAW,EACX,iBAAiB,EACjB,QAAQ,EACR,kBAAkB,EAClB,OAAO,EACP,OAAO,GACR,MAAM,mBAAmB,CAAC;AAE3B,OAAO,EACL,UAAU,EACV,YAAY,EACZ,OAAO,EACP,WAAW,EACX,WAAW,EACX,eAAe,GAChB,MAAM,mBAAmB,CAAC;AAE3B,OAAO,EACL,eAAe,EACf,eAAe,EACf,cAAc,EACd,oBAAoB,EACpB,mBAAmB,EACnB,sBAAsB,EACtB,qBAAqB,GACtB,MAAM,iBAAiB,CAAC;AAEzB,OAAO,EACL,eAAe,EACf,sBAAsB,EACtB,eAAe,EACf,aAAa,EACb,UAAU,EACV,WAAW,EACX,iBAAiB,GAClB,MAAM,aAAa,CAAC;AAErB,OAAO,EACL,mBAAmB,EACnB,gBAAgB,EAChB,aAAa,EACb,sBAAsB,EACtB,YAAY,EACZ,UAAU,EACV,YAAY,EACZ,UAAU,EACV,WAAW,GACZ,MAAM,YAAY,CAAC;AAEpB,OAAO,EAAE,SAAS,EAAE,QAAQ,EAAE,YAAY,EAAE,MAAM,UAAU,CAAC"}

package/dist/index.js

@@ -0,0 +1,6 @@
+export { createNode, isNodeOfType, walkAst, getNodeText, countTokens, findNodesOfType, } from "./document-ast.js";
+export { extensionToMime, mimeToExtension, detectMimeType, isSupportedExtension, isSupportedMimeType, getSupportedExtensions, getSupportedMimeTypes, } from "./mime-types.js";
+export { MarkItDownError, UnsupportedFormatError, ConversionError, FileReadError, ParseError, PluginError, CancellationError, } from "./errors.js";
+export { uint8ArrayToDataUrl, blobToUint8Array, readInputData, detectMimeTypeFromData, truncateText, generateId, mergeOptions, AbortError, checkSignal, } from "./utils.js";
+export { parseHTML, parseXML, serializeXML } from "./dom.js";
+//# sourceMappingURL=index.js.map

package/dist/index.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AA4DA,OAAO,EACL,UAAU,EACV,YAAY,EACZ,OAAO,EACP,WAAW,EACX,WAAW,EACX,eAAe,GAChB,MAAM,mBAAmB,CAAC;AAE3B,OAAO,EACL,eAAe,EACf,eAAe,EACf,cAAc,EACd,oBAAoB,EACpB,mBAAmB,EACnB,sBAAsB,EACtB,qBAAqB,GACtB,MAAM,iBAAiB,CAAC;AAEzB,OAAO,EACL,eAAe,EACf,sBAAsB,EACtB,eAAe,EACf,aAAa,EACb,UAAU,EACV,WAAW,EACX,iBAAiB,GAClB,MAAM,aAAa,CAAC;AAErB,OAAO,EACL,mBAAmB,EACnB,gBAAgB,EAChB,aAAa,EACb,sBAAsB,EACtB,YAAY,EACZ,UAAU,EACV,YAAY,EACZ,UAAU,EACV,WAAW,GACZ,MAAM,YAAY,CAAC;AAEpB,OAAO,EAAE,SAAS,EAAE,QAAQ,EAAE,YAAY,EAAE,MAAM,UAAU,CAAC"}

package/dist/mime-types.d.ts

@@ -0,0 +1,9 @@
+import type { MimeType, FileExtension } from "./types.js";
+export declare function extensionToMime(extension: FileExtension): MimeType | undefined;
+export declare function mimeToExtension(mimeType: MimeType): FileExtension | undefined;
+export declare function detectMimeType(fileName: string): MimeType | undefined;
+export declare function isSupportedExtension(extension: FileExtension): boolean;
+export declare function isSupportedMimeType(mimeType: MimeType): boolean;
+export declare function getSupportedExtensions(): FileExtension[];
+export declare function getSupportedMimeTypes(): MimeType[];
+//# sourceMappingURL=mime-types.d.ts.map

package/dist/mime-types.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"mime-types.d.ts","sourceRoot":"","sources":["../src/mime-types.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,QAAQ,EAAE,aAAa,EAAE,MAAM,YAAY,CAAC;AAyC1D,wBAAgB,eAAe,CAAC,SAAS,EAAE,aAAa,GAAG,QAAQ,GAAG,SAAS,CAG9E;AAED,wBAAgB,eAAe,CAAC,QAAQ,EAAE,QAAQ,GAAG,aAAa,GAAG,SAAS,CAE7E;AAED,wBAAgB,cAAc,CAAC,QAAQ,EAAE,MAAM,GAAG,QAAQ,GAAG,SAAS,CAKrE;AAED,wBAAgB,oBAAoB,CAAC,SAAS,EAAE,aAAa,GAAG,OAAO,CAGtE;AAED,wBAAgB,mBAAmB,CAAC,QAAQ,EAAE,QAAQ,GAAG,OAAO,CAE/D;AAED,wBAAgB,sBAAsB,IAAI,aAAa,EAAE,CAExD;AAED,wBAAgB,qBAAqB,IAAI,QAAQ,EAAE,CAElD"}

package/dist/mime-types.js

@@ -0,0 +1,63 @@
+const EXTENSION_TO_MIME = {
+    ".pdf": "application/pdf",
+    ".docx": "application/vnd.openxmlformats-officedocument.wordprocessingml.document",
+    ".doc": "application/msword",
+    ".pptx": "application/vnd.openxmlformats-officedocument.presentationml.presentation",
+    ".ppt": "application/vnd.ms-powerpoint",
+    ".xlsx": "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet",
+    ".xls": "application/vnd.ms-excel",
+    ".csv": "text/csv",
+    ".tsv": "text/tab-separated-values",
+    ".json": "application/json",
+    ".xml": "application/xml",
+    ".html": "text/html",
+    ".htm": "text/html",
+    ".mhtml": "multipart/related",
+    ".epub": "application/epub+zip",
+    ".txt": "text/plain",
+    ".md": "text/markdown",
+    ".markdown": "text/markdown",
+    ".zip": "application/zip",
+    ".png": "image/png",
+    ".jpg": "image/jpeg",
+    ".jpeg": "image/jpeg",
+    ".webp": "image/webp",
+    ".gif": "image/gif",
+    ".tiff": "image/tiff",
+    ".tif": "image/tiff",
+    ".bmp": "image/bmp",
+    ".mp3": "audio/mpeg",
+    ".wav": "audio/wav",
+    ".m4a": "audio/mp4",
+    ".ogg": "audio/ogg",
+    ".flac": "audio/flac",
+};
+const MIME_TO_EXTENSION = Object.fromEntries(Object.entries(EXTENSION_TO_MIME).map(([ext, mime]) => [mime, ext]));
+export function extensionToMime(extension) {
+    const ext = extension.startsWith(".") ? extension : `.${extension}`;
+    return EXTENSION_TO_MIME[ext.toLowerCase()];
+}
+export function mimeToExtension(mimeType) {
+    return MIME_TO_EXTENSION[mimeType.toLowerCase()];
+}
+export function detectMimeType(fileName) {
+    const dotIndex = fileName.lastIndexOf(".");
+    if (dotIndex === -1)
+        return undefined;
+    const ext = fileName.substring(dotIndex);
+    return extensionToMime(ext);
+}
+export function isSupportedExtension(extension) {
+    const ext = extension.startsWith(".") ? extension : `.${extension}`;
+    return ext.toLowerCase() in EXTENSION_TO_MIME;
+}
+export function isSupportedMimeType(mimeType) {
+    return mimeType.toLowerCase() in MIME_TO_EXTENSION;
+}
+export function getSupportedExtensions() {
+    return Object.keys(EXTENSION_TO_MIME);
+}
+export function getSupportedMimeTypes() {
+    return Object.keys(MIME_TO_EXTENSION);
+}
+//# sourceMappingURL=mime-types.js.map

package/dist/mime-types.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"mime-types.js","sourceRoot":"","sources":["../src/mime-types.ts"],"names":[],"mappings":"AAEA,MAAM,iBAAiB,GAAoC;IACzD,MAAM,EAAE,iBAAiB;IACzB,OAAO,EAAE,yEAAyE;IAClF,MAAM,EAAE,oBAAoB;IAC5B,OAAO,EAAE,2EAA2E;IACpF,MAAM,EAAE,+BAA+B;IACvC,OAAO,EAAE,mEAAmE;IAC5E,MAAM,EAAE,0BAA0B;IAClC,MAAM,EAAE,UAAU;IAClB,MAAM,EAAE,2BAA2B;IACnC,OAAO,EAAE,kBAAkB;IAC3B,MAAM,EAAE,iBAAiB;IACzB,OAAO,EAAE,WAAW;IACpB,MAAM,EAAE,WAAW;IACnB,QAAQ,EAAE,mBAAmB;IAC7B,OAAO,EAAE,sBAAsB;IAC/B,MAAM,EAAE,YAAY;IACpB,KAAK,EAAE,eAAe;IACtB,WAAW,EAAE,eAAe;IAC5B,MAAM,EAAE,iBAAiB;IACzB,MAAM,EAAE,WAAW;IACnB,MAAM,EAAE,YAAY;IACpB,OAAO,EAAE,YAAY;IACrB,OAAO,EAAE,YAAY;IACrB,MAAM,EAAE,WAAW;IACnB,OAAO,EAAE,YAAY;IACrB,MAAM,EAAE,YAAY;IACpB,MAAM,EAAE,WAAW;IACnB,MAAM,EAAE,YAAY;IACpB,MAAM,EAAE,WAAW;IACnB,MAAM,EAAE,WAAW;IACnB,MAAM,EAAE,WAAW;IACnB,OAAO,EAAE,YAAY;CACtB,CAAC;AAEF,MAAM,iBAAiB,GAAkC,MAAM,CAAC,WAAW,CACzE,MAAM,CAAC,OAAO,CAAC,iBAAiB,CAAC,CAAC,GAAG,CAAC,CAAC,CAAC,GAAG,EAAE,IAAI,CAAC,EAAE,EAAE,CAAC,CAAC,IAAI,EAAE,GAAG,CAAC,CAAC,CACpE,CAAC;AAEF,MAAM,UAAU,eAAe,CAAC,SAAwB;IACtD,MAAM,GAAG,GAAG,SAAS,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,IAAI,SAAS,EAAE,CAAC;IACpE,OAAO,iBAAiB,CAAC,GAAG,CAAC,WAAW,EAAE,CAAC,CAAC;AAC9C,CAAC;AAED,MAAM,UAAU,eAAe,CAAC,QAAkB;IAChD,OAAO,iBAAiB,CAAC,QAAQ,CAAC,WAAW,EAAE,CAAC,CAAC;AACnD,CAAC;AAED,MAAM,UAAU,cAAc,CAAC,QAAgB;IAC7C,MAAM,QAAQ,GAAG,QAAQ,CAAC,WAAW,CAAC,GAAG,CAAC,CAAC;IAC3C,IAAI,QAAQ,KAAK,CAAC,CAAC;QAAE,OAAO,SAAS,CAAC;IACtC,MAAM,GAAG,GAAG,QAAQ,CAAC,SAAS,CAAC,QAAQ,CAAC,CAAC;IACzC,OAAO,eAAe,CAAC,GAAG,CAAC,CAAC;AAC9B,CAAC;AAED,MAAM,UAAU,oBAAoB,CAAC,SAAwB;IAC3D,MAAM,GAAG,GAAG,SAAS,CAAC,UAAU,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,SAAS,CAAC,CAAC,CAAC,IAAI,SAAS,EAAE,CAAC;IACpE,OAAO,GAAG,CAAC,WAAW,EAAE,IAAI,iBAAiB,CAAC;AAChD,CAAC;AAED,MAAM,UAAU,mBAAmB,CAAC,QAAkB;IACpD,OAAO,QAAQ,CAAC,WAAW,EAAE,IAAI,iBAAiB,CAAC;AACrD,CAAC;AAED,MAAM,UAAU,sBAAsB;IACpC,OAAO,MAAM,CAAC,IAAI,CAAC,iBAAiB,CAAC,CAAC;AACxC,CAAC;AAED,MAAM,UAAU,qBAAqB;IACnC,OAAO,MAAM,CAAC,IAAI,CAAC,iBAAiB,CAAC,CAAC;AACxC,CAAC"}

package/dist/types.d.ts

@@ -0,0 +1,171 @@
+import type { AnyNode } from "./document-ast.js";
+export type MimeType = string;
+export type FileExtension = string;
+export type ConverterId = string;
+/** Standardized document metadata schema */
+export interface DocumentMetadata {
+    title?: string;
+    author?: string;
+    createdAt?: string;
+    modifiedAt?: string;
+    pages?: number;
+    language?: string;
+    format?: string;
+    source?: string;
+    description?: string;
+    subject?: string;
+    keywords?: string[];
+    wordCount?: number;
+    characterCount?: number;
+    pageCount?: number;
+    /** @deprecated Use createdAt instead */
+    createdDate?: Date;
+    /** @deprecated Use modifiedAt instead */
+    modifiedDate?: Date;
+    customProperties?: Record<string, unknown>;
+}
+/** Chunk metadata for RAG workflows */
+export interface ChunkMetadata {
+    chunkId: string;
+    page?: number;
+    headingPath: string[];
+    tokenCount: number;
+    sourceFile?: string;
+    startIndex: number;
+    endIndex: number;
+}
+/** Document chunk for RAG */
+export interface DocumentChunk {
+    id: string;
+    content: string;
+    metadata: ChunkMetadata;
+    ast: AnyNode;
+}
+/** Chunking configuration */
+export interface ChunkingOptions {
+    enabled?: boolean;
+    strategy?: "heading" | "page" | "semantic" | "fixed";
+    maxTokens?: number;
+    overlap?: number;
+    headingDepth?: number;
+    sourceFile?: string;
+}
+export interface ConversionInput {
+    data: Uint8Array | ArrayBuffer | Blob | string;
+    mimeType?: MimeType;
+    fileName?: string;
+    options?: ConversionOptions;
+}
+export interface ConversionOptions {
+    outputFormat?: OutputFormat;
+    includeMetadata?: boolean;
+    includeImages?: boolean;
+    includeTables?: boolean;
+    maxTokens?: number;
+    customProperties?: Record<string, unknown>;
+    onProgress?: ProgressCallback;
+    signal?: AbortSignal;
+    /** Chunking configuration for RAG workflows */
+    chunking?: ChunkingOptions;
+}
+export type OutputFormat = "markdown" | "json" | "plaintext" | "html";
+export interface ConversionResult {
+    markdown: string;
+    metadata: DocumentMetadata;
+    assets: AssetInfo[];
+    tables: TableData[];
+    images: ImageInfo[];
+    headings: HeadingInfo[];
+    ast?: AnyNode;
+    format: OutputFormat;
+    converterId: ConverterId;
+    stats: ConversionStats;
+    /** Document chunks for RAG workflows (only populated when chunking is enabled) */
+    chunks?: DocumentChunk[];
+}
+export interface AssetInfo {
+    id: string;
+    type: "image" | "font" | "stylesheet" | "other";
+    data?: Uint8Array;
+    mimeType: string;
+    fileName?: string;
+    url?: string;
+}
+export interface TableData {
+    headers: string[];
+    rows: string[][];
+    caption?: string;
+}
+export interface ImageInfo {
+    id: string;
+    data?: Uint8Array;
+    mimeType: string;
+    alt?: string;
+    caption?: string;
+    width?: number;
+    height?: number;
+}
+export interface HeadingInfo {
+    level: number;
+    text: string;
+    id?: string;
+}
+export interface ConversionStats {
+    startTime: number;
+    endTime: number;
+    duration: number;
+    inputSize: number;
+    outputSize: number;
+    pagesProcessed?: number;
+    chunksProcessed?: number;
+}
+export type ProgressCallback = (progress: ProgressInfo) => void;
+export interface ProgressInfo {
+    phase: "parsing" | "converting" | "rendering" | "chunking";
+    percentage: number;
+    message?: string;
+}
+export interface Converter {
+    readonly id: ConverterId;
+    readonly supportedMimeTypes: MimeType[];
+    readonly supportedExtensions: FileExtension[];
+    canConvert(input: ConversionInput): Promise<boolean>;
+    convert(input: ConversionInput): Promise<ConversionResult>;
+}
+export interface Plugin {
+    readonly id: string;
+    readonly name: string;
+    readonly version: string;
+    converters?: Converter[];
+    middleware?: ConversionMiddleware[];
+    initialize?(context: PluginContext): Promise<void>;
+    destroy?(): Promise<void>;
+}
+export interface PluginContext {
+    registry: ConverterRegistry;
+    logger: Logger;
+    config: Record<string, unknown>;
+}
+export interface ConversionMiddleware {
+    name: string;
+    priority: number;
+    beforeConvert?(input: ConversionInput): Promise<ConversionInput>;
+    afterConvert?(result: ConversionResult): Promise<ConversionResult>;
+}
+export interface ConverterRegistry {
+    register(converter: Converter): void;
+    unregister(id: ConverterId): void;
+    get(id: ConverterId): Converter | undefined;
+    getByMimeType(mimeType: MimeType): Converter | undefined;
+    getByExtension(extension: FileExtension): Converter | undefined;
+    list(): Converter[];
+    canConvert(input: ConversionInput): Promise<Converter | undefined>;
+    getMiddlewares(): ConversionMiddleware[];
+}
+export interface Logger {
+    debug(message: string, ...args: unknown[]): void;
+    info(message: string, ...args: unknown[]): void;
+    warn(message: string, ...args: unknown[]): void;
+    error(message: string, ...args: unknown[]): void;
+}
+//# sourceMappingURL=types.d.ts.map

package/dist/types.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":"AAAA,OAAO,KAAK,EAAE,OAAO,EAAE,MAAM,mBAAmB,CAAC;AAEjD,MAAM,MAAM,QAAQ,GAAG,MAAM,CAAC;AAC9B,MAAM,MAAM,aAAa,GAAG,MAAM,CAAC;AACnC,MAAM,MAAM,WAAW,GAAG,MAAM,CAAC;AAEjC,4CAA4C;AAC5C,MAAM,WAAW,gBAAgB;IAC/B,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,MAAM,EAAE,CAAC;IACpB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,cAAc,CAAC,EAAE,MAAM,CAAC;IACxB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,wCAAwC;IACxC,WAAW,CAAC,EAAE,IAAI,CAAC;IACnB,yCAAyC;IACzC,YAAY,CAAC,EAAE,IAAI,CAAC;IACpB,gBAAgB,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CAC5C;AAED,uCAAuC;AACvC,MAAM,WAAW,aAAa;IAC5B,OAAO,EAAE,MAAM,CAAC;IAChB,IAAI,CAAC,EAAE,MAAM,CAAC;IACd,WAAW,EAAE,MAAM,EAAE,CAAC;IACtB,UAAU,EAAE,MAAM,CAAC;IACnB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,UAAU,EAAE,MAAM,CAAC;IACnB,QAAQ,EAAE,MAAM,CAAC;CAClB;AAED,6BAA6B;AAC7B,MAAM,WAAW,aAAa;IAC5B,EAAE,EAAE,MAAM,CAAC;IACX,OAAO,EAAE,MAAM,CAAC;IAChB,QAAQ,EAAE,aAAa,CAAC;IACxB,GAAG,EAAE,OAAO,CAAC;CACd;AAED,6BAA6B;AAC7B,MAAM,WAAW,eAAe;IAC9B,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,QAAQ,CAAC,EAAE,SAAS,GAAG,MAAM,GAAG,UAAU,GAAG,OAAO,CAAC;IACrD,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,YAAY,CAAC,EAAE,MAAM,CAAC;IACtB,UAAU,CAAC,EAAE,MAAM,CAAC;CACrB;AAED,MAAM,WAAW,eAAe;IAC9B,IAAI,EAAE,UAAU,GAAG,WAAW,GAAG,IAAI,GAAG,MAAM,CAAC;IAC/C,QAAQ,CAAC,EAAE,QAAQ,CAAC;IACpB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,OAAO,CAAC,EAAE,iBAAiB,CAAC;CAC7B;AAED,MAAM,WAAW,iBAAiB;IAChC,YAAY,CAAC,EAAE,YAAY,CAAC;IAC5B,eAAe,CAAC,EAAE,OAAO,CAAC;IAC1B,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,gBAAgB,CAAC,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;IAC3C,UAAU,CAAC,EAAE,gBAAgB,CAAC;IAC9B,MAAM,CAAC,EAAE,WAAW,CAAC;IACrB,+CAA+C;IAC/C,QAAQ,CAAC,EAAE,eAAe,CAAC;CAC5B;AAED,MAAM,MAAM,YAAY,GAAG,UAAU,GAAG,MAAM,GAAG,WAAW,GAAG,MAAM,CAAC;AAEtE,MAAM,WAAW,gBAAgB;IAC/B,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,EAAE,gBAAgB,CAAC;IAC3B,MAAM,EAAE,SAAS,EAAE,CAAC;IACpB,MAAM,EAAE,SAAS,EAAE,CAAC;IACpB,MAAM,EAAE,SAAS,EAAE,CAAC;IACpB,QAAQ,EAAE,WAAW,EAAE,CAAC;IACxB,GAAG,CAAC,EAAE,OAAO,CAAC;IACd,MAAM,EAAE,YAAY,CAAC;IACrB,WAAW,EAAE,WAAW,CAAC;IACzB,KAAK,EAAE,eAAe,CAAC;IACvB,kFAAkF;IAClF,MAAM,CAAC,EAAE,aAAa,EAAE,CAAC;CAC1B;AAED,MAAM,WAAW,SAAS;IACxB,EAAE,EAAE,MAAM,CAAC;IACX,IAAI,EAAE,OAAO,GAAG,MAAM,GAAG,YAAY,GAAG,OAAO,CAAC;IAChD,IAAI,CAAC,EAAE,UAAU,CAAC;IAClB,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,GAAG,CAAC,EAAE,MAAM,CAAC;CACd;AAED,MAAM,WAAW,SAAS;IACxB,OAAO,EAAE,MAAM,EAAE,CAAC;IAClB,IAAI,EAAE,MAAM,EAAE,EAAE,CAAC;IACjB,OAAO,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,MAAM,WAAW,SAAS;IACxB,EAAE,EAAE,MAAM,CAAC;IACX,IAAI,CAAC,EAAE,UAAU,CAAC;IAClB,QAAQ,EAAE,MAAM,CAAC;IACjB,GAAG,CAAC,EAAE,MAAM,CAAC;IACb,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;CACjB;AAED,MAAM,WAAW,WAAW;IAC1B,KAAK,EAAE,MAAM,CAAC;IACd,IAAI,EAAE,MAAM,CAAC;IACb,EAAE,CAAC,EAAE,MAAM,CAAC;CACb;AAED,MAAM,WAAW,eAAe;IAC9B,SAAS,EAAE,MAAM,CAAC;IAClB,OAAO,EAAE,MAAM,CAAC;IAChB,QAAQ,EAAE,MAAM,CAAC;IACjB,SAAS,EAAE,MAAM,CAAC;IAClB,UAAU,EAAE,MAAM,CAAC;IACnB,cAAc,CAAC,EAAE,MAAM,CAAC;IACxB,eAAe,CAAC,EAAE,MAAM,CAAC;CAC1B;AAED,MAAM,MAAM,gBAAgB,GAAG,CAAC,QAAQ,EAAE,YAAY,KAAK,IAAI,CAAC;AAEhE,MAAM,WAAW,YAAY;IAC3B,KAAK,EAAE,SAAS,GAAG,YAAY,GAAG,WAAW,GAAG,UAAU,CAAC;IAC3D,UAAU,EAAE,MAAM,CAAC;IACnB,OAAO,CAAC,EAAE,MAAM,CAAC;CAClB;AAED,MAAM,WAAW,SAAS;IACxB,QAAQ,CAAC,EAAE,EAAE,WAAW,CAAC;IACzB,QAAQ,CAAC,kBAAkB,EAAE,QAAQ,EAAE,CAAC;IACxC,QAAQ,CAAC,mBAAmB,EAAE,aAAa,EAAE,CAAC;IAC9C,UAAU,CAAC,KAAK,EAAE,eAAe,GAAG,OAAO,CAAC,OAAO,CAAC,CAAC;IACrD,OAAO,CAAC,KAAK,EAAE,eAAe,GAAG,OAAO,CAAC,gBAAgB,CAAC,CAAC;CAC5D;AAED,MAAM,WAAW,MAAM;IACrB,QAAQ,CAAC,EAAE,EAAE,MAAM,CAAC;IACpB,QAAQ,CAAC,IAAI,EAAE,MAAM,CAAC;IACtB,QAAQ,CAAC,OAAO,EAAE,MAAM,CAAC;IACzB,UAAU,CAAC,EAAE,SAAS,EAAE,CAAC;IACzB,UAAU,CAAC,EAAE,oBAAoB,EAAE,CAAC;IACpC,UAAU,CAAC,CAAC,OAAO,EAAE,aAAa,GAAG,OAAO,CAAC,IAAI,CAAC,CAAC;IACnD,OAAO,CAAC,IAAI,OAAO,CAAC,IAAI,CAAC,CAAC;CAC3B;AAED,MAAM,WAAW,aAAa;IAC5B,QAAQ,EAAE,iBAAiB,CAAC;IAC5B,MAAM,EAAE,MAAM,CAAC;IACf,MAAM,EAAE,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,CAAC;CACjC;AAED,MAAM,WAAW,oBAAoB;IACnC,IAAI,EAAE,MAAM,CAAC;IACb,QAAQ,EAAE,MAAM,CAAC;IACjB,aAAa,CAAC,CAAC,KAAK,EAAE,eAAe,GAAG,OAAO,CAAC,eAAe,CAAC,CAAC;IACjE,YAAY,CAAC,CAAC,MAAM,EAAE,gBAAgB,GAAG,OAAO,CAAC,gBAAgB,CAAC,CAAC;CACpE;AAED,MAAM,WAAW,iBAAiB;IAChC,QAAQ,CAAC,SAAS,EAAE,SAAS,GAAG,IAAI,CAAC;IACrC,UAAU,CAAC,EAAE,EAAE,WAAW,GAAG,IAAI,CAAC;IAClC,GAAG,CAAC,EAAE,EAAE,WAAW,GAAG,SAAS,GAAG,SAAS,CAAC;IAC5C,aAAa,CAAC,QAAQ,EAAE,QAAQ,GAAG,SAAS,GAAG,SAAS,CAAC;IACzD,cAAc,CAAC,SAAS,EAAE,aAAa,GAAG,SAAS,GAAG,SAAS,CAAC;IAChE,IAAI,IAAI,SAAS,EAAE,CAAC;IACpB,UAAU,CAAC,KAAK,EAAE,eAAe,GAAG,OAAO,CAAC,SAAS,GAAG,SAAS,CAAC,CAAC;IACnE,cAAc,IAAI,oBAAoB,EAAE,CAAC;CAC1C;AAED,MAAM,WAAW,MAAM;IACrB,KAAK,CAAC,OAAO,EAAE,MAAM,EAAE,GAAG,IAAI,EAAE,OAAO,EAAE,GAAG,IAAI,CAAC;IACjD,IAAI,CAAC,OAAO,EAAE,MAAM,EAAE,GAAG,IAAI,EAAE,OAAO,EAAE,GAAG,IAAI,CAAC;IAChD,IAAI,CAAC,OAAO,EAAE,MAAM,EAAE,GAAG,IAAI,EAAE,OAAO,EAAE,GAAG,IAAI,CAAC;IAChD,KAAK,CAAC,OAAO,EAAE,MAAM,EAAE,GAAG,IAAI,EAAE,OAAO,EAAE,GAAG,IAAI,CAAC;CAClD"}

package/dist/types.js

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=types.js.map

package/dist/types.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"types.js","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":""}

package/dist/utils.d.ts

@@ -0,0 +1,12 @@
+export declare function uint8ArrayToDataUrl(data: Uint8Array, mimeType: string): string;
+export declare function blobToUint8Array(blob: Blob): Promise<Uint8Array>;
+export declare function readInputData(input: Uint8Array | ArrayBuffer | Blob | string): Promise<Uint8Array>;
+export declare function detectMimeTypeFromData(data: Uint8Array): string | undefined;
+export declare function truncateText(text: string, maxLength: number): string;
+export declare function generateId(): string;
+export declare function mergeOptions<T extends Record<string, unknown>>(defaults: T, overrides?: Partial<T>): T;
+export declare class AbortError extends Error {
+    constructor(message?: string);
+}
+export declare function checkSignal(signal?: AbortSignal): void;
+//# sourceMappingURL=utils.d.ts.map

package/dist/utils.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"utils.d.ts","sourceRoot":"","sources":["../src/utils.ts"],"names":[],"mappings":"AAAA,wBAAgB,mBAAmB,CAAC,IAAI,EAAE,UAAU,EAAE,QAAQ,EAAE,MAAM,GAAG,MAAM,CAS9E;AAED,wBAAsB,gBAAgB,CAAC,IAAI,EAAE,IAAI,GAAG,OAAO,CAAC,UAAU,CAAC,CAEtE;AAED,wBAAsB,aAAa,CACjC,KAAK,EAAE,UAAU,GAAG,WAAW,GAAG,IAAI,GAAG,MAAM,GAC9C,OAAO,CAAC,UAAU,CAAC,CAWrB;AAED,wBAAgB,sBAAsB,CAAC,IAAI,EAAE,UAAU,GAAG,MAAM,GAAG,SAAS,CAuB3E;AAED,wBAAgB,YAAY,CAAC,IAAI,EAAE,MAAM,EAAE,SAAS,EAAE,MAAM,GAAG,MAAM,CAGpE;AAED,wBAAgB,UAAU,IAAI,MAAM,CAEnC;AAED,wBAAgB,YAAY,CAAC,CAAC,SAAS,MAAM,CAAC,MAAM,EAAE,OAAO,CAAC,EAC5D,QAAQ,EAAE,CAAC,EACX,SAAS,CAAC,EAAE,OAAO,CAAC,CAAC,CAAC,GACrB,CAAC,CAGH;AAED,qBAAa,UAAW,SAAQ,KAAK;gBACvB,OAAO,SAA0B;CAI9C;AAED,wBAAgB,WAAW,CAAC,MAAM,CAAC,EAAE,WAAW,GAAG,IAAI,CAItD"}

package/dist/utils.js

@@ -0,0 +1,73 @@
+export function uint8ArrayToDataUrl(data, mimeType) {
+    let binary = "";
+    const chunkSize = 8192;
+    for (let i = 0; i < data.length; i += chunkSize) {
+        const chunk = data.subarray(i, i + chunkSize);
+        binary += String.fromCharCode(...chunk);
+    }
+    const base64 = btoa(binary);
+    return `data:${mimeType};base64,${base64}`;
+}
+export async function blobToUint8Array(blob) {
+    return new Uint8Array(await blob.arrayBuffer());
+}
+export async function readInputData(input) {
+    if (typeof input === "string") {
+        return new TextEncoder().encode(input);
+    }
+    if (input instanceof Uint8Array) {
+        return input;
+    }
+    if (input instanceof ArrayBuffer) {
+        return new Uint8Array(input);
+    }
+    return blobToUint8Array(input);
+}
+export function detectMimeTypeFromData(data) {
+    if (data.length < 4)
+        return undefined;
+    if (data[0] === 0x25 && data[1] === 0x50 && data[2] === 0x44 && data[3] === 0x46) {
+        return "application/pdf";
+    }
+    if (data[0] === 0x50 && data[1] === 0x4b && data[2] === 0x03 && data[3] === 0x04) {
+        return "application/zip";
+    }
+    if (data[0] === 0x89 && data[1] === 0x50 && data[2] === 0x4e && data[3] === 0x47) {
+        return "image/png";
+    }
+    if (data[0] === 0xff && data[1] === 0xd8 && data[2] === 0xff) {
+        return "image/jpeg";
+    }
+    if (data[0] === 0x52 && data[1] === 0x49 && data[2] === 0x46 && data[3] === 0x46) {
+        return "image/webp";
+    }
+    if (data[0] === 0x47 && data[1] === 0x49 && data[2] === 0x46 && data[3] === 0x38) {
+        return "image/gif";
+    }
+    return undefined;
+}
+export function truncateText(text, maxLength) {
+    if (text.length <= maxLength)
+        return text;
+    return text.slice(0, maxLength - 3) + "...";
+}
+export function generateId() {
+    return Math.random().toString(36).substring(2, 11);
+}
+export function mergeOptions(defaults, overrides) {
+    if (!overrides)
+        return defaults;
+    return { ...defaults, ...overrides };
+}
+export class AbortError extends Error {
+    constructor(message = "Operation was aborted") {
+        super(message);
+        this.name = "AbortError";
+    }
+}
+export function checkSignal(signal) {
+    if (signal?.aborted) {
+        throw new AbortError();
+    }
+}
+//# sourceMappingURL=utils.js.map

package/dist/utils.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"utils.js","sourceRoot":"","sources":["../src/utils.ts"],"names":[],"mappings":"AAAA,MAAM,UAAU,mBAAmB,CAAC,IAAgB,EAAE,QAAgB;IACpE,IAAI,MAAM,GAAG,EAAE,CAAC;IAChB,MAAM,SAAS,GAAG,IAAI,CAAC;IACvB,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,IAAI,CAAC,MAAM,EAAE,CAAC,IAAI,SAAS,EAAE,CAAC;QAChD,MAAM,KAAK,GAAG,IAAI,CAAC,QAAQ,CAAC,CAAC,EAAE,CAAC,GAAG,SAAS,CAAC,CAAC;QAC9C,MAAM,IAAI,MAAM,CAAC,YAAY,CAAC,GAAG,KAAK,CAAC,CAAC;IAC1C,CAAC;IACD,MAAM,MAAM,GAAG,IAAI,CAAC,MAAM,CAAC,CAAC;IAC5B,OAAO,QAAQ,QAAQ,WAAW,MAAM,EAAE,CAAC;AAC7C,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,gBAAgB,CAAC,IAAU;IAC/C,OAAO,IAAI,UAAU,CAAC,MAAM,IAAI,CAAC,WAAW,EAAE,CAAC,CAAC;AAClD,CAAC;AAED,MAAM,CAAC,KAAK,UAAU,aAAa,CACjC,KAA+C;IAE/C,IAAI,OAAO,KAAK,KAAK,QAAQ,EAAE,CAAC;QAC9B,OAAO,IAAI,WAAW,EAAE,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC;IACzC,CAAC;IACD,IAAI,KAAK,YAAY,UAAU,EAAE,CAAC;QAChC,OAAO,KAAK,CAAC;IACf,CAAC;IACD,IAAI,KAAK,YAAY,WAAW,EAAE,CAAC;QACjC,OAAO,IAAI,UAAU,CAAC,KAAK,CAAC,CAAC;IAC/B,CAAC;IACD,OAAO,gBAAgB,CAAC,KAAK,CAAC,CAAC;AACjC,CAAC;AAED,MAAM,UAAU,sBAAsB,CAAC,IAAgB;IACrD,IAAI,IAAI,CAAC,MAAM,GAAG,CAAC;QAAE,OAAO,SAAS,CAAC;IAEtC,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACjF,OAAO,iBAAiB,CAAC;IAC3B,CAAC;IACD,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACjF,OAAO,iBAAiB,CAAC;IAC3B,CAAC;IACD,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACjF,OAAO,WAAW,CAAC;IACrB,CAAC;IACD,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QAC7D,OAAO,YAAY,CAAC;IACtB,CAAC;IACD,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACjF,OAAO,YAAY,CAAC;IACtB,CAAC;IACD,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,IAAI,IAAI,CAAC,CAAC,CAAC,KAAK,IAAI,EAAE,CAAC;QACjF,OAAO,WAAW,CAAC;IACrB,CAAC;IAED,OAAO,SAAS,CAAC;AACnB,CAAC;AAED,MAAM,UAAU,YAAY,CAAC,IAAY,EAAE,SAAiB;IAC1D,IAAI,IAAI,CAAC,MAAM,IAAI,SAAS;QAAE,OAAO,IAAI,CAAC;IAC1C,OAAO,IAAI,CAAC,KAAK,CAAC,CAAC,EAAE,SAAS,GAAG,CAAC,CAAC,GAAG,KAAK,CAAC;AAC9C,CAAC;AAED,MAAM,UAAU,UAAU;IACxB,OAAO,IAAI,CAAC,MAAM,EAAE,CAAC,QAAQ,CAAC,EAAE,CAAC,CAAC,SAAS,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC;AACrD,CAAC;AAED,MAAM,UAAU,YAAY,CAC1B,QAAW,EACX,SAAsB;IAEtB,IAAI,CAAC,SAAS;QAAE,OAAO,QAAQ,CAAC;IAChC,OAAO,EAAE,GAAG,QAAQ,EAAE,GAAG,SAAS,EAAE,CAAC;AACvC,CAAC;AAED,MAAM,OAAO,UAAW,SAAQ,KAAK;IACnC,YAAY,OAAO,GAAG,uBAAuB;QAC3C,KAAK,CAAC,OAAO,CAAC,CAAC;QACf,IAAI,CAAC,IAAI,GAAG,YAAY,CAAC;IAC3B,CAAC;CACF;AAED,MAAM,UAAU,WAAW,CAAC,MAAoB;IAC9C,IAAI,MAAM,EAAE,OAAO,EAAE,CAAC;QACpB,MAAM,IAAI,UAAU,EAAE,CAAC;IACzB,CAAC;AACH,CAAC"}

package/package.json

@@ -0,0 +1,29 @@
+{
+  "name": "@markitdownjs/shared",
+  "version": "0.1.0",
+  "description": "Shared types, AST nodes, and utilities for MarkItDownJS",
+  "license": "MIT",
+  "type": "module",
+  "main": "./dist/index.js",
+  "module": "./dist/index.js",
+  "types": "./dist/index.d.ts",
+  "exports": {
+    ".": {
+      "import": "./dist/index.js",
+      "types": "./dist/index.d.ts"
+    }
+  },
+  "files": [
+    "dist"
+  ],
+  "devDependencies": {
+    "linkedom": "^0.18.12",
+    "typescript": "^5.5.0"
+  },
+  "scripts": {
+    "build": "tsc --project tsconfig.json",
+    "dev": "tsc --watch --project tsconfig.json",
+    "typecheck": "tsc --noEmit",
+    "clean": "rm -rf dist tsconfig.tsbuildinfo"
+  }
+}