@polotno/pdf-export 0.1.37 → 0.1.39

package/lib/text/parser.d.ts

@@ -1,46 +0,0 @@
-import { TextElement, TextSegment, ListParagraphMeta } from './types.js';
-export declare function decodeHtmlEntities(text: string): string;
-/**
- * Normalize rich text HTML by converting block-level line breaks into newline characters
- * while preserving inline formatting tags.
- */
-export declare function normalizeRichText(text: string): string;
-/**
- * Parse HTML text into styled segments
- */
-export declare function parseHTMLToSegments(html: string, baseElement: TextElement): TextSegment[];
-/**
- * Represents a parsed HTML token (either text content or a tag)
- */
-export interface Token {
-    type: 'text' | 'tag';
-    content: string;
-    decodedContent?: string;
-    tagName?: string;
-    isClosing?: boolean;
-}
-/**
- * Parse HTML into tokens (text and tags)
- */
-export declare function tokenizeHTML(html: string): Token[];
-export interface ParagraphDefinition {
-    html: string;
-    listMeta?: ListParagraphMeta;
-}
-export declare function parseHtmlToParagraphs(html: string): ParagraphDefinition[];
-/**
- * Reconstruct HTML from tokens while maintaining proper tag nesting across line breaks
- * @param tokens - Array of parsed HTML tokens
- * @param openTags - Tags that were opened in previous lines and should be carried forward
- * @returns Reconstructed HTML string and the updated list of open tags
- */
-export declare function tokensToHTML(tokens: Token[], openTags: Array<{
-    name: string;
-    fullTag: string;
-}>): {
-    html: string;
-    openTags: Array<{
-        name: string;
-        fullTag: string;
-    }>;
-};

package/lib/text/parser.js

@@ -1,415 +0,0 @@
-import { decode as decodeEntities } from 'html-entities';
-export function decodeHtmlEntities(text) {
-    if (!text) {
-        return text;
-    }
-    const decoded = decodeEntities(text);
-    // Don't replace tabs here - we'll handle them with expandTabsToTabStops
-    return decoded;
-}
-/**
- * Normalize rich text HTML by converting block-level line breaks into newline characters
- * while preserving inline formatting tags.
- */
-export function normalizeRichText(text) {
-    if (!text) {
-        return text;
-    }
-    let normalized = text.replace(/\r\n/g, '\n').replace(/\r/g, '\n');
-    // Convert explicit HTML break tags into newline characters
-    normalized = normalized.replace(/<br\s*\/?>/gi, '\n');
-    // Treat paragraph boundaries as newlines and drop opening tags
-    normalized = normalized.replace(/<\/p\s*>/gi, '\n');
-    normalized = normalized.replace(/<p[^>]*>/gi, '');
-    // Collapse excessive consecutive newlines produced by HTML cleanup
-    normalized = normalized.replace(/\n{3,}/g, '\n\n');
-    // Trim stray leading/trailing newlines introduced by paragraph conversion
-    normalized = normalized.replace(/^\n+/, '').replace(/\n+$/, '');
-    // Decode common HTML non-breaking space entities into their unicode counterpart
-    normalized = normalized.replace(/&(nbsp|#160|#xA0);/gi, '\u00A0');
-    // Strip zero-width characters that can create missing-glyph boxes in PDF output
-    normalized = normalized.replace(/[\u200B\u200C\u200D\uFEFF\u2060]/g, '');
-    return normalized;
-}
-/**
- * Parse HTML text into styled segments
- */
-export function parseHTMLToSegments(html, baseElement) {
-    const segments = [];
-    const tagStack = [];
-    // Regex to match tags and text content
-    const regex = /<(\/?)(strong|b|em|i|u|span)([^>]*)>|([^<]+)/gi;
-    let match;
-    while ((match = regex.exec(html)) !== null) {
-        if (match[4]) {
-            // Text content
-            const text = decodeHtmlEntities(match[4]);
-            // Calculate current styles from tag stack
-            let bold = false;
-            let italic = false;
-            let underline = false;
-            let color = undefined;
-            for (const tag of tagStack) {
-                if (tag.tag === 'strong' || tag.tag === 'b')
-                    bold = true;
-                if (tag.tag === 'em' || tag.tag === 'i')
-                    italic = true;
-                if (tag.tag === 'u')
-                    underline = true;
-                if (tag.color)
-                    color = tag.color;
-            }
-            segments.push({
-                text,
-                bold,
-                italic,
-                underline,
-                color,
-            });
-        }
-        else {
-            // Tag
-            const isClosing = match[1] === '/';
-            const tagName = match[2].toLowerCase();
-            const attributes = match[3];
-            if (isClosing) {
-                // Remove from stack
-                const index = tagStack.findIndex((t) => t.tag === tagName);
-                if (index !== -1) {
-                    tagStack.splice(index, 1);
-                }
-            }
-            else {
-                // Add to stack
-                const tagData = { tag: tagName };
-                // Parse color from span style attribute
-                if (attributes) {
-                    const colorMatch = /style=["'](?:[^"']*)?color:\s*([^;"']+)/i.exec(attributes);
-                    if (colorMatch) {
-                        tagData.color = colorMatch[1].trim();
-                    }
-                }
-                tagStack.push(tagData);
-            }
-        }
-    }
-    return segments;
-}
-/**
- * Parse HTML into tokens (text and tags)
- */
-export function tokenizeHTML(html) {
-    const tokens = [];
-    const regex = /<(\/?)(strong|b|em|i|u|span)([^>]*)>|([^<]+)/gi;
-    let match;
-    while ((match = regex.exec(html)) !== null) {
-        if (match[4]) {
-            // Text content
-            const decodedContent = decodeHtmlEntities(match[4]);
-            tokens.push({
-                type: 'text',
-                content: match[4],
-                decodedContent,
-            });
-        }
-        else {
-            // Tag
-            const isClosing = match[1] === '/';
-            const tagName = match[2].toLowerCase();
-            tokens.push({
-                type: 'tag',
-                content: match[0],
-                tagName: tagName,
-                isClosing: isClosing,
-            });
-        }
-    }
-    return tokens;
-}
-const VOID_ELEMENTS = new Set(['br']);
-function parseAttributes(raw) {
-    const attributes = {};
-    const attrRegex = /([^\s=]+)(?:\s*=\s*(?:"([^"]*)"|'([^']*)'|([^\s"'=<>`]+)))?/g;
-    let attrMatch;
-    while ((attrMatch = attrRegex.exec(raw)) !== null) {
-        const name = attrMatch[1].toLowerCase();
-        const value = attrMatch[2] ?? attrMatch[3] ?? attrMatch[4] ?? '';
-        attributes[name] = value;
-    }
-    return attributes;
-}
-function parseSimpleHTML(html) {
-    const root = {
-        type: 'element',
-        tagName: 'root',
-        attributes: {},
-        children: [],
-    };
-    const stack = [root];
-    const tagRegex = /<\/?([a-zA-Z0-9:-]+)([^>]*)>/g;
-    let lastIndex = 0;
-    let match;
-    while ((match = tagRegex.exec(html)) !== null) {
-        if (match.index > lastIndex) {
-            const textContent = html.slice(lastIndex, match.index);
-            if (textContent) {
-                stack[stack.length - 1].children.push({
-                    type: 'text',
-                    content: textContent,
-                });
-            }
-        }
-        const fullMatch = match[0];
-        const isClosing = fullMatch.startsWith('</');
-        const tagName = match[1].toLowerCase();
-        const attrChunk = match[2] || '';
-        if (isClosing) {
-            for (let i = stack.length - 1; i >= 0; i--) {
-                if (stack[i].tagName === tagName) {
-                    stack.length = i;
-                    break;
-                }
-            }
-        }
-        else {
-            const attributes = parseAttributes(attrChunk);
-            const elementNode = {
-                type: 'element',
-                tagName,
-                attributes,
-                children: [],
-            };
-            const parent = stack[stack.length - 1];
-            parent.children.push(elementNode);
-            const selfClosing = fullMatch.endsWith('/>') || VOID_ELEMENTS.has(tagName);
-            if (!selfClosing) {
-                stack.push(elementNode);
-            }
-        }
-        lastIndex = tagRegex.lastIndex;
-    }
-    if (lastIndex < html.length) {
-        const remaining = html.slice(lastIndex);
-        if (remaining) {
-            stack[stack.length - 1].children.push({
-                type: 'text',
-                content: remaining,
-            });
-        }
-    }
-    return root;
-}
-function serializeAttributes(attributes) {
-    const entries = Object.entries(attributes);
-    if (!entries.length) {
-        return '';
-    }
-    return (' ' +
-        entries
-            .map(([key, value]) => value === '' ? key : `${key}="${value.replace(/"/g, '&quot;')}"`)
-            .join(' '));
-}
-function shouldSkipNode(node) {
-    if (node.type === 'element' && node.tagName === 'span') {
-        const classAttr = node.attributes['class'] || node.attributes['className'] || '';
-        if (/\bql-cursor\b/.test(classAttr)) {
-            return true;
-        }
-    }
-    return false;
-}
-function serializeNodes(nodes) {
-    return nodes
-        .map((node) => {
-        if (shouldSkipNode(node)) {
-            return '';
-        }
-        if (node.type === 'text') {
-            return node.content.replace(/\uFEFF/g, '');
-        }
-        const attrs = serializeAttributes(node.attributes);
-        if (VOID_ELEMENTS.has(node.tagName)) {
-            return `<${node.tagName}${attrs}>`;
-        }
-        return `<${node.tagName}${attrs}>${serializeNodes(node.children)}</${node.tagName}>`;
-    })
-        .join('');
-}
-function getIndentLevelFromAttributes(attributes) {
-    const classAttr = attributes['class'] || attributes['className'] || '';
-    const match = classAttr.match(/ql-indent-(\d+)/);
-    return match ? parseInt(match[1], 10) : 0;
-}
-function detectIndentLevel(node) {
-    const directIndent = getIndentLevelFromAttributes(node.attributes);
-    if (directIndent > 0) {
-        return directIndent;
-    }
-    for (const child of node.children) {
-        if (child.type === 'element' &&
-            child.tagName !== 'ul' &&
-            child.tagName !== 'ol') {
-            const childIndent = detectIndentLevel(child);
-            if (childIndent > 0) {
-                return childIndent;
-            }
-        }
-    }
-    return 0;
-}
-function nodesToParagraphs(nodes, baseIndentLevel = 0) {
-    const paragraphs = [];
-    let pendingInline = [];
-    const flushInline = () => {
-        if (pendingInline.length === 0) {
-            return;
-        }
-        const html = serializeNodes(pendingInline);
-        paragraphs.push({ html });
-        pendingInline = [];
-    };
-    for (const node of nodes) {
-        if (shouldSkipNode(node)) {
-            continue;
-        }
-        if (node.type === 'text') {
-            const newlineRegex = /\n+/g;
-            let lastIndex = 0;
-            let match;
-            while ((match = newlineRegex.exec(node.content)) !== null) {
-                const chunk = node.content.slice(lastIndex, match.index);
-                if (chunk.length > 0) {
-                    pendingInline.push({
-                        type: 'text',
-                        content: chunk,
-                    });
-                }
-                flushInline();
-                const extraBreaks = match[0].length - 1;
-                for (let extra = 0; extra < extraBreaks; extra++) {
-                    paragraphs.push({ html: '' });
-                }
-                lastIndex = newlineRegex.lastIndex;
-            }
-            const rest = node.content.slice(lastIndex);
-            if (rest.length > 0) {
-                pendingInline.push({
-                    type: 'text',
-                    content: rest,
-                });
-            }
-            continue;
-        }
-        if (node.tagName === 'p') {
-            flushInline();
-            const html = serializeNodes(node.children);
-            paragraphs.push({
-                html,
-            });
-            continue;
-        }
-        if (node.tagName === 'br') {
-            flushInline();
-            paragraphs.push({ html: '' });
-            continue;
-        }
-        if (node.tagName === 'ul' || node.tagName === 'ol') {
-            flushInline();
-            paragraphs.push(...collectListParagraphs(node, baseIndentLevel));
-            continue;
-        }
-        pendingInline.push(node);
-    }
-    flushInline();
-    return paragraphs;
-}
-function collectListParagraphs(listNode, baseIndentLevel) {
-    const paragraphs = [];
-    const listIndent = baseIndentLevel + getIndentLevelFromAttributes(listNode.attributes);
-    let counter = 1;
-    for (const child of listNode.children) {
-        if (child.type !== 'element' || child.tagName !== 'li') {
-            continue;
-        }
-        const liIndentFromAttribute = getIndentLevelFromAttributes(child.attributes);
-        const detectedIndent = detectIndentLevel(child);
-        const indentLevel = liIndentFromAttribute > 0
-            ? listIndent + liIndentFromAttribute
-            : listIndent + detectedIndent;
-        const itemParagraphs = nodesToParagraphs(child.children, indentLevel);
-        let markerAssigned = false;
-        for (const paragraph of itemParagraphs) {
-            if (!paragraph.listMeta) {
-                paragraph.listMeta = {
-                    type: listNode.tagName,
-                    index: counter,
-                    indentLevel,
-                    displayMarker: !markerAssigned,
-                };
-                markerAssigned = true;
-            }
-            paragraphs.push(paragraph);
-        }
-        if (!markerAssigned) {
-            paragraphs.push({
-                html: '',
-                listMeta: {
-                    type: listNode.tagName,
-                    index: counter,
-                    indentLevel,
-                    displayMarker: true,
-                },
-            });
-        }
-        if (listNode.tagName === 'ol') {
-            counter += 1;
-        }
-    }
-    return paragraphs;
-}
-export function parseHtmlToParagraphs(html) {
-    const root = parseSimpleHTML(html);
-    return nodesToParagraphs(root.children);
-}
-/**
- * Reconstruct HTML from tokens while maintaining proper tag nesting across line breaks
- * @param tokens - Array of parsed HTML tokens
- * @param openTags - Tags that were opened in previous lines and should be carried forward
- * @returns Reconstructed HTML string and the updated list of open tags
- */
-export function tokensToHTML(tokens, openTags) {
-    let html = '';
-    const tagStack = [...openTags]; // Clone the open tags
-    // Prepend any open tags
-    for (const tag of openTags) {
-        html += tag.fullTag;
-    }
-    // Process tokens
-    for (const token of tokens) {
-        if (token.type === 'text') {
-            html += token.content;
-        }
-        else if (token.type === 'tag') {
-            html += token.content;
-            if (token.isClosing) {
-                // Remove from stack
-                const idx = tagStack.findIndex((t) => t.name === token.tagName);
-                if (idx !== -1) {
-                    tagStack.splice(idx, 1);
-                }
-            }
-            else {
-                // Add to stack
-                tagStack.push({
-                    name: token.tagName,
-                    fullTag: token.content,
-                });
-            }
-        }
-    }
-    // Close any remaining open tags for this line
-    for (let i = tagStack.length - 1; i >= 0; i--) {
-        html += `</${tagStack[i].name}>`;
-    }
-    return { html, openTags: tagStack };
-}

package/lib/text/render.d.ts

@@ -1,8 +0,0 @@
-import { TextElement, TextLine, LineRenderContext, SegmentRenderOptions, RenderSegment } from './types.js';
-export declare function drawListMarker(doc: PDFKit.PDFDocument, element: TextElement, line: TextLine, lineXOffset: number, lineYOffset: number, fonts: Record<string, boolean>, color: string, opacity: number, mode: 'fill' | 'stroke', textOptions: PDFKit.Mixins.TextOptions): Promise<void>;
-export declare function prepareLineForRendering(doc: PDFKit.PDFDocument, element: TextElement, line: TextLine, lineIndex: number, yOffset: number, lineHeightPx: number, cumulativeYOffset: number, textOptions: PDFKit.Mixins.TextOptions, fonts: Record<string, boolean>, markerColor: string, markerOpacity: number, markerMode: 'fill' | 'stroke'): Promise<LineRenderContext>;
-export declare function renderSegmentsForLine(doc: PDFKit.PDFDocument, element: TextElement, line: TextLine, renderSegments: RenderSegment[], context: LineRenderContext, textOptions: PDFKit.Mixins.TextOptions, options: SegmentRenderOptions): Promise<void>;
-export declare function renderPDFX1aStroke(doc: PDFKit.PDFDocument, element: TextElement, textLines: TextLine[], yOffset: number, lineHeightPx: number, textOptions: PDFKit.Mixins.TextOptions, fonts: Record<string, boolean>): Promise<void>;
-export declare function renderStandardStroke(doc: PDFKit.PDFDocument, element: TextElement, textLines: TextLine[], yOffset: number, lineHeightPx: number, textOptions: PDFKit.Mixins.TextOptions, fonts: Record<string, boolean>): Promise<void>;
-export declare function renderTextFill(doc: PDFKit.PDFDocument, element: TextElement, textLines: TextLine[], yOffset: number, lineHeightPx: number, textOptions: PDFKit.Mixins.TextOptions, fonts: Record<string, boolean>): Promise<void>;
-export declare function renderTextBackground(doc: PDFKit.PDFDocument, element: TextElement, verticalAlignmentOffset: number, textOptions: PDFKit.Mixins.TextOptions): void;