@fiduswriter/document 0.1.0-alpha.4 → 0.1.0-alpha.5

package/dist/importer/docx/convert.js

@@ -1,11 +1,29 @@
 import { MathMLToLaTeX } from "mathml-to-latex";
-import { xmlDOM } from "../../exporter/tools/xml.js";
-import { randomCommentId, randomFigureId, randomHeadingId } from "../../schema/common/index.js";
+import { randomFigureId, randomHeadingId } from "../../schema/common/index.js";
 import { isDocxBibliographyField, isDocxCitationField, isDocxSdtBibliography, isDocxSdtCitation, parseDocxFieldCitation, parseDocxSdtCitation } from "./citations.js";
 import { normalizeText } from "./helpers.js";
 import { omml2mathml } from "./omml2mathml.js";
 import { DocxParser } from "./parse.js";
+import { gettext } from "fwtoolkit";
+function attr(node, name) {
+    if (node && typeof node === "object" && "getAttribute" in node) {
+        return String(node.getAttribute(name) || "");
+    }
+    return "";
+}
 export class DocxConvert {
+    zip;
+    importId;
+    template;
+    bibliography;
+    images;
+    parser;
+    tracks;
+    currentTracks;
+    currentFields;
+    currentCommentIds;
+    sourcesXml;
+    referenceTargets;
     constructor(zip, importId, template, bibliography) {
         this.zip = zip;
         this.importId = importId;
@@ -18,6 +36,7 @@ export class DocxConvert {
         this.currentFields = [];
         this.currentCommentIds = [];
         this.sourcesXml = null;
+        this.referenceTargets = {};
     }
     async init() {
         await this.parser.init();
@@ -73,21 +92,21 @@ export class DocxConvert {
                 { type: "text", text: gettext("Untitled") }
             ]
         });
-        title.containerNodes.forEach(node => {
+        title.containerNodes.forEach((node) => {
             node.parentElement.removeChild(node);
         });
         document.attrs.title =
-            title.content.map(node => node.textContent).join("") ||
+            title.content.map((node) => node.textContent).join("") ||
                 gettext("Untitled");
         // Extract metadata sections
         const metadata = this.extractMetadata(body);
         metadata.forEach(({ type, content }) => {
-            const templatePart = templateParts.find(part => part.attrs.metadata === type);
+            const templatePart = templateParts.find((part) => part.attrs.metadata === type);
             const attrs = {};
-            if (templatePart.attrs.hidden) {
-                attrs.hidden = false;
-            }
             if (templatePart) {
+                if (templatePart.attrs?.hidden) {
+                    attrs.hidden = false;
+                }
                 document.content.push({
                     type: templatePart.type,
                     attrs: {
@@ -97,7 +116,7 @@ export class DocxConvert {
                     content: content.content
                 });
                 // Remove paragraphs from content so they are not added to body
-                content.containerNodes.forEach(node => {
+                content.containerNodes.forEach((node) => {
                     node.parentElement?.removeChild(node);
                 });
             }
@@ -105,7 +124,7 @@ export class DocxConvert {
         // Extract main content sections
         const sections = this.groupContentIntoSections(body);
         // Map sections to template parts
-        sections.forEach(section => {
+        sections.forEach((section) => {
             const templatePart = this.findMatchingTemplatePart(section.title, templateParts);
             if (templatePart) {
                 document.content.push({
@@ -126,17 +145,15 @@ export class DocxConvert {
         });
         // Add remaining content to body section
         const unassignedContent = sections
-            .filter(section => !this.findMatchingTemplatePart(section.title, templateParts))
-            .flatMap(section => section.content);
+            .filter((section) => !this.findMatchingTemplatePart(section.title, templateParts))
+            .flatMap((section) => section.content);
         if (unassignedContent.length) {
-            const bodyTemplatePart = templateParts.find(part => !part.attrs.metadata && part.type === "richtext_part");
+            const bodyTemplatePart = templateParts.find((part) => !part.attrs.metadata && part.type === "richtext_part");
             document.content.push({
                 type: "richtext_part",
                 attrs: {
-                    title: bodyTemplatePart
-                        ? bodyTemplatePart.attrs.title
-                        : "Body",
-                    id: bodyTemplatePart ? bodyTemplatePart.attrs.id : "body",
+                    title: bodyTemplatePart?.attrs?.title || "Body",
+                    id: bodyTemplatePart?.attrs?.id || "body",
                     marks: ["strong", "em", "link"]
                 },
                 content: unassignedContent
@@ -190,8 +207,8 @@ export class DocxConvert {
         }
         const properties = this.parser.customDoc.queryAll("property");
         const contributors = [];
-        properties.forEach(prop => {
-            const name = prop.getAttribute("name");
+        properties.forEach((prop) => {
+            const name = attr(prop, "name");
             if (!name || !name.startsWith("fidus_contributor_")) {
                 return;
             }
@@ -232,7 +249,7 @@ export class DocxConvert {
             }
         });
         const byRole = {};
-        contributors.forEach(contributor => {
+        contributors.forEach((contributor) => {
             if (!contributor) {
                 return;
             }
@@ -249,9 +266,9 @@ export class DocxConvert {
         // Try to find author information in metadata
         const authorNodes = body
             .queryAll("w:pStyle", { "w:val": "Author" })
-            .map(pStyle => pStyle.closest("w:p"))
-            .filter(p => p);
-        authorNodes.forEach(authorNode => {
+            .map((pStyle) => pStyle.closest("w:p"))
+            .filter((p) => p);
+        authorNodes.forEach((authorNode) => {
             const authorText = this.getTextContent(authorNode);
             const [firstname = "", lastname = ""] = authorText.split(" ", 2);
             authors.push({
@@ -271,7 +288,7 @@ export class DocxConvert {
             };
         }
         // Also check Creator in document properties
-        const creator = this.parser.coreDoc.query("dc:creator")?.textContent;
+        const creator = this.parser.coreDoc?.query("dc:creator")?.textContent;
         if (creator) {
             const [firstname = "", lastname = ""] = creator.split(" ", 2);
             return {
@@ -295,18 +312,18 @@ export class DocxConvert {
         // Look for section with Abstract style or heading
         const abstractNodes = body
             .queryAll("w:pStyle", { "w:val": "Abstract" })
-            .map(pStyle => pStyle.closest("w:p"))
-            .filter(p => p);
+            .map((pStyle) => pStyle.closest("w:p"))
+            .filter((p) => p);
         if (abstractNodes.length) {
             return {
-                content: abstractNodes.map(abstractNode => this.convertBlock(abstractNode)),
+                content: abstractNodes.map((abstractNode) => this.convertBlock(abstractNode)),
                 containerNodes: abstractNodes
             };
         }
         const extractedPart = this.extractPartOnTitle(body, ["Abstract"]);
         if (extractedPart.content.length) {
             return {
-                content: extractedPart.content.map(abstractNode => this.convertBlock(abstractNode)),
+                content: extractedPart.content.map((abstractNode) => this.convertBlock(abstractNode)),
                 containerNodes: extractedPart.content.concat([
                     extractedPart.header
                 ])
@@ -319,8 +336,8 @@ export class DocxConvert {
         // Look for keywords section or metadata
         let keywordNodes = body
             .queryAll("w:pStyle", { "w:val": "Keywords" })
-            .map(pStyle => pStyle.closest("w:p"))
-            .filter(p => p);
+            .map((pStyle) => pStyle.closest("w:p"))
+            .filter((p) => p);
         if (!keywordNodes.length) {
             // If no keywords section is found, look for a title called "Keywords"
             const extractedPart = this.extractPartOnTitle(body, ["Keywords", "Keywords:", "Keyword"], 1);
@@ -332,11 +349,11 @@ export class DocxConvert {
         if (keywordNodes) {
             return {
                 content: keywordNodes
-                    .map(keywordsNode => this.getTextContent(keywordsNode))
-                    .flatMap(str => str.split(/[,;|:]+/)) // Split on multiple separators
-                    .map(keyword => keyword.trim()) // Trim whitespace
-                    .filter(keyword => keyword.length > 0)
-                    .map(keyword => ({
+                    .map((keywordsNode) => this.getTextContent(keywordsNode))
+                    .flatMap((str) => str.split(/[,;|:]+/)) // Split on multiple separators
+                    .map((keyword) => keyword.trim()) // Trim whitespace
+                    .filter((keyword) => keyword.length > 0)
+                    .map((keyword) => ({
                     type: "tag",
                     attrs: {
                         tag: keyword
@@ -366,9 +383,9 @@ export class DocxConvert {
                 "Heading9"
             ]
         })
-            .map(pStyle => pStyle.closest("w:p"))
-            .filter(p => p);
-        const header = headingPars.find(p => titleWords.includes(this.getTextContent(p).trim()));
+            .map((pStyle) => pStyle.closest("w:p"))
+            .filter((p) => p);
+        const header = headingPars.find((p) => titleWords.includes(this.getTextContent(p).trim()));
         const content = [];
         if (header && header.nextSibling) {
             //const content = []
@@ -395,7 +412,7 @@ export class DocxConvert {
             content: []
         };
         const skippedBlocks = [];
-        body.children.forEach(node => {
+        body.children.forEach((node) => {
             if (skippedBlocks.includes(node)) {
                 return;
             }
@@ -445,12 +462,12 @@ export class DocxConvert {
             return null;
         }
         // Try exact match first
-        let matchingPart = templateParts.find(part => part.type === "richtext_part" &&
+        let matchingPart = templateParts.find((part) => part.type === "richtext_part" &&
             !part.attrs.metadata &&
             part.attrs.title.toLowerCase() === sectionTitle.toLowerCase());
         if (!matchingPart) {
             // Try fuzzy matching
-            matchingPart = templateParts.find(part => part.type === "richtext_part" &&
+            matchingPart = templateParts.find((part) => part.type === "richtext_part" &&
                 !part.attrs.metadata &&
                 this.isSimilarTitle(part.attrs.title, sectionTitle));
         }
@@ -465,15 +482,15 @@ export class DocxConvert {
     getTextContent(node) {
         return node
             .queryAll("w:t")
-            .map(t => t.textContent)
+            .map((t) => t.textContent)
             .join("");
     }
     extractTitle(body) {
         // First try to find paragraph with Title style
         const titlePars = body
             .queryAll("w:pStyle", { "w:val": "Title" })
-            .map(pStyle => pStyle.closest("w:p"))
-            .filter(p => p);
+            .map((pStyle) => pStyle.closest("w:p"))
+            .filter((p) => p);
         if (titlePars.length) {
             return {
                 content: this.convertInline(titlePars[0]),
@@ -495,8 +512,8 @@ export class DocxConvert {
                 "Heading9"
             ]
         })
-            .map(pStyle => pStyle.closest("w:p"))
-            .filter(p => p);
+            .map((pStyle) => pStyle.closest("w:p"))
+            .filter((p) => p);
         if (headingPars.length) {
             return {
                 content: this.convertInline(headingPars[0]),
@@ -589,22 +606,22 @@ export class DocxConvert {
         const insertion = node.query("w:pPr")?.query("w:ins");
         const tracks = [];
         if (insertion) {
-            const date = new Date(insertion.getAttribute("w:date"));
+            const date = new Date(attr(insertion, "w:date"));
             const date10 = Math.floor(date.getTime() / 60000) * 10;
             tracks.push({
                 type: "insertion",
                 user: 0, // Default user ID
-                username: insertion.getAttribute("w:author"),
+                username: attr(insertion, "w:author"),
                 date: date10
             });
         }
         if (deletion) {
-            const date = new Date(deletion.getAttribute("w:date"));
+            const date = new Date(attr(deletion, "w:date"));
             const date10 = Math.floor(date.getTime() / 60000) * 10;
             tracks.push({
                 type: "deletion",
                 user: 0, // Default user ID
-                username: deletion.getAttribute("w:author"),
+                username: attr(deletion, "w:author"),
                 date: date10
             });
         }
@@ -615,11 +632,11 @@ export class DocxConvert {
     }
     getParaStyle(node) {
         const pStyle = node.query("w:pStyle");
-        const styleId = pStyle?.getAttribute("w:val");
+        const styleId = attr(pStyle, "w:val");
         const style = this.parser.styles[styleId] || {};
         const numPr = node.query("w:numPr");
-        const numId = numPr?.query("w:numId")?.getAttribute("w:val");
-        const ilvl = parseInt(numPr?.query("w:ilvl")?.getAttribute("w:val") || "0");
+        const numId = attr(numPr?.query("w:numId"), "w:val");
+        const ilvl = parseInt(attr(numPr?.query("w:ilvl"), "w:val") || "0");
         return {
             ...style,
             numbering: numId
@@ -633,7 +650,7 @@ export class DocxConvert {
     }
     convertParagraph(node) {
         const pStyle = node.query("w:pStyle");
-        const styleId = pStyle?.getAttribute("w:val");
+        const styleId = attr(pStyle, "w:val");
         // Check if this is a code block (Code style or inherited from one)
         if (styleId &&
             (this.parser.isCodeStyle?.(styleId) || styleId === "Code")) {
@@ -696,7 +713,7 @@ export class DocxConvert {
         if (!blip) {
             return null;
         }
-        const rId = blip.getAttribute("r:embed");
+        const rId = attr(blip, "r:embed");
         const rel = this.parser.relationships[rId];
         if (!rel) {
             return null;
@@ -709,8 +726,8 @@ export class DocxConvert {
         // <a:ext cx="5753598" cy="4463556" />
         //
         const size = drawing.query("a:ext");
-        const width = parseInt(size.getAttribute("cx") || 0) / 9525; // In EMUs
-        const height = parseInt(size.getAttribute("cy") || 0) / 9525; // In EMUs
+        const width = parseInt(attr(size, "cx") || "0") / 9525; // In EMUs
+        const height = parseInt(attr(size, "cy") || "0") / 9525; // In EMUs
         const imageId = Math.floor(Math.random() * 1000000);
         this.images[imageId] = {
             id: imageId,
@@ -752,7 +769,7 @@ export class DocxConvert {
     convertInline(node) {
         const content = [];
         // We'll process all inline nodes in document order
-        node.children.forEach(child => {
+        node.children.forEach((child) => {
             let contentReceiver = content;
             const currentField = this.currentFields[this.currentFields.length - 1];
             if (currentField) {
@@ -774,7 +791,7 @@ export class DocxConvert {
                 if (fieldChar) {
                     let currentField;
                     let rendercurrentField = false;
-                    const type = fieldChar.getAttribute("w:fldCharType");
+                    const type = attr(fieldChar, "w:fldCharType");
                     if (type === "begin") {
                         currentField = {
                             status: "instruction",
@@ -807,14 +824,14 @@ export class DocxConvert {
                         currentField.data = fldDataNode.textContent || null;
                     }
                     if (rendercurrentField && currentField) {
-                        this.renderField(currentField).forEach(node => contentReceiver.push(node));
+                        this.renderField(currentField).forEach((node) => contentReceiver.push(node));
                     }
                     return;
                 }
                 // Process footnote references
                 const footnoteRef = child.query("w:footnoteReference");
                 if (footnoteRef) {
-                    const footnoteId = footnoteRef.getAttribute("w:id");
+                    const footnoteId = attr(footnoteRef, "w:id");
                     if (this.parser.footnotes[footnoteId]) {
                         contentReceiver.push(this.convertFootnote(footnoteId));
                     }
@@ -823,7 +840,7 @@ export class DocxConvert {
                 // Process endnote references
                 const endnoteRef = child.query("w:endnoteReference");
                 if (endnoteRef) {
-                    const endnoteId = endnoteRef.getAttribute("w:id");
+                    const endnoteId = attr(endnoteRef, "w:id");
                     if (this.parser.endnotes[endnoteId]) {
                         contentReceiver.push(this.convertFootnote(endnoteId, true));
                     }
@@ -852,14 +869,14 @@ export class DocxConvert {
                 });
             }
             else if (child.tagName === "w:commentRangeStart") {
-                const commentId = child.getAttribute("w:id");
+                const commentId = attr(child, "w:id");
                 if (commentId && this.parser.comments[commentId]) {
                     this.currentCommentIds.push(commentId);
                 }
                 return;
             }
             else if (child.tagName === "w:commentRangeEnd") {
-                const commentId = child.getAttribute("w:id");
+                const commentId = attr(child, "w:id");
                 if (commentId) {
                     const index = this.currentCommentIds.indexOf(commentId);
                     if (index !== -1) {
@@ -875,14 +892,14 @@ export class DocxConvert {
             }
             else if (child.tagName === "w:hyperlink") {
                 // Process hyperlink
-                const rId = child.getAttribute("r:id");
-                const anchor = child.getAttribute("w:anchor");
+                const rId = attr(child, "r:id");
+                const anchor = attr(child, "w:anchor");
                 const relationship = rId ? this.parser.relationships[rId] : null;
                 const href = relationship?.target || (anchor ? `#${anchor}` : null);
                 if (href) {
                     const runs = child.queryAll("w:r");
                     const text = runs
-                        .map(run => run.query("w:t")?.textContent || "")
+                        .map((run) => run.query("w:t")?.textContent || "")
                         .join("");
                     if (text) {
                         // Check if this is an internal link (bookmark reference) that should be a cross-reference
@@ -942,9 +959,9 @@ export class DocxConvert {
     findReferenceTargets(document) {
         const targets = {};
         // Find bookmarks
-        document.queryAll("w:bookmarkStart").forEach(bookmark => {
-            const id = bookmark.getAttribute("w:id");
-            const name = bookmark.getAttribute("w:name");
+        document.queryAll("w:bookmarkStart").forEach((bookmark) => {
+            const id = attr(bookmark, "w:id");
+            const name = attr(bookmark, "w:name");
             if (id && name) {
                 targets[name] = {
                     id: name,
@@ -953,8 +970,8 @@ export class DocxConvert {
             }
         });
         // Find headings (with styles like Heading1, Heading2, etc.)
-        document.queryAll("w:pStyle").forEach(pStyle => {
-            const val = pStyle.getAttribute("w:val");
+        document.queryAll("w:pStyle").forEach((pStyle) => {
+            const val = attr(pStyle, "w:val");
             if (val && val.match(/^Heading\d+$/)) {
                 const paragraph = pStyle.closest("w:p");
                 if (paragraph) {
@@ -981,11 +998,11 @@ export class DocxConvert {
             : this.parser.footnotes[id].content;
         // Convert the footnote content to our document model
         const content = [];
-        footnoteContent.forEach(block => {
+        footnoteContent.forEach((block) => {
             if (block.type === "paragraph") {
                 content.push({
                     type: "paragraph",
-                    content: block.content.map(node => {
+                    content: block.content.map((node) => {
                         if (node.type === "text") {
                             return {
                                 type: "text",
@@ -1171,42 +1188,42 @@ export class DocxConvert {
                 /^ubuntu mono/i,
                 /^droid sans mono/i
             ];
-            const isMonospace = monospacePatterns.some(pattern => pattern.test(formatting.fontFamily));
+            const isMonospace = monospacePatterns.some((pattern) => pattern.test(formatting.fontFamily));
             if (isMonospace) {
                 marks.push({ type: "code" });
             }
         }
         if (insertion) {
-            const date = new Date(insertion.getAttribute("w:date"));
+            const date = new Date(attr(insertion, "w:date"));
             const date10 = Math.floor(date.getTime() / 600000) * 10;
             marks.push({
                 type: "insertion",
                 attrs: {
                     user: 0,
-                    username: insertion.getAttribute("w:author"),
+                    username: attr(insertion, "w:author"),
                     date: date10,
                     approved: false
                 }
             });
         }
         if (deletion) {
-            const date = new Date(deletion.getAttribute("w:date"));
+            const date = new Date(attr(deletion, "w:date"));
             const date10 = Math.floor(date.getTime() / 600000) * 10;
             marks.push({
                 type: "deletion",
                 attrs: {
                     user: 0,
-                    username: deletion.getAttribute("w:author"),
+                    username: attr(deletion, "w:author"),
                     date: date10
                 }
             });
         }
         return marks;
     }
-    getCurrentMarks(formatting, insertion, deletion) {
+    getCurrentMarks(formatting, insertion = null, deletion = null) {
         const marks = this.createMarksFromFormatting(formatting, insertion, deletion);
         // Add comment marks for any active comment IDs
-        this.currentCommentIds.forEach(commentId => {
+        this.currentCommentIds.forEach((commentId) => {
             marks.push({
                 type: "comment",
                 attrs: {
@@ -1220,7 +1237,7 @@ export class DocxConvert {
         return Boolean(doc.query("w:ins") || doc.query("w:del"));
     }
     detectLanguage(doc) {
-        return doc.query("w:lang")?.getAttribute("w:val") || "en-US";
+        return attr(doc.query("w:lang"), "w:val") || "en-US";
     }
 }
 //# sourceMappingURL=convert.js.map