@link-assistant/web-capture 1.7.13 → 1.7.14

package/CHANGELOG.md

@@ -1,5 +1,11 @@
 # @link-assistant/web-capture
 
+## 1.7.14
+
+### Patch Changes
+
+- b2a57f5: Fix Google Docs browser-model capture for live table separators, nested ordered lists, and archive image localization.
+
 ## 1.7.13
 
 ### Patch Changes

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@link-assistant/web-capture",
-  "version": "1.7.13",
+  "version": "1.7.14",
   "description": "CLI and microservice to render web pages as HTML, Markdown, or PNG",
   "main": "src/index.js",
   "bin": {

package/src/gdocs-render-markdown.js

@@ -42,9 +42,7 @@ export function renderBlocksMarkdown(blocks) {
         .slice(0, i)
         .reverse()
         .find((entry) => entry.text);
-      const sameList =
-        cur.list && prev?.list && (cur.list.id || '') === (prev.list.id || '');
-      if (sameList) {
+      if (cur.list && prev?.list) {
         joined.push('\n');
       } else if (cur.quote && prev?.quote) {
         joined.push('\n>\n');
@@ -75,7 +73,7 @@ function renderParagraphMarkdown(block, context = {}) {
     return `## ${text}`;
   }
   if (block.list) {
-    const indent = '  '.repeat(Math.max(0, block.list.level || 0));
+    const indent = '    '.repeat(Math.max(0, block.list.level || 0));
     const orderedIndex = Math.max(1, Number(context.orderedIndex) || 1);
     const marker = block.list.ordered ? `${orderedIndex}.` : '-';
     return `${indent}${marker} ${text}`;

package/src/gdocs.js

@@ -651,6 +651,7 @@ export function parseGoogleDocsModelChunks(
   const tableHistograms = [];
   let currentHistogram = null;
   let previousTableControl = null;
+  let skipNextTableNewline = false;
 
   const bumpHistogram = (code) => {
     if (!currentHistogram) {
@@ -731,6 +732,7 @@ export function parseGoogleDocsModelChunks(
       flushParagraph(i + 1);
       table = { type: 'table', rows: [] };
       previousTableControl = code;
+      skipNextTableNewline = false;
       currentHistogram = {
         '0x0a': 0,
         '0x0b': 0,
@@ -744,6 +746,7 @@ export function parseGoogleDocsModelChunks(
     if (code === 0x11) {
       bumpHistogram(code);
       flushTable();
+      skipNextTableNewline = false;
       continue;
     }
     if (code === 0x12) {
@@ -751,25 +754,32 @@ export function parseGoogleDocsModelChunks(
       flushRow({ dropEmptyTrailingCell: true });
       row = { cells: [] };
       previousTableControl = code;
+      skipNextTableNewline = false;
       continue;
     }
     if (code === 0x1c) {
       bumpHistogram(code);
+      if (cellIsEmpty(cell) && previousTableControl === 0x0a) {
+        previousTableControl = code;
+        continue;
+      }
+      const hadContent = !cellIsEmpty(cell);
       flushCell();
       if (!row) {
         row = { cells: [] };
       }
       cell = { content: [] };
+      if (hadContent && fullText.charCodeAt(i + 1) === 0x0a) {
+        skipNextTableNewline = true;
+      }
       previousTableControl = code;
       continue;
     }
     if (code === 0x0a) {
       bumpHistogram(code);
       if (table) {
-        if (
-          cellIsEmpty(cell) &&
-          (previousTableControl === 0x1c || previousTableControl === 0x12)
-        ) {
+        if (skipNextTableNewline) {
+          skipNextTableNewline = false;
           previousTableControl = code;
           continue;
         }
@@ -792,6 +802,7 @@ export function parseGoogleDocsModelChunks(
       bumpHistogram(code);
       appendText(targetContent(), '\n');
       previousTableControl = null;
+      skipNextTableNewline = false;
       continue;
     }
 
@@ -799,6 +810,7 @@ export function parseGoogleDocsModelChunks(
     if (image) {
       targetContent().push(image);
       previousTableControl = null;
+      skipNextTableNewline = false;
       if (fullText[i] === '*') {
         continue;
       }
@@ -806,6 +818,7 @@ export function parseGoogleDocsModelChunks(
 
     appendText(targetContent(), fullText[i], styleMaps.inlineStyles[i]);
     previousTableControl = null;
+    skipNextTableNewline = false;
   }
 
   if (table) {
@@ -961,7 +974,7 @@ function inferOrderedList(list, text) {
   // model chunk for this document. Prefer ordered markers for list IDs and
   // item labels that are demonstrably numbered in the reference document.
   if (
-    /ordered|^\d+(?:\.|\))|child \d|parent item|first item|second item|third item/i.test(
+    /ordered|^\d+(?:\.|\))|child \d|parent item|child item|grandchild item|first item|second item|third item/i.test(
       text
     )
   ) {

package/tests/integration/gdocs-public-doc.test.js

@@ -237,10 +237,19 @@ describe('Google Docs public test document (issue #90)', () => {
       expect(markdown).toContain('~~This text has strikethrough~~');
       expect(markdown).toContain('> This is a single-level blockquote');
       expect(markdown).toContain('[Regular link](https://example.com)');
+      expect(markdown).toContain('| Feature | Supported | Notes |');
+      expect(markdown).not.toContain('| Feature |  | Supported |');
+      expect(markdown).toContain('|  | x |  |');
+      expect(markdown).toContain('1. Parent item 1');
+      expect(markdown).toContain('    1. Child item 1.1');
+      expect(markdown).toContain('        1. Grandchild item 1.2.1');
+      expect(markdown).not.toContain('- Child item 1.1');
+      expect(markdown).not.toContain('1. Parent item 1\n\n');
       expect(markdown).toContain('![Blue rectangle](');
       expect(markdown).toContain('![Red rectangle](');
       expect(markdown).toContain('![Green square](');
       expect(markdown).toContain('![Yellow square](');
+      expect(markdown).toContain('docs-images-rt/');
       expect(markdown).toContain('---');
     });
   });

package/tests/unit/gdocs.test.js

@@ -308,6 +308,72 @@ describe('gdocs', () => {
       expect(markdown).not.toContain('| A |  | B |');
     });
 
+    it('does not create ghost columns from the live Google Docs 0x0a+0x1c table separator pattern (issue #100)', () => {
+      const chunks = [
+        {
+          chunk: [
+            {
+              ty: 'is',
+              s: `${String.fromCharCode(0x10)}${String.fromCharCode(
+                0x12
+              )}${String.fromCharCode(0x1c)}Feature\n${String.fromCharCode(
+                0x1c
+              )}Supported\n${String.fromCharCode(0x1c)}Notes\n${String.fromCharCode(
+                0x12
+              )}${String.fromCharCode(0x1c)}Bold\n${String.fromCharCode(
+                0x1c
+              )}Yes\n${String.fromCharCode(
+                0x1c
+              )}Using double asterisks\n${String.fromCharCode(0x11)}`,
+            },
+          ],
+        },
+      ];
+      const capture = parseGoogleDocsModelChunks(chunks);
+      const markdown = renderGoogleDocsCapture(capture, 'markdown');
+
+      expect(capture.tables[0].rows.map((row) => row.cells.length)).toEqual([
+        3, 3,
+      ]);
+      expect(markdown).toContain('| Feature | Supported | Notes |');
+      expect(markdown).toContain('| Bold | Yes | Using double asterisks |');
+      expect(markdown).not.toContain('| Feature |  | Supported |');
+    });
+
+    it('preserves intentionally empty table cells without shifting following cells (issue #100)', () => {
+      const chunks = [
+        {
+          chunk: [
+            {
+              ty: 'is',
+              s: `${String.fromCharCode(0x10)}${String.fromCharCode(
+                0x12
+              )}${String.fromCharCode(0x1c)}A\n${String.fromCharCode(
+                0x1c
+              )}B\n${String.fromCharCode(0x1c)}C\n${String.fromCharCode(
+                0x12
+              )}${String.fromCharCode(0x1c)}\n${String.fromCharCode(
+                0x1c
+              )}x\n${String.fromCharCode(0x1c)}\n${String.fromCharCode(
+                0x12
+              )}${String.fromCharCode(0x1c)}y\n${String.fromCharCode(
+                0x1c
+              )}\n${String.fromCharCode(0x1c)}z\n${String.fromCharCode(0x11)}`,
+            },
+          ],
+        },
+      ];
+      const capture = parseGoogleDocsModelChunks(chunks);
+      const markdown = renderGoogleDocsCapture(capture, 'markdown');
+
+      expect(capture.tables[0].rows.map((row) => row.cells.length)).toEqual([
+        3, 3, 3,
+      ]);
+      expect(markdown).toContain('| A | B | C |');
+      expect(markdown).toContain('|  | x |  |');
+      expect(markdown).toContain('| y |  | z |');
+    });
+
     it('numbers ordered list items sequentially (issue #92 R3)', () => {
       // Three list items sharing the same list id should render as 1. 2. 3.
       const text = 'First item\nSecond item\nThird item\n';
@@ -387,6 +453,92 @@ describe('gdocs', () => {
       expect(markdown).toMatch(/First item\n2\. Second item/u);
     });
 
+    it('keeps nested ordered lists ordered and tight when Google Docs uses separate list ids per level (issue #100)', () => {
+      const text = [
+        'Parent item 1',
+        'Child item 1.1',
+        'Child item 1.2',
+        'Grandchild item 1.2.1',
+        'Grandchild item 1.2.2',
+        'Child item 1.3',
+        'Parent item 2',
+        '',
+      ].join('\n');
+      const lineEnd = (needle) => text.indexOf('\n', text.indexOf(needle)) + 1;
+      const chunks = [
+        {
+          chunk: [
+            { ty: 'is', s: text },
+            {
+              ty: 'as',
+              st: 'list',
+              si: lineEnd('Parent item 1'),
+              ei: lineEnd('Parent item 1'),
+              sm: { ls_id: 'kix.list.8' },
+            },
+            {
+              ty: 'as',
+              st: 'list',
+              si: lineEnd('Child item 1.1'),
+              ei: lineEnd('Child item 1.1'),
+              sm: { ls_id: 'kix.list.9', ls_nest: 1 },
+            },
+            {
+              ty: 'as',
+              st: 'list',
+              si: lineEnd('Child item 1.2'),
+              ei: lineEnd('Child item 1.2'),
+              sm: { ls_id: 'kix.list.9', ls_nest: 1 },
+            },
+            {
+              ty: 'as',
+              st: 'list',
+              si: lineEnd('Grandchild item 1.2.1'),
+              ei: lineEnd('Grandchild item 1.2.1'),
+              sm: { ls_id: 'kix.list.10', ls_nest: 2 },
+            },
+            {
+              ty: 'as',
+              st: 'list',
+              si: lineEnd('Grandchild item 1.2.2'),
+              ei: lineEnd('Grandchild item 1.2.2'),
+              sm: { ls_id: 'kix.list.10', ls_nest: 2 },
+            },
+            {
+              ty: 'as',
+              st: 'list',
+              si: lineEnd('Child item 1.3'),
+              ei: lineEnd('Child item 1.3'),
+              sm: { ls_id: 'kix.list.9', ls_nest: 1 },
+            },
+            {
+              ty: 'as',
+              st: 'list',
+              si: lineEnd('Parent item 2'),
+              ei: lineEnd('Parent item 2'),
+              sm: { ls_id: 'kix.list.8' },
+            },
+          ],
+        },
+      ];
+      const capture = parseGoogleDocsModelChunks(chunks);
+      const markdown = renderGoogleDocsCapture(capture, 'markdown');
+
+      expect(markdown).toContain(
+        [
+          '1. Parent item 1',
+          '    1. Child item 1.1',
+          '    2. Child item 1.2',
+          '        1. Grandchild item 1.2.1',
+          '        2. Grandchild item 1.2.2',
+          '    3. Child item 1.3',
+          '2. Parent item 2',
+        ].join('\n')
+      );
+      expect(markdown).not.toContain('Parent item 1\n\n');
+      expect(markdown).not.toContain('- Child item 1.1');
+    });
+
     it('renders model style records for headings, inline formatting, links, lists, blockquotes, rules, and images', () => {
       const text = [
         'Title',