extern/pdf-to-markdown
1
0
Fork 0
mirror of https://github.com/jzillmann/pdf-to-markdown.git synced 2025-02-05 04:19:20 +01:00
Code Issues Packages Projects Releases Wiki Activity

Improve markdown rendering

Browse Source
This commit is contained in:
Johannes Zillmann 2024-03-28 11:56:16 -06:00
parent 55ae236928
commit 5ab8730a4b
1 changed files with 20 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
src/convert/MarkdownConverter.ts
View File

@ -1,24 +1,38 @@
import Item from 'src/Item'; import { groupByLine } from '../support/groupingUtils';
import { Converter } from 'src/convert'; import Item from '../Item';
import { Converter } from '../convert';
import ChangeTracker from '../debug/ChangeTracker';
import EvaluationTracker from '../debug/EvaluationTracker';
import LineItemMerger from '../debug/LineItemMerger';
export default class MarkdownConverter implements Converter { export default class MarkdownConverter implements Converter {
convert(items: Item[]) { convert(items: Item[]) {
let content = ''; let content = '';
items.forEach((item) => { const itemsByLine = groupByLine(items);
const types = item.data['types'] || []; const lineMerger = new LineItemMerger();
const itemText = item.data['str']; itemsByLine.forEach((lineItems) => {
const lineItem = lineMerger.merge(new EvaluationTracker(), new ChangeTracker(), ['types'], lineItems);
const types = lineItem.data['types'] || [];
const itemText = lineItem.data['str'];
if (types.includes('H1')) { if (types.includes('H1')) {
content += '# ' + itemText + '\n'; content += '# ' + itemText + '\n';
} else if (types.includes('H2')) { } else if (types.includes('H2')) {
content += '## ' + itemText + '\n'; content += '## ' + itemText + '\n';
} else if (types.includes('H3')) { } else if (types.includes('H3')) {
content += '## ' + itemText + '\n'; content += '### ' + itemText + '\n';
} else if (types.includes('H4')) {
content += '#### ' + itemText + '\n';
} else if (types.includes('H5')) {
content += '##### ' + itemText + '\n';
} else if (types.includes('H6')) {
content += '###### ' + itemText + '\n';
} else { } else {
content += itemText; content += itemText;
} }
content += '\n'; content += '\n';
}); });
return content; return content;
} }
} }