extern/pdf-to-markdown
1
0
Fork 0
mirror of https://github.com/jzillmann/pdf-to-markdown.git synced 2025-06-27 04:51:37 +02:00
Code Issues Packages Projects Releases Wiki Activity

Base markdown converter on blocks

Browse Source
This commit is contained in:
Johannes Zillmann 2024-04-02 19:57:53 -06:00
parent 16e5a62951
commit 78f44a0ad9
2 changed files with 29 additions and 21 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

46
src/convert/MarkdownConverter.ts
View File

@ -1,38 +1,42 @@
import { groupByLine } from '../support/groupingUtils';
import { groupByBlock, groupByLine } from '../support/groupingUtils';
import Item from '../Item';
import { Converter } from '../convert';
import ChangeTracker from '../debug/ChangeTracker';
import EvaluationTracker from '../debug/EvaluationTracker';
import LineItemMerger from '../debug/LineItemMerger';
import { TextType, headlineLevel } from '../text-types';
export default class MarkdownConverter implements Converter {
convert(items: Item[]) {
let content = '';
const itemsByLine = groupByLine(items);
const lineMerger = new LineItemMerger();
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')) {
content += '# ' + itemText + '\n';
} else if (types.includes('H2')) {
content += '## ' + itemText + '\n';
} else if (types.includes('H3')) {
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 {
content += itemText;
}
groupByBlock(items).forEach((blockItems) => {
const types = blockItems[0].data['types'] || [];
groupByLine(blockItems).forEach((lineItems) => {
const evaluationTracker = new EvaluationTracker();
const changeTracker = new ChangeTracker();
const lineItem = lineMerger.merge(evaluationTracker, changeTracker, ['types'], lineItems);
const itemText = lineItem.data['str'];
content += elementToText(itemText, types[0]);
});
content += '\n';
});
return content;
}
}
function elementToText(text: string, type: TextType) {
switch (type) {
case 'H1':
case 'H2':
case 'H3':
case 'H4':
case 'H5':
case 'H6':
return '#'.repeat(headlineLevel(type)) + ' ' + text + '\n';
default:
return text + '\n';
}
}

4
src/support/groupingUtils.ts
View File

@ -72,6 +72,10 @@ export function groupByPage(items: Item[]): Item[][] {
return groupBy(items, (item) => item.page);
}
export function groupByBlock(items: Item[]): Item[][] {
return groupByElement(items, 'block');
}
export function groupByLine(items: Item[]): Item[][] {
return groupByElement(items, 'line');
}