extern/pdf-to-markdown
1
0
Fork 0
mirror of https://github.com/jzillmann/pdf-to-markdown.git synced 2024-11-21 23:33:31 +01:00
Code Issues Packages Projects Releases Wiki Activity

[WIP] don’t make paragraph bolds to headline

Browse Source
This commit is contained in:
Johannes Zillmann 2017-03-24 08:06:54 +01:00
parent e19294f35f
commit 81518a857b
1 changed files with 3 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/javascript/models/transformations/textitem/DetectHeaders.jsx
View File

@ -91,16 +91,19 @@ export default class DetectHeaders extends ToTextItemTransformation {
if (smallesHeadlineLevel < 6) { if (smallesHeadlineLevel < 6) {
const nextHeadlineType = headlineByLevel(smallesHeadlineLevel + 1); const nextHeadlineType = headlineByLevel(smallesHeadlineLevel + 1);
parseResult.pages.forEach(page => { parseResult.pages.forEach(page => {
var lastItem;
page.items.forEach(textItem => { page.items.forEach(textItem => {
if (!textItem.type if (!textItem.type
&& textItem.height == mostUsedHeight && textItem.height == mostUsedHeight
&& textItem.font !== mostUsedFont && textItem.font !== mostUsedFont
&& (!lastItem || lastItem.y < textItem.y || (lastItem.type && lastItem.type.headline) || (lastItem.y - textItem.y > mostUsedDistance * 2))
&& textItem.text === textItem.text.toUpperCase() && textItem.text === textItem.text.toUpperCase()
) { ) {
detectedHeaders++; detectedHeaders++;
textItem.annotation = DETECTED_ANNOTATION; textItem.annotation = DETECTED_ANNOTATION;
textItem.type = nextHeadlineType; textItem.type = nextHeadlineType;
} }
lastItem = textItem;
}); });
}); });
} }