extern/pdf-to-markdown
1
0
Fork 0
mirror of https://github.com/jzillmann/pdf-to-markdown.git synced 2025-02-09 06:19:39 +01:00
Code Issues Packages Projects Releases Wiki Activity

Add font info to pdf page view

Browse Source
This commit is contained in:
Johannes Zillmann 2017-02-14 20:28:01 +01:00
parent ab5705cd27
commit 92a4337387
3 changed files with 29 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
src/javascript/components/LoadingView.jsx
View File

@ -55,6 +55,17 @@ export default class LoadingView extends React.Component {
pdfDocument.getPage(j).then(function(page) { pdfDocument.getPage(j).then(function(page) {
var scale = 1.0; var scale = 1.0;
var viewport = page.getViewport(scale); var viewport = page.getViewport(scale);
// pdfDocument.getMetadata().then(function(data) {
// console.debug(data);
// });
// page.getOperatorList().then(function(opList) {
// console.debug(opList);
// // var svgGfx = new svgLib.SVGGraphics(page.commonObjs, page.objs);
// // return svgGfx.getSVG(opList, viewport).then(function (svg) {
// // container.appendChild(svg);
// // });
// });
page.getTextContent().then(function(textContent) { page.getTextContent().then(function(textContent) {
const textItems = textContent.items.map(function(item) { const textItems = textContent.items.map(function(item) {
const tx = PDFJS.Util.transform( // eslint-disable-line no-undef const tx = PDFJS.Util.transform( // eslint-disable-line no-undef
@ -64,12 +75,17 @@ export default class LoadingView extends React.Component {
const fontHeight = Math.sqrt((tx[2] * tx[2]) + (tx[3] * tx[3])); const fontHeight = Math.sqrt((tx[2] * tx[2]) + (tx[3] * tx[3]));
const dividedHeight = item.height / fontHeight; const dividedHeight = item.height / fontHeight;
const style = textContent.styles[item.fontName];
return new TextItem({ return new TextItem({
x: item.transform[4], x: item.transform[4],
y: item.transform[5], y: item.transform[5],
width: item.width, width: item.width,
height: dividedHeight <= 1 ? item.height : dividedHeight, height: dividedHeight <= 1 ? item.height : dividedHeight,
text: item.str text: item.str,
font: item.fontName,
fontAscent: style.ascent,
fontDescent: style.descent
}); });
}); });
anouncePageParsedFunction(page.pageIndex, textItems); anouncePageParsedFunction(page.pageIndex, textItems);
@ -79,9 +95,6 @@ export default class LoadingView extends React.Component {
}); });
} }
render() { render() {
const {parsedPages, pdfPages} = this.state; const {parsedPages, pdfPages} = this.state;
var percentDone = 0; var percentDone = 0;

9
src/javascript/components/debug/PdfPageView.jsx
View File

@ -41,6 +41,10 @@ export default class PdfPageView extends React.Component {
<th> <th>
Height Height
</th> </th>
<th>
Font
<br/>(asc/desc)
</th>
</tr> </tr>
</thead> </thead>
<tbody> <tbody>
@ -75,6 +79,11 @@ export default class PdfPageView extends React.Component {
<td> <td>
{ textItem.height } { textItem.height }
</td> </td>
<td>
{ textItem.font }
<br/>
{ textItem.fontAscent + ' / ' + textItem.fontDescent }
</td>
</tr> </tr>
) } ) }
</tbody> </tbody>

3
src/javascript/models/TextItem.jsx
View File

@ -7,6 +7,9 @@ export default class TextItem {
this.width = options.width; this.width = options.width;
this.height = options.height; this.height = options.height;
this.text = options.text; this.text = options.text;
this.font = options.font;
this.fontAscent = options.fontAscent;
this.fontDescent = options.fontDescent;
this.annotation = options.annotation; this.annotation = options.annotation;
this.markdownElement = options.markdownElement; this.markdownElement = options.markdownElement;
} }