Invert the Y coordinate of each item to make PDF.js happy

g0v · Jul 13, 2015 · 01c4dda · 01c4dda
1 parent 091c598
commit 01c4dda
Show file tree

Hide file tree

Showing 2 changed files with 11 additions and 1 deletion.
diff --git a/src/index.js b/src/index.js
@@ -13,7 +13,8 @@ export default function parse(pdfData, options = {}) {
         return;
       }
 
-      machine.push(word.text, pageIdx + 1, [word.xMin, word.yMin]);
+      // y coordinate must be inverted, since PDF.js uses bottom-left as (0,0).
+      machine.push(word.text, pageIdx + 1, [word.xMin, page.height - word.yMin]);
     });
   }
 

diff --git a/test/unit/indexSpec.js b/test/unit/indexSpec.js
@@ -53,4 +53,13 @@ describe('parse', () => {
     expect(parsedData[0].items[0].text.slice(0, 4)).to.equal('產業城市');
 
   });
+
+  it('should invert Y coordinate of each bounding box', () => {
+    // PDF.js uses bottom-left corner as (0,0), thus inverting the Y-coord here.
+
+    const pdfData = require('../fixture/桃園1040417.json'),
+          parsedData = parser(pdfData);
+
+    expect(parsedData[0].coord[1]).to.equal(842 - 98.279297);
+  });
 });