From 0f16bb582b95942d20900ac032ca9dcba8f545b7 Mon Sep 17 00:00:00 2001
From: Balearica <admin@scribeocr.com>
Date: Mon, 28 Oct 2024 04:50:38 -0700
Subject: [PATCH] Prevented calcSuppFontInfo from running with offscreen text

---
 js/fontSupp.js | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/js/fontSupp.js b/js/fontSupp.js
index d0947cc..63a0dac 100644
--- a/js/fontSupp.js
+++ b/js/fontSupp.js
@@ -86,6 +86,9 @@ export const calcSuppFontInfo = async (ocrArr) => {
         if (word.font) {
           if (skipFonts.has(word.font)) {
             continue;
+          // Printing words off screen is a common method of hiding text in PDFs.
+          } else if (word.bbox.left < 0 || word.bbox.top < 0 || word.bbox.right > page.dims.width || word.bbox.bottom > page.dims.height) {
+            continue;
           } else if (!calcFonts.has(word.font)) {
             const sansSerifUnknown = determineSansSerif(word.font) === 'Default';
             if (sansSerifUnknown || !word.visualCoords) {