Updated PDFs to render in higher resolution

scribeocr · Sep 10, 2024 · 0dd8801 · 0dd8801
1 parent a23a329
commit 0dd8801
Show file tree

Hide file tree

Showing 2 changed files with 2 additions and 2 deletions.
diff --git a/js/containers/imageContainer.js b/js/containers/imageContainer.js
@@ -525,7 +525,7 @@ export class ImageCache {
 
     // For reasons that are unclear, a small number of pages have been rendered into massive files
     // so a hard-cap on resolution must be imposed.
-    const pageDPI = ImageCache.pdfDims300.map((x) => 300 * 2000 / x.width, 2000);
+    const pageDPI = ImageCache.pdfDims300.map((x) => 300 * Math.min(x.width, 3500) / x.width);
 
     // In addition to capping the resolution, also switch the width/height
     ImageCache.pdfDims300.forEach((x, i) => {

diff --git a/js/extractPDFText.js b/js/extractPDFText.js
@@ -21,7 +21,7 @@ const extractInternalPDFTextRaw = async () => {
   };
 
   const stextArr = /** @type {Array<string>} */ ([]);
-  const pageDPI = ImageCache.pdfDims300.map((x) => 300 * 2000 / x.width, 2000);
+  const pageDPI = ImageCache.pdfDims300.map((x) => 300 * Math.min(x.width, 3500) / x.width);
   const resArr = pageDPI.map(async (x, i) => {
     // While using `pageTextJSON` would save some parsing, unfortunately that format only includes line-level granularity.
     // The XML format is the only built-in mupdf format that includes character-level granularity.