mu-ref/pdf-inspector
8
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

fix(detector): raise text-ops threshold on pages with images

Browse Source 
Pages with images AND very few text operators (<10) are image pages
with overlay text (headers/footers), not real text pages. This fixes
image-heavy PDFs like newspaper ads being misclassified as TextBased.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Abimael Martell
2026-03-09 14:37:00 -07:00
co-authored by Claude Opus 4.6
parent f2869fff30
commit f3c00678b2
1 changed files with 6 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/detector.rs
+6 -1
View File
@@ -197,7 +197,12 @@ pub(crate) fn detect_from_document(
pages_actually_sampled += 1; pages_actually_sampled += 1;
let is_image_dominated = analysis.image_count > 10 let is_image_dominated = analysis.image_count > 10
&& analysis.image_count > analysis.text_operator_count * 3; && analysis.image_count > analysis.text_operator_count * 3;
if analysis.text_operator_count >= config.min_text_ops_per_page let effective_min_ops = if analysis.has_images || analysis.image_count > 0 {
config.min_text_ops_per_page.max(10)
} else {
config.min_text_ops_per_page
};
if analysis.text_operator_count >= effective_min_ops
&& !is_image_dominated && !is_image_dominated
&& analysis.unique_text_chars >= 5 && analysis.unique_text_chars >= 5
&& !analysis.has_vector_text && !analysis.has_vector_text