From db8615e5ff02b1ce08ddac173af2f9be9919eaf2 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?=E9=9B=B7=E7=94=B5=E8=8A=BD=E8=A1=A3?= Date: Tue, 6 Oct 2026 19:13:23 -0400 Subject: [PATCH] Read China screens below the logo in the harness OCR --- bin/omarchy-iso-test | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/bin/omarchy-iso-test b/bin/omarchy-iso-test index 2228ca5..2266247 100755 --- a/bin/omarchy-iso-test +++ b/bin/omarchy-iso-test @@ -261,9 +261,10 @@ ocr_screen() { screendump "$shot" [[ -s $shot ]] || return 0 magick "$shot" -colorspace gray -negate -resize 150% "$prepped" 2>/dev/null || return 0 - # tesseract spaces out CJK characters, so China runs also emit each line without spaces. + # The logo and the black strip under foot throw tesseract off CJK, so China runs also read between them, without spaces. if [[ $REGION == "china" ]]; then - tesseract "$prepped" - --psm 6 -l "$OCR_LANGS" 2>/dev/null | sed 'p; s/ //g' || true + magick "$shot" -gravity north -chop 0x35% -gravity south -chop 0x4% -colorspace gray -negate -resize 150% "$prepped.text.png" 2>/dev/null || return 0 + { tesseract "$prepped" - --psm 6 -l "$OCR_LANGS"; tesseract "$prepped.text.png" - --psm 6 -l "$OCR_LANGS"; } 2>/dev/null | sed 'p; s/ //g' || true else tesseract "$prepped" - --psm 6 -l "$OCR_LANGS" 2>/dev/null || true fi