thumbnailify! because mp4 doesn't specify thumbnails, i'm sick of the blank black, and discord just uses the first frame

bash/cartographize.sh

@@ -1,8 +1,8 @@
 #!/bin/bash
 
-#
+echo "#"
-# phase 1: take screenshots
+echo "# phase 1: take screenshots"
-#
+echo "#"
 
 if [ ! -d "screenshots" ]; then
   mkdir screenshots
@@ -13,12 +13,12 @@ do
 	filename=$(basename -- "$file")
 	#extension="${filename##*.}"
 	filename="${filename%.*}"
-	ffmpeg -ss 0 -i $file -vf fps=0.5 ./screenshots/$filename%d.png
+	ffmpeg -ss 0 -i $file -vf fps=0.5 ./screenshots/$filename-%d.png
 done
 
-#
+echo "#"
-# phase 2: img preprocess
+echo "# phase 2: img preprocess"
-#
+echo "#"
 
 if [ ! -d "preproccessed" ]; then
   mkdir preproccessed
@@ -28,12 +28,12 @@ do
 	filename=$(basename -- "$file")
 	extension="${filename##*.}"
 	filename="${filename%.*}"
-	convert "$file" -gravity SouthWest -crop 560x60+0+0  -fuzz 12% +transparent "#edf1c9" "preproccessed/$filename.png"
+	convert "$file" -gravity SouthWest -crop 560x48+0+8  -fuzz 12% +transparent "#edf1c9" "preproccessed/$filename.png"
 done
 
-#
+echo "#"
-# phase 3: ocr
+echo "# phase 3: ocr"
-#
+echo "#"
 
 if [ ! -d "ocr" ]; then
   mkdir ocr
@@ -43,14 +43,14 @@ do
 	filename=$(basename -- "$file")
 	extension="${filename##*.}"
 	filename="${filename%.*}"
-	tesseract $file "ocr/$filename"
+	tesseract $file "ocr/$filename" -c tessedit_char_whitelist=" 0123456789KM/HNW."	
+	sed -i 's/ //g' "ocr/$filename.txt"
+
 done
 
-#
+echo "#"
-# phase 4: prep for geotagging
+echo "# phase 4: prep for geotagging"
-#
+echo "#"
-
-
 if [ ! -d "tagged" ]; then
   mkdir tagged
 fi
@@ -60,5 +60,38 @@ do
 	extension="${filename##*.}"
 	filename="${filename%.*}"
 	convert $file "tagged/$filename.jpg"
-	exiftool -overwrite_original -TagsFromFile ./knowngood.jpg -All:All "tagged/$filename.jpg"
+done
-done
+exiftool -overwrite_original -TagsFromFile ./knowngood.jpg -All:All "tagged/*.jpg"
+
+echo "#"
+echo "# phase 5: if ocr was good, geotag. Otherwise, reject."
+echo "#"
+goodreads=0
+badreads=0
+if [ ! -d "badread" ]; then
+  mkdir badread
+fi
+find "ocr" -iname "*.txt" -print0 | while read -d $'\0' file
+do
+	filename=$(basename -- "$file")
+	extension="${filename##*.}"
+	filename="${filename%.*}"
+	
+	n=$(grep -Poh "(?<=N)[0-9]{2}\\.[0-9]+" "$file")
+	w=$(grep -Poh "(?<=W)[0-9]{2}\\.[0-9]+" "$file")
+
+	if [ -n "$n" ] && [ -n "$w" ]; then
+		goodreads=$(($goodreads+1))
+		echo "$filename was OCR'd well; $goodreads good so far"
+		exiftool -overwrite_original -gpslatitude="$n" -gpslongitude="-$w" -GPSLatitudeRef="North" -GPSLongitudeRef="West" "tagged/$filename.jpg"
+	else
+		badreads=$(($badreads+1))
+		echo "$filename is not perfect; $badreads bad so far"
+		mv "tagged/$filename.jpg" "badread/$filename.jpg"
+		mv "ocr/$filename.txt" "badread/$filename.txt"
+		mv "preproccessed/$filename.png" "badread/$filename.png"
+	fi
+done
+
+echo "$goodreads good, $badreads bad. accuracy:"
+bc <<< "scale=4; goodreads=$goodreads; badreads=$badreads; ratio=100*goodreads/(goodreads+badreads);ratio"

bash/thumbnailify.sh

@@ -0,0 +1,34 @@
+#!/bin/bash
+
+if [ ! -f "$@" ]; then
+	echo "file to thumbnailify plz"
+fi
+
+fullfilename=$(basename -- "$@")
+extension="${fullfilename##*.}"
+filename="${fullfilename%.*}"
+
+pushd $(dirname "$@")
+
+if [ -f "$filename""_thumbless.$extension" ]; then
+	echo "there's already a _thumbless in here... I think i'm already done?"
+	exit 1
+fi
+
+if [ -f "testframe.mp4" ] || [ -f "thumbnailism.txt" ]; then
+	echo "my temp files are already here... I think i'm redundant?"
+	exit 1
+fi
+
+echo "file testframe.mp4" > thumbnailism.txt
+echo "file $fullfilename" >> thumbnailism.txt
+
+ffmpeg -i "$fullfilename" -ss 1.5 -to 1.51666 testframe.mp4
+ffmpeg -f concat -i thumbnailism.txt -c copy output.mp4
+
+mv "$fullfilename" "$filename""_thumbless.$extension"
+mv output.mp4 "$fullfilename"
+rm thumbnailism.txt
+rm testframe.mp4
+
+popd