split-kochabo.sh: add support for obstbox recipes

Signed-off-by: Florian Pritz <bluewind@xinu.at>

1 files changed, 50 insertions, 14 deletions

diff --git a/split-kochabo.sh b/split-kochabo.sh
index 2878c4d..94081ec 100755
--- a/split-kochabo.sh
+++ b/split-kochabo.sh
@@ -3,10 +3,14 @@
 # This script can extract recipes from the PDFs mailed to customers of
 # kochabo.at.
 # Recipes from 2013 need extra processing for the file name in get_name().
+# Recipes before 2014-18 (that's week 18) also need less processing (no obstbox recipe)
 #
 
 set -e
 
+TMPDIR="`mktemp -d "/tmp/${0##*/}.XXXXXX"`"
+trap "rm -rf '${TMPDIR}'" EXIT TERM
+
 extract_pages() {
 	local first=$1
 	local last=$2
@@ -16,31 +20,63 @@ extract_pages() {
 	gs -dNOPAUSE -dQUIET -dBATCH -dFirstPage=$first -dLastPage=$last -sOutputFile="$output" -sDEVICE=pdfwrite "$@"
 }
 
+get_obstbox_name() {
+	pdftotext "$1" /dev/stdout | sed -n '3p' | paste -sd " "
+}
+
 get_name() {
 	pdftotext "$1" /dev/stdout | grep -B10 "Nährwertangaben pro Person" | sed '/[1-5]$/d; $d' | paste -sd " "
 	# needed for older files with all uppercase names
 	#| awk '{print tolower($0)}' sed 's/\b\(.\)/\u\1/g; s#\bMit\b#mit#g; s#\bUnd\b#und#g; s#\bAuf\b#auf#g; s#\s\+$##;'
 }
 
+extract_obstbox() {
+	tmpfile="$TMPDIR/rezept-obstbox.pdf"
+
+	extract_pages 4 4 "$tmpfile" "$file"
+	name="$(get_obstbox_name "$tmpfile")"
+
+	if [[ -z $name ]]; then
+		rm "$tmpfile"
+		continue
+	fi
+
+	mv "$tmpfile" "extracted/$name.pdf"
+
+	pdfimages -j "extracted/$name.pdf" "$TMPDIR/$name"
+	mv "$TMPDIR/$name-002.jpg" "extracted/$name.jpg"
+	rm -f $TMPDIR/*.{jpg,ppm}
+}
+
+extract_recipe() {
+	local i=$1
+	local tmpfile="$TMPDIR/rezept-$i.pdf"
+
+	# extract 2 pages for each $i starting on page 5
+	extract_pages $((skip_pages+((i-1)*2)+1)) $((skip_pages+((i-1)*2)+2)) "$tmpfile" "$file"
+	name="$(get_name "$tmpfile")"
+
+	if [[ -z $name ]]; then
+		rm "$tmpfile"
+		continue
+	fi
+
+	mv "$tmpfile" "extracted/$name.pdf"
+
+	pdfimages -j "extracted/$name.pdf" "$TMPDIR/$name"
+	mv "$TMPDIR/$name-000.jpg" "extracted/$name.jpg"
+	rm -f $TMPDIR/*.{jpg,ppm}
+}
+
 mkdir -p extracted
 
 skip_pages=4
 for file; do
-	for i in 1 2 3 4 5; do
-		extract_pages $((skip_pages+((i-1)*2)+1)) $((skip_pages+((i-1)*2)+2)) rezept-$i.pdf "$file"
-		src=rezept-$i.pdf
-		name="$(get_name $src)"
-
-		if [[ -z $name ]]; then
-			rm "rezept-$i.pdf"
-			continue
-		fi
+	extract_obstbox
 
-		mv "$src" "extracted/$name.pdf"
-
-		pdfimages -j "extracted/$name.pdf" "extracted/$name"
-		rm -f extracted/*-{001,002,003,004}.{jpg,ppm}
-		mv "extracted/$name-000.jpg" "extracted/$name.jpg"
+	# extract normal recipes
+	for i in 1 2 3 4 5; do
+		extract_recipe $i
 	done
 done