From 573b1df2daa5ab99e1ea6cae4c57f608bfd06fcc Mon Sep 17 00:00:00 2001
From: Stefan Armbruster <stefan@armbruster-it.de>
Date: Mon, 29 Oct 2018 13:55:03 +0100
Subject: [PATCH] Remove empty pages based on cmd line option

---
 README.md    |  8 ++++++
 scan         | 37 +++++++++++++++++++--------
 scan_perpage | 72 ++++++++++++++++++++++++++++++----------------------
 3 files changed, 77 insertions(+), 40 deletions(-)

diff --git a/README.md b/README.md
index 1a4287f..2bbe87f 100644
--- a/README.md
+++ b/README.md
@@ -24,6 +24,7 @@ Tested and run regularly on Fedora, but should work on other distributions with
 * pnmtops (netpbm-progs)
 * ps2pdf (ghostscript)
 * pdfunite
+* ImageMagick (if --skip-empty-pages) is used
 
 ### Optional
 
@@ -68,9 +69,16 @@ OPTIONS
    Run post-processing deskew and black edge detection (requires unpaper)
  --ocr
    Run OCR to make the PDF searchable (requires tesseract)
+ --skip-empty-pages
+   remove empty pages from resulting PDF document (e.g. one sided doc in duplex mode)
 
 OUTPUT
  -o, --output <outputfile>
    Output to named file default=scan.pdf
  -l, --outputlist <outputfile-1...outputfile-n> Output to named files for each scanned page, can be used with append
 ```
+
+## Contributors
+
+* [Raman Gupta](https://github.com/rocketraman/)
+* [Stefan Armbruster](https://github.com/sarmbruster/)
diff --git a/scan b/scan
index 3d49ffb..356c951 100755
--- a/scan
+++ b/scan
@@ -12,6 +12,7 @@ SCRIPT="$DIR/scan_perpage"
 DUPLEX=0
 UNPAPER=0
 SEARCHABLE=0
+LANGUAGE=eng
 MAXPAGE=
 TRUNCPAGE=0
 HELP=0
@@ -23,6 +24,9 @@ PGWIDTHIN=
 CROP=0
 DESKEW=0
 VERBOSE=0
+SKIP_EMPTY_PAGES=0
+TMP_DIR=`mktemp -d`
+trap "rm -rf $TMP_DIR" 0
 
 # Parse command-line options
 while [ $# -gt 0 ]; do
@@ -58,6 +62,10 @@ while [ $# -gt 0 ]; do
 
   --searchable|--ocr) SEARCHABLE=1 ;;
 
+  --language) LANGUAGE=$1 ;;
+
+  --skip-empty-pages) SKIP_EMPTY_PAGES=1 ;;
+
   -o|--output) shift; OUTPUT="$1" ;;
 
   -l|--outputlist) shift; USEARRAY=1; OUTPUT=(); OUTPUT+=("$1") ;;
@@ -94,6 +102,8 @@ if [ $HELP -eq 1 ]; then
   echo "   Custom Page Height in mm"
   echo " -pw, --page-width"
   echo "   Custom Page Width in mm"
+  echo " -x, --device"
+  echo "  Override scanner device name, defaulting to `fujitsu`"
   echo " --crop"
   echo "   Crop to contents (driver must support this)"
   echo " --deskew"
@@ -102,6 +112,10 @@ if [ $HELP -eq 1 ]; then
   echo "   Run post-processing deskew and black edge detection (requires unpaper)"
   echo " --ocr"
   echo "   Run OCR to make the PDF searchable (requires tesseract)"
+  echo " --language <lang>"
+  echo "   which language to use for OCR"
+  echo " --skip-empty-pages"
+  echo "   remove empty pages from resulting PDF document (e.g. one sided doc in duplex mode)"
   echo ""
   echo "OUTPUT"
   echo " -o, --output <outputfile>"
@@ -175,10 +189,12 @@ fi
 export VERBOSE
 export UNPAPER
 export SEARCHABLE
+export LANGUAGE
 export RESOLUTION
 export PGWIDTHIN
 export PGHEIGHTIN
 export PS2PDF_OPTS
+export SKIP_EMPTY_PAGES
 
 if [ $VERBOSE = 1 ]; then
   LOCKFILE=$(mktemp)
@@ -186,12 +202,13 @@ if [ $VERBOSE = 1 ]; then
   export LOCKFILE
 fi;
 
+
 echo >&2 "Scanning..."
 #eval strace -f -o /tmp/scan-trace.txt scanadf -d $DEVICE $MAXPAGE $PGHEIGHT $PGWIDTH -S $SCRIPT --script-wait --resolution $RESOLUTION --mode $MODE $DESKEW $CROP $SOURCE -o scan-%04d
-eval scanadf -d "$DEVICE" $MAXPAGE $PGHEIGHT $PGWIDTH -S $SCRIPT --script-wait --resolution $RESOLUTION --mode $MODE $DESKEW $CROP $SOURCE -o scan-%04d
+eval scanadf -d "$DEVICE" $MAXPAGE $PGHEIGHT $PGWIDTH -S $SCRIPT --script-wait --resolution $RESOLUTION --mode $MODE $DESKEW $CROP $SOURCE -o $TMP_DIR/scan-%04d
 
 shopt -s extglob nullglob
-pdffiles=(scan-[0-9]*.pdf)
+pdffiles=($TMP_DIR/scan-[0-9]*.pdf)
 numscans=${#pdffiles[@]}
 if [ $numscans -gt 0 ]; then
   echo "Processing $numscans pages"
@@ -213,13 +230,13 @@ if [ $numscans -gt 0 ]; then
           if [ -f "${OUTPUT[$index]}.orig" ]; then
             pdffiles+=("${OUTPUT[$index]}.orig")
           fi
-          pdffiles+=(scan-*(0)$scanno.pdf)
-          pdfunite "${pdffiles[@]}" "${OUTPUT[$index]}" && rm scan-*(0)$scanno.pdf
+          pdffiles+=($TMP_DIR/scan-*(0)$scanno.pdf)
+          pdfunite "${pdffiles[@]}" "${OUTPUT[$index]}" && rm $TMP_DIR/scan-*(0)$scanno.pdf
         else
-          mv scan-*(0)$scanno.pdf "${OUTPUT[$index]}"
+          mv $TMP_DIR/scan-*(0)$scanno.pdf "${OUTPUT[$index]}"
         fi
       else
-        mv scan-*(0)$scanno.pdf "${OUTPUT[$index]}"
+        mv $TMP_DIR/scan-*(0)$scanno.pdf "${OUTPUT[$index]}"
       fi
       let "index = $index + 1"
     done
@@ -232,13 +249,13 @@ if [ $numscans -gt 0 ]; then
     if [ -f "${OUTPUT}.orig" ]; then
       pdffiles+=("${OUTPUT}.orig")
     fi
-    pdffiles+=(scan-[0-9]*.pdf)
-    pdfunite "${pdffiles[@]}" "$OUTPUT" && rm scan-[0-9]*.pdf
+    pdffiles+=($TMP_DIR/scan-[0-9]*.pdf)
+    pdfunite "${pdffiles[@]}" "$OUTPUT" && rm $TMP_DIR/scan-[0-9]*.pdf
   else
     if [ $USEARRAY = 1 ]; then
-      mv scan-0*.pdf "${OUTPUT[0]}"
+      mv $TMP_DIR/scan-0*.pdf "${OUTPUT[0]}"
     else
-      mv scan-0*.pdf "$OUTPUT"
+      mv $TMP_DIR/scan-0*.pdf "$OUTPUT"
     fi
   fi
   echo ""
diff --git a/scan_perpage b/scan_perpage
index d546ee2..366de62 100755
--- a/scan_perpage
+++ b/scan_perpage
@@ -9,9 +9,11 @@ usage()
   echo "Set the following environment variables:"
   echo "  UNPAPER"
   echo "  SEARCHABLE"
+  echo "  LANGUAGE"
   echo "  RESOLUTION"
   echo "  PGWIDTHIN"
   echo "  PGHEIGHTIN"
+  echo "  SKIP_EMPTY_PAGES"
   echo "  PS2PDF_OPTS (optional)"
   echo "  VERBOSE (optional)"
   echo "  LOCKFILE (required if VERBOSE=1)"
@@ -38,53 +40,63 @@ if [ $# -lt 1 ]; then
   exit 1
 fi
 
-if [ "$UNPAPER" == "" -o "$SEARCHABLE" == "" -o "$RESOLUTION" == "" -o "$RESOLUTION" == "" ]; then
+if [ "$UNPAPER" == "" -o "$SEARCHABLE" == "" -o "$RESOLUTION" == "" -o "$RESOLUTION" == "" -o "$SKIP_EMPTY_PAGES" == "" ]; then
   usage
   exit 1
 fi
 
-IMAGE_FILE=$1
+IMAGE_PATH=$1
+IMAGE_DIR=`dirname $1`
+IMAGE_FILE=`basename $1`
 
 process_page() {
   log ""
   log "-------------------------------------------------------------------------------"
   log "Post-processing scanned page ${IMAGE_FILE}, deskew=$UNPAPER, searchable=$SEARCHABLE..."
 
+  
+  [[ $SKIP_EMPTY_PAGES -eq 1 ]] && PERCENTAGE_WHITE=`convert "$IMAGE_PATH" -fuzz 0% -negate -threshold 0 -negate -format "%[fx:100*mean]" info:` || PERCENTAGE_WHITE=0
+  log "$IMAGE_PATH has $PERCENTAGE_WHITE % white"
+    
   PP_PREFIX=
-  if [ $UNPAPER -eq 1 ]; then
-    log "Applying unpaper post-processing to image data..."
-    PP_PREFIX="unpaper-"
-    if [ $VERBOSE = 1 ]; then
-      UNPAPERVERBOSE="-v"
+  if (( $(echo "$PERCENTAGE_WHITE < 99.99" |bc -l) )); then
+    if [ $UNPAPER -eq 1 ]; then
+      log "Applying unpaper post-processing to image data..."
+      PP_PREFIX="unpaper-"
+      if [ $VERBOSE = 1 ]; then
+        UNPAPERVERBOSE="-v"
+      fi
+      #unpaper $UNPAPERVERBOSE --no-mask-scan --overwrite --dpi $RESOLUTION --no-blackfilter $IMAGE_FILE $PP_PREFIX$IMAGE_FILE | logstdout
+      unpaper $UNPAPERVERBOSE --overwrite --dpi $RESOLUTION $IMAGE_PATH $IMAGE_DIR/$PP_PREFIX$IMAGE_FILE | logstdout
     fi
-    #unpaper $UNPAPERVERBOSE --no-mask-scan --overwrite --dpi $RESOLUTION --no-blackfilter $IMAGE_FILE $PP_PREFIX$IMAGE_FILE | logstdout
-    unpaper $UNPAPERVERBOSE --overwrite --dpi $RESOLUTION $IMAGE_FILE $PP_PREFIX$IMAGE_FILE | logstdout
-  fi
-  log ""
-  log "-------------------------------------------------------------------------------"
-  log "Converting image data to pdf..."
-  if [ $SEARCHABLE -eq 1 ]; then
-    log "Converting image data to searchable pdf..."
-    tesseract $PP_PREFIX$IMAGE_FILE ${IMAGE_FILE%.*} -l eng pdf | logstdout
-  else
+    log ""
+    log "-------------------------------------------------------------------------------"
     log "Converting image data to pdf..."
-    PNMVERBOSE=
-    if [ $VERBOSE = 1 ]; then
-      PNMVERBOSE="-verbose"
-    fi
-    if [ "$PGWIDTHIN" == "" -o "$PGHEIGHTIN" == "" ]; then
-      PAGEOPTS="-equalpixels -dpi=$RESOLUTION -noturn"
+    if [ $SEARCHABLE -eq 1 ]; then
+      log "Converting image data to searchable pdf..."
+      tesseract $IMAGE_DIR/$PP_PREFIX$IMAGE_FILE $IMAGE_DIR/${IMAGE_FILE%.*} -l $LANGUAGE pdf | logstdout
     else
-      PAGEOPTS="-imagewidth $PGWIDTHIN -imageheight $PGHEIGHTIN"
-    fi
-    log "Using page options: $PAGEOPTS"
-    pnmtops $PNMVERBOSE $PAGEOPTS $PP_PREFIX$IMAGE_FILE | ps2pdf $PS2PDF_OPTS - > ${IMAGE_FILE%.*}.pdf | logstdout
+      log "Converting image data to pdf..."
+      PNMVERBOSE=
+      if [ $VERBOSE = 1 ]; then
+        PNMVERBOSE="-verbose"
+      fi
+      if [ "$PGWIDTHIN" == "" -o "$PGHEIGHTIN" == "" ]; then
+        PAGEOPTS="-equalpixels -dpi=$RESOLUTION -noturn"
+      else
+        PAGEOPTS="-imagewidth $PGWIDTHIN -imageheight $PGHEIGHTIN"
+      fi
+      log "Using page options: $PAGEOPTS"
+      pnmtops $PNMVERBOSE $PAGEOPTS $IMAGE_DIR/$PP_PREFIX$IMAGE_FILE | ps2pdf $PS2PDF_OPTS - > $IMAGE_DIR/${IMAGE_FILE%.*}.pdf | logstdout
+    fi    
+  else
+    log "skipping empty page $IMAGE_FILE"
   fi
   
   status=$?
-  rm $IMAGE_FILE
-  if [ -f $PP_PREFIX$IMAGE_FILE ]; then
-    rm $PP_PREFIX$IMAGE_FILE
+  rm $IMAGE_PATH
+  if [ -f $IMAGE_DIR/$PP_PREFIX$IMAGE_FILE ]; then
+    rm $IMAGE_DIR/$PP_PREFIX$IMAGE_FILE
   fi
 
   log ""