X-Git-Url: https://git.ucc.asn.au/?p=dja%2Fscandal.git;a=blobdiff_plain;f=scan2pages.sh;h=662d87749a541ea3142f7aceafa0df76ad03903b;hp=995c5a6dbbcaaae2b9b06405634619ac4d19f957;hb=eab68b895041b14605257517b214d14474037d4b;hpb=3d2d99e7b6ce1522ad2c30ccfe24d3271c68cb47 diff --git a/scan2pages.sh b/scan2pages.sh index 995c5a6..662d877 100755 --- a/scan2pages.sh +++ b/scan2pages.sh @@ -2,19 +2,22 @@ function usage { cat >&2 << __EOF__ -Usage: $0 [-vfpsmut] [-d depth] pdffile outdir +Usage: $0 [-vfpsmutb] [-D DPI] [-d depth] [-F format] pdffile outdir Convert pdffile - a pdf of scanned facing pages - to a set of images in outdir, each with only one page. OPTIONS: -v: Be verbose. - -d depth: Use given depth. Default is 1. Unpaper can only handle up to 8. - -D dpi: Use given DPI. Colour scans on library scanners are usually 300, bw 400. + -d depth: Use given depth. Default is 1 for black and white, 8 for colour. + Unpaper can only handle up to 8. + -D dpi: Use given DPI. Colour scans on library scanners are usually 300, bw 400. + -F extension: use the output format signified by the extension. Default is png -s: skip masking/trimming. Overrides -m. -f: Forceably redo everything. -p: Forceably redo pdf conversion. Implies options below, equivalent to -f. -m: Forceably redo masking/trimming and other preprocessing. Impiles options below. -u: Forceably redo unpaper processing. Implies option below. IGNORED -t: Forceably redo final trimming and cleaning. IGNORED + -b: Optimize for BeBook One. __EOF__ } @@ -34,18 +37,21 @@ forceunpaper= forceclean= convertflags= unpaperflags= -depthflags="-depth 1" +depthset= +depth= dpi= skipmask= verbose= -while getopts 'vd:sfpmuc' OPTION +extension="png" +bebook= +while getopts 'vd:D:sfpmucF:b' OPTION do case $OPTION in v) verbose=1 convertflags="$convertflags -verbose" unpaperflags="$unpaperflags -v --time" ;; - d) depthflags="-depth $OPTARG" + d) depth="$OPTARG" ;; D) dpi=$OPTARG ;; @@ -61,9 +67,10 @@ do ;; t) forceclean=1 ;; -# b) bflag=1 -# bval="$OPTARG" -# ;; + F) extension="$OPTARG" + ;; + b) bebook=1 + ;; ?) usage exit 2 ;; @@ -111,6 +118,18 @@ if [[ "$dpi" = "" ]]; then fi fi +# guess at depth, unless it's been set +if [[ "$depth" = "" ]]; then + if [[ $( convert $file[0] -format "%k" info: || exit 1 ) == 2 ]]; then + depth=1; + else + depth=8 + fi + if [[ $verbose == 1 ]]; then + echo "Depth guessed at $depth." + fi +fi + # process pages 1 by 1 to avoid convert gobbling all the memory for pg in `$my_seq 1 $pages`; do echo "Processing page $pg." @@ -120,7 +139,7 @@ for pg in `$my_seq 1 $pages`; do # convert from pdf origpnm=$dir/pg-${pgn}.pnm if [[ ! $([ -e $origpnm ]) || $forcepdf ]]; then - convert $convertflags $depthflags -density $dpi $file[$(expr $pg - 1)] \ + convert $convertflags -depth $depth -density $dpi $file[$(expr $pg - 1)] \ $origpnm || exit 1 fi; @@ -147,12 +166,15 @@ for pg in `$my_seq 1 $pages`; do unppnm=$dir/upg-${pgn}-%01d.pnm unpaper $unpaperflags --layout double --overwrite --no-blackfilter -ni 10 -op 2 $preppnm $unppnm || exit 1 - # final convert and clean - convert $convertflags $dir/upg-${pgn}-?.pnm $dir/upg-${pgn}-%01d.tiff || exit 1 - + #final convert and clean w/ bebook optimisation + if [[ $bebook ]]; then + convert $convertflags $dir/upg-${pgn}-?.pnm -depth 4 -resize 600x800 $dir/final-${pgn}-%01d.${extension} || exit 1 + else + convert $convertflags $dir/upg-${pgn}-?.pnm $dir/final-${pgn}-%01d.${extension} || exit 1 + fi done mkdir $dir/pages -mv $dir/upg-*.tiff $dir/pages +mv $dir/final-*.${extension} $dir/pages