Mercurial > dottes
view makeAltoRecorderCFingering.sh @ 735:68f926e61d16
Revise Markdown handling of character entities.
Ideally I'd like the ABC character entities to survive Markdown and then
get converted. But because they start with '\', they don't. So I have
no alternative but to convert them to HTML entities, which Markdown then
converts to UTF-8.
author | Jim Hague <jim.hague@acm.org> |
---|---|
date | Thu, 12 Oct 2017 13:32:24 +0100 |
parents | 2f6e05d0aba0 |
children | ce5c7214f9aa |
line wrap: on
line source
#!/bin/bash # # Transpose a book for alto recorder with C fingering. if [ $# != 1 ]; then echo "Usage: makeAltoRecorderCFingering.sh <book dir name>" exit 1 fi # Transpose up (return 0) if bottom note was < F (< C for recorder). transposeup() { (($3 < 103)) } dir=`pwd` booke=$dir/$1 outdir=$dir/$1-AltoRecorderCFingering mkdir -p $outdir # Copy book component items. cp $booke/*.txt $outdir echo "Alto Recorder (C Fingering)" > $outdir/instrument.txt find $booke -name "*.abc" | sort | while read filename do name=`basename $filename .abc` range=`./abcrange.py $filename` # Transpose concert pitch down a fifth. # If there are any notes below 'F' (recorder 'C'), transpose # up a seventh instead. transpose=-5 if transposeup $range; then transpose=7 fi # There's no point in having transposed chords. Remove from the # abc before transposing. Some badly formed chord items can give # erroneous output from abc2abc (like, strings of binary gibberish). sed -e "s/\"[^\"]*\"//g" $filename > $outdir/$name.abc.tmp # Transpose. By default abc2abc will report errors in the output, # but this messes up output formatting so stop it. abc2abc $outdir/$name.abc.tmp -e -t $transpose > $outdir/$name.abc rm $outdir/$name.abc.tmp done