#!/bin/bash

# A simple script of perl pattern replacement calls to normalize symbol
# labels for CROHME 2013.
#
# R. Zanibbi

if [ $# -lt 1 ]
then
	echo "CROHMELib CROHME .inkml Symbol Label Normalizer"
	echo "Copyright (c) R. Zanibbi, H. Mouchere, 2012-2013"
	echo ""
	echo "Usage: normalizeSymbols <inkml_dir>"
	echo ""
	echo "Applies a series of symbol pattern replacements to all files"
	echo "in <inkml_dir>. This is currently an ad-hoc list of patterns"
	echo "and replacements (see script source for details)."
	echo ""
	exit 0
fi


dirIn=$1
if ! [ -d $dirIn ]
then
	echo "Directory \"$dirIn\" does not exist."
	exit 1
fi

# Replacing unicode patterns (e.g. from MfrDB) 

perl -p -i -e "s/\342\213\205/\./g;
	s/\342\206\222/rightarrow/g;
	s/\342\210\221/sum/g;
	s/\342\210\253/int/g;
	s/\317\200/pi/g;
	s/\342\211\244/leq/g;
	s/\342\211\245/geq/g;
	s/\342\210\236/infty/g;
	s/\342\200\262/prime/g;
	s/\303\227/times/g;
	s/\342\210\225/\//g;
	s/\316\261/alpha/g;
	s/\316\262/beta/g;
	s/\302\261/pm/g;" $dirIn/*.inkml


# Bugs (or Unicode duplicates?)
perl -p -i -e "s/\342\205\206/d/g;s/\342\205\207/e/g;s/\342\205\210/i/g" $dirIn/*.inkml


# Unicode name identification in KME2G3 files.
perl -p -i -e "s/\354\235\264\355\203\235\355\227\214/Name_Removed/g" $dirIn/*.inkml

# Replacing other patterns
perl -p -i -e "s/ctdot/\./g;s/>le</>leq</g;s/>exist</>exists</g" $dirIn/*.inkml
