#!/bin/bash
# Check common misspellings
# input file format:
# word->word1, ...
# Source: http://en.wikipedia.org/wiki/Wikipedia:Lists_of_common_misspellings/For_machines

hunspell=../../src/tools/hunspell
hlang=${HUNSPELL:-en_US}
alang=${ASPELL:-en_US}
input=${INPUT:-List_of_common_misspellings.txt}

# remove bad words recognised by Hunspell as good
cat $input | sed 's/[-]>/	/' | $hunspell -d $hlang -1 -L |

# remove items with dash for Aspell
grep '^[^-]*	' |

# remove spaces from end of lines
sed 's/ *$//' >$input.1

# remove bad words recognised by Aspell as good
cut -f 1 -d '	' $input.1 | aspell -l $alang --list |
awk 'FILENAME=="-"{a[$1]=1;next}a[$1]{print$0}' - $input.1 |

# change commas with tabs
sed 's/, */	/g' >$input.2

# remove lines with unrecognised suggestions (except suggestion with spaces)
cut -d '	' -f 2- $input.2 | tr "\t" "\n" | grep -v ' ' >x.1
cat x.1 | $hunspell -l -d $hlang >x.2
cat x.1 | aspell -l $alang --list >>x.2
cat x.2 | awk 'BEGIN{FS="\t"}
FILENAME=="-"{a[$1]=1;next}a[$2]!=1 && a[$3]!=1{print $0}' - $input.2 >$input.3

cut -f 1 -d '	' $input.3 | aspell -l $alang -a | grep -v ^$ | sed -n '2,$p' |
sed 's/^.*: //;s/, /	/g' >$input.4

cat $input.3 | $hunspell -d $hlang -a -1 | grep -v ^$ | sed -n '2,$p' |
sed 's/^.*: //;s/, /	/g' >$input.5