igor@0: #!/bin/sh
igor@0: 
igor@0: cat <<HELP > /dev/null
igor@0: 
igor@0: Поддержка нескольких языков:
igor@0: 
igor@0:     new-words -l lang URL
igor@0: 
igor@0: Например, для немецких текстов:
igor@0: 
igor@0:     new-words -l de URL
igor@0: 
igor@0: Или, предварительно создав соответствующую ссылку:
igor@0: 
igor@0:     de-words URL
igor@0: 
igor@0: HELP
igor@0: 
igor@0: WORK_DIR=~/.new-words/
igor@0: LANGUAGE=en
igor@0: VOCABULARY=${LANGUAGE}.txt
igor@0: TEMP1=`mktemp /tmp/news-words-XXXXXXXXXX`
igor@0: TEMP2=`mktemp /tmp/news-words-XXXXXXXXXX`
igor@0: editor=${EDITOR:-vim}
igor@0: 
igor@0: my_name="`echo $0 | sed s/-words// | sed s@.*/@@`"
igor@0: [ "${my_name}" = "new" ] || VOCABULARY="$my_name".txt
igor@0: if [ "$1" = "-l" ]
igor@0: then
igor@0:     LANGUAGE="$2"
igor@0:     VOCABULARY="$LANGUAGE".txt
igor@0:     shift 2
igor@0: fi
igor@0: 
igor@0: get_words()
igor@0: {
igor@0: tr ' ' '\n' | sed 's/--/ /g' | tr A-Z a-z \
igor@0: | tr -d '*\r,.-:#@()+=—<>$;"?!|·[]^%&'"'" \
igor@0: | tr ' ' '\n' | grep_v_english_perl \
igor@0: | grep -x '[a-zA-Z0-9_-]*' \
igor@0: | sort | uniq -c | awk '{if ($2!="") print;}' | sort -rn
igor@0: }
igor@0: 
igor@0: grep_v_english()
igor@0: {
igor@0: [ -e "$VOCABULARY" ] || touch "$VOCABULARY"
igor@0: eval $(cat $VOCABULARY | tr -d "'" | xargs -n10 echo | tr ' ' '|' | sed 's/^/egrep -xv "RRRRRRR|/' | sed 's/$/"/' | tr '\n' '|')cat
igor@0: }
igor@0: 
igor@0: grep_v_english_perl()
igor@0: {
igor@0:     PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@0:     cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@0: open(VOC, $ENV{VOCABULARY})
igor@0:  or die "Can't open VOCABULARY";
igor@0: while (<VOC>){
igor@0:     chomp;
igor@0:     s/'//g;
igor@0:     $voc{$_}="1";
igor@0: }
igor@0: while(<>) {
igor@0:     chomp;
igor@0:     if (not defined($voc{$_})) { print "$_\n"; }
igor@0: }
igor@0: PERL_SCRIPT
igor@0:     [ -e "$VOCABULARY" ] || touch "$VOCABULARY"
igor@0:     export VOCABULARY 
igor@0:     perl $PERL_SCRIPT_TEMP_NAME
igor@0:     rm $PERL_SCRIPT_TEMP_NAME
igor@0: }
igor@0: 
igor@0: text_from_url()
igor@0: {
igor@0: lynx -dump "$1" | perl -p -e 's@http://[a-zA-Z&_.:/0-9%?=,#+()\[\]~-]*@@'
igor@0: }
igor@0: 
igor@0: mkdir -p $WORK_DIR
igor@0: cd $WORK_DIR
igor@0: if echo "$1" | grep -q http: 
igor@0: then 
igor@0:     text_from_url "$1" | get_words | tee "$TEMP1" > "$TEMP2"
igor@0: elif [ "$#" != 0 ]
igor@0: then
igor@0:     cat "$1" | get_words | tee "$TEMP1" > "$TEMP2"
igor@0: else 
igor@0:     get_words | tee "$TEMP1" > "$TEMP2"
igor@0: fi
igor@0: 
igor@0: if [ "$editor" = vim ]
igor@0: then
igor@0:     vim -c 'set keywordprg='"$LANGUAGE" -c 'set iskeyword=@,48-57,/,.,-,_,+,,,#,$,%,~,=' "$TEMP2" < /dev/tty > /dev/tty
igor@0: else
igor@0:     echo 2
igor@0:     $editor "$TEMP2"
igor@0: fi
igor@0: diff "$TEMP1" "$TEMP2" | awk '{print $3}' | sort -u >> "$VOCABULARY"
igor@0: rm -f "$TEMP1" "$TEMP2"