new-words

annotate new-words.sh @ 0:b21fb3f1fcb7

find changesets by author, revision, files, or words in the commit message
new-words.sh добавлен в репозиторий

Пока здесь есть только маленький README-файл
и скрипт new-words.sh, который делает минимум работы.
Подробности: http://xgu.ru/wiki/new-words
author igor@book.xt.vpn
date Sun Feb 21 19:34:18 2010 +0200 (2010-02-21)
parents
children 4c9076f87241
rev   line source
igor@0 1 #!/bin/sh
igor@0 2
igor@0 3 cat <<HELP > /dev/null
igor@0 4
igor@0 5 Поддержка нескольких языков:
igor@0 6
igor@0 7 new-words -l lang URL
igor@0 8
igor@0 9 Например, для немецких текстов:
igor@0 10
igor@0 11 new-words -l de URL
igor@0 12
igor@0 13 Или, предварительно создав соответствующую ссылку:
igor@0 14
igor@0 15 de-words URL
igor@0 16
igor@0 17 HELP
igor@0 18
igor@0 19 WORK_DIR=~/.new-words/
igor@0 20 LANGUAGE=en
igor@0 21 VOCABULARY=${LANGUAGE}.txt
igor@0 22 TEMP1=`mktemp /tmp/news-words-XXXXXXXXXX`
igor@0 23 TEMP2=`mktemp /tmp/news-words-XXXXXXXXXX`
igor@0 24 editor=${EDITOR:-vim}
igor@0 25
igor@0 26 my_name="`echo $0 | sed s/-words// | sed s@.*/@@`"
igor@0 27 [ "${my_name}" = "new" ] || VOCABULARY="$my_name".txt
igor@0 28 if [ "$1" = "-l" ]
igor@0 29 then
igor@0 30 LANGUAGE="$2"
igor@0 31 VOCABULARY="$LANGUAGE".txt
igor@0 32 shift 2
igor@0 33 fi
igor@0 34
igor@0 35 get_words()
igor@0 36 {
igor@0 37 tr ' ' '\n' | sed 's/--/ /g' | tr A-Z a-z \
igor@0 38 | tr -d '*\r,.-:#@()+=—<>$;"?!|·[]^%&'"'" \
igor@0 39 | tr ' ' '\n' | grep_v_english_perl \
igor@0 40 | grep -x '[a-zA-Z0-9_-]*' \
igor@0 41 | sort | uniq -c | awk '{if ($2!="") print;}' | sort -rn
igor@0 42 }
igor@0 43
igor@0 44 grep_v_english()
igor@0 45 {
igor@0 46 [ -e "$VOCABULARY" ] || touch "$VOCABULARY"
igor@0 47 eval $(cat $VOCABULARY | tr -d "'" | xargs -n10 echo | tr ' ' '|' | sed 's/^/egrep -xv "RRRRRRR|/' | sed 's/$/"/' | tr '\n' '|')cat
igor@0 48 }
igor@0 49
igor@0 50 grep_v_english_perl()
igor@0 51 {
igor@0 52 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@0 53 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@0 54 open(VOC, $ENV{VOCABULARY})
igor@0 55 or die "Can't open VOCABULARY";
igor@0 56 while (<VOC>){
igor@0 57 chomp;
igor@0 58 s/'//g;
igor@0 59 $voc{$_}="1";
igor@0 60 }
igor@0 61 while(<>) {
igor@0 62 chomp;
igor@0 63 if (not defined($voc{$_})) { print "$_\n"; }
igor@0 64 }
igor@0 65 PERL_SCRIPT
igor@0 66 [ -e "$VOCABULARY" ] || touch "$VOCABULARY"
igor@0 67 export VOCABULARY
igor@0 68 perl $PERL_SCRIPT_TEMP_NAME
igor@0 69 rm $PERL_SCRIPT_TEMP_NAME
igor@0 70 }
igor@0 71
igor@0 72 text_from_url()
igor@0 73 {
igor@0 74 lynx -dump "$1" | perl -p -e 's@http://[a-zA-Z&_.:/0-9%?=,#+()\[\]~-]*@@'
igor@0 75 }
igor@0 76
igor@0 77 mkdir -p $WORK_DIR
igor@0 78 cd $WORK_DIR
igor@0 79 if echo "$1" | grep -q http:
igor@0 80 then
igor@0 81 text_from_url "$1" | get_words | tee "$TEMP1" > "$TEMP2"
igor@0 82 elif [ "$#" != 0 ]
igor@0 83 then
igor@0 84 cat "$1" | get_words | tee "$TEMP1" > "$TEMP2"
igor@0 85 else
igor@0 86 get_words | tee "$TEMP1" > "$TEMP2"
igor@0 87 fi
igor@0 88
igor@0 89 if [ "$editor" = vim ]
igor@0 90 then
igor@0 91 vim -c 'set keywordprg='"$LANGUAGE" -c 'set iskeyword=@,48-57,/,.,-,_,+,,,#,$,%,~,=' "$TEMP2" < /dev/tty > /dev/tty
igor@0 92 else
igor@0 93 echo 2
igor@0 94 $editor "$TEMP2"
igor@0 95 fi
igor@0 96 diff "$TEMP1" "$TEMP2" | awk '{print $3}' | sort -u >> "$VOCABULARY"
igor@0 97 rm -f "$TEMP1" "$TEMP2"