igor@0: #!/bin/sh
igor@20: 
igor@20: show_usage()
igor@20: {
igor@20: cat <<HELP > /dev/stderr
igor@0: 
igor@16: USAGE: 
igor@16: 
igor@16:     new-words [ -l lang ] [ -s ] [ ARG ] 
igor@16: 
igor@16: SWITCHES: 
igor@16: 
igor@20:     -h          print this screen
igor@20:     -k          put higher words that are similar to the known words (only for English)
igor@16:     -l lang     override language settings
igor@21:     -n          non-interactive mode (don't run vi)
igor@21:     -m          don't add marks (and don't save marks added by user)
igor@20:     -s          show the text statistics (percentage of known words and so on) and exit
igor@16: 
igor@20: The language of the text can be specified also
igor@20: by name of the program new-words (correspondent link must be created before).
igor@20: For example, these calls are equivalent:
igor@0: 
igor@20:     de-words URL
igor@0:     new-words -l de URL
igor@0: 
igor@20: HELP
igor@20: }
igor@0: 
igor@20: if [ "$1" = "-h" ]
igor@20: then
igor@20:     show_usage
igor@20:     exit 0
igor@20: fi
igor@0: 
igor@0: WORK_DIR=~/.new-words/
igor@17: TEMP1=`mktemp /tmp/new-words-XXXXXXXXXX-temp1`
igor@17: TEMP2=`mktemp /tmp/new-words-XXXXXXXXXX-temp2`
igor@17: export ORIGINAL_TEXT=`mktemp /tmp/new-words-XXXXXXXXXX-orig`
igor@0: editor=${EDITOR:-vim}
igor@0: 
igor@21: #----------------------------------------------------
igor@21: # command line options processing
igor@21: 
igor@2: LANGUAGE=en
igor@2: my_name="`echo $0 | sed s@.*/@@ | sed s/-.*// `"
igor@22: for arg
igor@22: do
igor@22:     if echo "$arg" | grep -q http://...wikipedia.org/wiki/
igor@22:     then
igor@22:     LANGUAGE="`echo $arg | sed s@http://@@ | sed s@.wikipedia.*@@`"
igor@22:     fi
igor@22: done
igor@2: [ "${my_name}" = "new" ] || LANGUAGE="$my_name"
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@2: VOCABULARY=${LANGUAGE}.txt
igor@2: NOTES_FILE=notes-${LANGUAGE}.txt
igor@0: 
igor@16: STAT_ONLY=NO
igor@16: if [ "$1" = "-s" ]
igor@16: then
igor@16:     STAT_ONLY=YES
igor@16:     shift
igor@16: fi
igor@16: 
igor@20: NEED_TO_USE_VOCABULARY_WHEN_SORT=NO
igor@20: if [ "$1" = "-k" ]
igor@20: then
igor@20:     NEED_TO_USE_VOCABULARY_WHEN_SORT=YES
igor@20:     shift
igor@20: fi
igor@20: 
igor@21: DONT_ADD_MARKS=NO
igor@21: if [ "$1" = "-m" ]
igor@21: then
igor@21:     DONT_ADD_MARKS=YES
igor@21:     shift
igor@21: fi
igor@21: 
igor@21: NON_INTERACTIVE_MODE=NO
igor@21: if [ "$1" = "-n" ]
igor@21: then
igor@21:     NON_INTERACTIVE_MODE=YES
igor@21:     shift
igor@21: fi
igor@21: 
igor@21: #----------------------------------------------------
igor@21: 
igor@0: get_words()
igor@0: {
igor@1: tr ' ' '\n' | sed 's/--/ /g' \
igor@3: | sed "s/'/__APOSTROPHE__/g" \
igor@3: | tr  '—·-' '-----' \
igor@3: | tr '*\r,.:#@()+=—<>$;"?!|·[]^%&' '                           ' \
igor@3: | tr ' ' '\n' \
igor@3: | grep -x '[[:alpha:]'"'"'-]*' \
igor@3: | tee "$1" \
igor@3: | grep_v_english_perl \
igor@0: | sort | uniq -c | awk '{if ($2!="") print;}' | sort -rn
igor@0: }
igor@0: 
igor@3: add_stat()
igor@3: {
igor@3:     before="$1"
igor@3:     after=${before}2
igor@3:     cat > "$after"
igor@3:     total="`wc -w $1 | awk '{print $1}'`"
igor@3:     total_unknown="`cat $after|awk '{s=s+$1}END{print s}'`"
igor@3:     total_known="`echo $total-$total_unknown|bc`"
igor@3:     percentage="`echo '100*('$total-$total_unknown')'/$total | bc -l | sed 's/\\.\(.\).*/.\1/'`"
igor@18:     #sentences="`cat $after | perl -e 'local $/; $_=<>; s@http://[a-zA-Z&_.:/0-9%?=,\#+()\[\]~-]*@@g; s@\n@@g; s@(Mr|Mrs)\.@\1POINT@g; @sentences=split /\\./;print $#sentences;'`"
igor@18:     sentences="`cat $ORIGINAL_TEXT | perl -e 'local $/; $_=<>; s/[^.]//msg; print length($_);'`"
igor@18: 
igor@18: 
igor@16:     if [ "$STAT_ONLY" = "YES" ]
igor@16:     then
igor@18:         echo "LANG  KNOWN%  UNKNOWN%  KNOWN     TOTAL     WPS  UWPS*10"
igor@18:         echo "$LANGUAGE    $percentage    `echo \(100-$percentage\) | bc -l`      $total_known    $total    `echo $total/$sentences|bc`   `echo 10*$total_unknown/$sentences|bc` "
igor@17:         rm $after
igor@16:         return 0
igor@16:     else 
igor@16:         echo "# $LANGUAGE, $percentage, <$total_known/$total>"
igor@16:     fi
igor@16: 
igor@14:     PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@14:     cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@14: my $total=shift(@ARGV);
igor@14: my $total_known=shift(@ARGV);
igor@14: my $s=0;
igor@16: my $mark_line=int($total_known*100/$total/5)*5;
igor@19: if ($mark_line>=90) { 
igor@19:     $mark_line=int($total_known*100/$total)+1;
igor@19: } else { $mark_line +=5; };
igor@14: while(<>)
igor@14: {
igor@14:     print;
igor@14:     /^\s*([0-9]*)\s*/;
igor@14:     $s+=$1;
igor@19:     if (($total_known+$s)*100/$total>=$mark_line) {
igor@14:         print "# $mark_line\n";
igor@14:         if ($mark_line>=90) { $mark_line+=1; } else { $mark_line +=5; };
igor@14:     }
igor@14: }
igor@14: PERL_SCRIPT
igor@14:     perl $PERL_SCRIPT_TEMP_NAME "$total" "$total_known" "$after"
igor@14:     rm $PERL_SCRIPT_TEMP_NAME
igor@3:     rm $after
igor@3: }
igor@3: 
igor@0: grep_v_english()
igor@0: {
igor@0: [ -e "$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@3:     #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 $PERL_SCRIPT_TEMP_NAME
igor@0: }
igor@0: 
igor@11: group_words()
igor@11: {
igor@22:     #if [ "$LANGUAGE" != "en" ]
igor@22:     #then
igor@22:     #    cat 
igor@22:     #    return
igor@22:     #fi
igor@11:     PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-group-words-XXXXXXXX`
igor@11:     cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@11: #!/usr/bin/perl
igor@11: 
igor@12: eval {
igor@12: # http://stackoverflow.com/questions/251694/how-can-i-check-if-i-have-a-perl-module-before-using-it
igor@12:     require String::Similarity;
igor@12:     String::Similarity->import();
igor@12: };
igor@12: unless($@)
igor@12: {
igor@12:     our $HAVE_String_Similarity=1;
igor@12: }
igor@12: 
igor@12: sub similar($$){
igor@12:     my $a=shift;
igor@12:     my $b=shift;
igor@12:     if ($HAVE_String_Similarity) {
igor@12:         return $Similarity{"$a $b"};
igor@12:     } 
igor@12:     else {
igor@12:         return 0;
igor@12:     }
igor@12: }
igor@12: 
igor@11: sub normalize($)
igor@11: {
igor@22:     if   ( $ENV{LANGUAGE} eq "en" ) { return normalize_english(shift); }
igor@22:     elsif ( $ENV{LANGUAGE} eq "de" ) { return normalize_german(shift); }
igor@22:     else { return shift ; }
igor@22: }
igor@22: 
igor@22: sub normalize_german($)
igor@22: {
igor@22:     $_=lc(shift);
igor@22: 
igor@22:     s/heit$//;  s/keit$//; s/tum$//; s/ung$//; s/nis$//;s/schaft$//; s/ist$//; 
igor@22:     s/en$//; s/er$//;
igor@22: 
igor@22:     s/lich$//; s/ig$//;
igor@22:     s/al$//; s/isch$//;
igor@22:     s/ell$//; s/haft$//;
igor@22: 
igor@22:     s/bar$//; s/sam$//; s/lich$//;
igor@22: 
igor@22:     @prefixes=qw(
igor@22:         ab an auf aus bei dazwischen ein fest heraus her hinaus hin los mit nach voraus vorbei vor weg weiter zurück zusammen zu
igor@22:         be emp ent er ge miss ver zer durch über um unter wieder);
igor@22:     for $pref (@prefixes) {
igor@22:         s/^$pref//;
igor@22:     }
igor@22: 
igor@22: 
igor@22:     return $_;
igor@22: }
igor@22: 
igor@22: sub normalize_english($)
igor@22: {
igor@11:     $_=lc(shift);
igor@12: 
igor@11:     s/s$//;
igor@22: 
igor@22:     s/ation$//;  s/ness$//; s/ship$//; s/ally$//; s/ance$//;s/ity$//; s/ment$//; 
igor@22: 
igor@11:     s/ed$//;
igor@12:     s/en$//;
igor@12:     s/er$//;
igor@12:     s/est$//;
igor@11:     s/ing$//;
igor@12: 
igor@12:     s/ism$//; s/ist$//; s/ful$//; s/able$//; s/ably$//;
igor@22:     s/ify$//; s/fy$//; s/ly$//;
igor@12:     s/ise$//; s/ize$//;
igor@12: 
igor@12:     s/e$//;
igor@11:     return $_;
igor@11: }
igor@11: 
igor@12: 
igor@11: sub compare($$)
igor@11: {
igor@11:     my $a=shift;
igor@11:     my $b=shift;
igor@11:     $a =~ s/^\s*//;
igor@11:     $b =~ s/^\s*//;
igor@11:     my ($a1, $a2)= split /\s+/,$a,2;
igor@11:     my ($b1, $b2)= split /\s+/,$b,2;
igor@11: 
igor@11:     my $cmp = $group_weight{normalize($a2)} <=> $group_weight{normalize($b2)};
igor@12: 
igor@11:     if ($cmp) {
igor@11:         return $cmp;
igor@11:     }
igor@11:     else {
igor@11:         if (normalize($a2) ne normalize($b2)) {
igor@11:             return normalize($a2) cmp normalize($b2);
igor@11:         }
igor@11:         else {
igor@11:             return $a1 <=> $b1;
igor@11:         }
igor@11:     }
igor@11: }
igor@11: 
igor@20: our %Vocabulary;
igor@20: open(VOC, $ENV{VOCABULARY})
igor@20:  or die "Can't open VOCABULARY";
igor@20: while (<VOC>){
igor@20:     chomp;
igor@20:     #s/'//g;
igor@20:     $Vocabulary{normalize($_)}="1";
igor@20: }
igor@20: close(VOC);
igor@20: 
igor@20: @lines=<STDIN>;
igor@11: for $L (@lines) {
igor@11:     chomp($L);
igor@11:     $l=$L;
igor@11:     $l =~ s/^\s*//;
igor@11:     my ($a, $b)=split(/\s+/,$l,2);
igor@11:     $group_weight{normalize($b)}+=$a;
igor@11: }
igor@20: if ($ENV{NEED_TO_USE_VOCABULARY_WHEN_SORT} eq "YES") {
igor@20:     for $k (keys %group_weight) {
igor@20:         if (defined($Vocabulary{$k})) {
igor@20:             $group_weight{$k} *= 2;
igor@20:         }
igor@20:     }
igor@20: }
igor@11: @lines2 = sort { compare($b,$a) } @lines;
igor@11: for $l (@lines2) {
igor@11:     print "$l\n";
igor@11: }
igor@11: PERL_SCRIPT
igor@20:     export VOCABULARY
igor@20:     export NEED_TO_USE_VOCABULARY_WHEN_SORT
igor@22:     export LANGUAGE
igor@11:     perl $PERL_SCRIPT_TEMP_NAME
igor@11:     rm $PERL_SCRIPT_TEMP_NAME
igor@11: }
igor@11: 
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@2: add_marks()
igor@2: {
igor@2:     PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@2:     cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@2: $file = $ARGV[0];
igor@2: our $dict;
igor@2: if (open(NOTES, $ENV{NOTES_FILE})) {
igor@2:     while(<NOTES>) {
igor@2:         chomp;
igor@2:         s/^\s+//;
igor@2:         my ($a,$b)=split /\s+/,$_,2;
igor@2:         $dict{$a}=$b;
igor@2:     }
igor@2: }
igor@2: if (open(F, $file)) {
igor@2:     @lines=<F>;
igor@2:     close(F);
igor@2: 
igor@2:     if (open(F, ">$file")) {
igor@2:         for (@lines) {
igor@2:             m/\s+\S+\s+(\S+)/;
igor@2:             $name=$1;
igor@3:             if (not /^#/ and defined($dict{$name})) {
igor@2:                 chomp;
igor@2:                 $mark=$dict{$name};
igor@2:                 $space=" "x(30-length($_));
igor@2:                 print F "$_$space$mark\n";
igor@2:             }
igor@2:             else {
igor@2:                 print F "$_";
igor@2:             }
igor@2:         }
igor@2:         close(F);
igor@2:     }
igor@2: }
igor@2: PERL_SCRIPT
igor@2:     [ -e "$NOTES_FILE" ] || touch "$NOTES_FILE"
igor@2:     export NOTES_FILE
igor@2:     perl $PERL_SCRIPT_TEMP_NAME "$1"
igor@2:     rm $PERL_SCRIPT_TEMP_NAME
igor@2: }
igor@2: 
igor@2: remove_marks()
igor@2: {
igor@2:     PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@2:     cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@2: $file = $ARGV[0];
igor@2: our %dict;
igor@2: if (open(F, $file)) {
igor@2:     @lines=<F>;
igor@2:     close(F);
igor@2: 
igor@2:     if (open(F, ">$file")) {
igor@2:         for (@lines) {
igor@2:             chomp;
igor@3:             if (not /^#/ and m/(\s+)(\S+)(\s+)(\S+)(\s+)(.*)/) {
igor@2:                 my $name=$4;
igor@2:                 my $comment=$6;
igor@2:                 $dict{$name}=$comment;
igor@2:                 print F "$1$2$3$4\n";
igor@2:             }
igor@2:             else {
igor@2:                 print F "$_\n";
igor@2:             }
igor@2:         }
igor@2:     }
igor@2: }
igor@21: if (($ENV{DONT_ADD_MARKS} ne "YES") and open(NOTES, $ENV{NOTES_FILE})) {
igor@2:     @lines=<NOTES>;
igor@2:     close(NOTES);
igor@2: 
igor@2:     if (open(NOTES, ">".$ENV{NOTES_FILE})) {
igor@2:         for (@lines) {
igor@2:             chomp;
igor@2:             s/^\s+//;
igor@2:             my ($a,$b)=split /\s+/,$_,2;
igor@2:             if (not defined($dict{$a}) || ($dict{$a} eq $b)) {
igor@2:                 print NOTES "$_\n";
igor@2:                 if (defined($dict{$a})) { unset($dict{$a}); }
igor@2:             }
igor@2:         }
igor@2:         for (keys %dict) {
igor@2:             $mark=$dict{$_};
igor@2:             $space=" "x(30-length($_));
igor@2:             print NOTES "$_$space$mark\n";
igor@2:         }
igor@2:     }
igor@2: }
igor@2: PERL_SCRIPT
igor@2:     [ -e "$NOTES_FILE" ] || touch "$NOTES_FILE"
igor@2:     export NOTES_FILE
igor@21:     export DONT_ADD_MARKS
igor@2:     perl $PERL_SCRIPT_TEMP_NAME "$1"
igor@2:     rm $PERL_SCRIPT_TEMP_NAME
igor@2: }
igor@2: 
igor@0: mkdir -p $WORK_DIR
igor@8: oldpwd="$PWD"
igor@0: cd $WORK_DIR
igor@0: if echo "$1" | grep -q http: 
igor@0: then 
igor@11:     text_from_url "$1"
igor@0: elif [ "$#" != 0 ]
igor@0: then
igor@13:     if echo $1 | grep -q ^/
igor@13:     then
igor@13:         cat "$1"
igor@13:     else
igor@13:         cat "$oldpwd/$1"
igor@13:     fi
igor@0: else 
igor@11:     cat
igor@11: fi \
igor@11:    | tee $ORIGINAL_TEXT \
igor@11:    | get_words ${TEMP1}-full \
igor@11:    | group_words \
igor@11:    | add_stat ${TEMP1}-full \
igor@11:    | tee "$TEMP1" > "$TEMP2"
igor@0: 
igor@16: if [ "$STAT_ONLY" = "YES" ]
igor@0: then
igor@16:     cat "$TEMP1"
igor@21: elif [ "$NON_INTERACTIVE_MODE" = "YES" ]
igor@21: then
igor@21:     cat "$TEMP1"
igor@0: else
igor@21:     [ "$DONT_ADD_MARKS" = "YES" ] || add_marks "$TEMP2"
igor@16:     if [ "$editor" = vim ]
igor@16:     then
igor@23:         vim -c 'set keywordprg='"$LANGUAGE" -c 'set iskeyword=@,48-57,/,.,-,_,+,,,#,$,%,~,=,48-255' "$TEMP2" < /dev/tty > /dev/tty
igor@16:     else
igor@16:         echo 2
igor@16:         $editor "$TEMP2"
igor@16:     fi
igor@16:     remove_marks "$TEMP2"
igor@0: fi
igor@2: 
igor@0: diff "$TEMP1" "$TEMP2" | awk '{print $3}' | sort -u >> "$VOCABULARY"
igor@5: rm -f "$TEMP1" "$TEMP2" "${TEMP1}-full" "$ORIGINAL_TEXT"