new-words

annotate new-words-py.sh @ 38:adbc809d3924

find changesets by author, revision, files, or words in the commit message
Transition to Python started

new-words-py.sh is a wrapper around
new-words.py version which is not finished yet.
author Igor Chubin <igor@chub.in>
date Sat Jan 22 23:42:31 2011 +0100 (2011-01-22)
parents
children a598e0d25784
rev   line source
igor@38 1 #!/bin/bash
igor@38 2
igor@38 3 show_usage()
igor@38 4 {
igor@38 5 cat <<HELP > /dev/stderr
igor@38 6
igor@38 7 USAGE:
igor@38 8
igor@38 9 new-words [ -l lang ] [ -s ] [ ARG ]
igor@38 10
igor@38 11 SWITCHES:
igor@38 12
igor@38 13 -h print this screen
igor@38 14 -c show compressed wordlist: one word per group
igor@38 15 -k put higher words that are similar to the known words (only for English)
igor@38 16 -l lang override language settings
igor@38 17 -n non-interactive mode (don't run vi)
igor@38 18 -N turn off known words filtering
igor@38 19 -a don't add marks (and don't save marks added by user)
igor@38 20 -p pages work with specified pages only (pages = start-stop/total )
igor@38 21 -s show the text statistics (percentage of known words and so on) and exit
igor@38 22 -S show your vocabulary statistics (number of words and word groups)
igor@38 23 -t tag tag known words with tag
igor@38 24 -T show list of active tags
igor@38 25 -m tag merge the words tagged with "tag" into the main vocabulary
igor@38 26 -M merge the words tagged with any tag into the main vocabulary
igor@38 27 -r tag remove subvocabulary for the "tag"
igor@38 28 -2 -3 find 2 and 3 words' sequences
igor@38 29
igor@38 30 The language of the text can be specified also
igor@38 31 by name of the program new-words (correspondent link must be created before).
igor@38 32 For example, these calls are equivalent:
igor@38 33
igor@38 34 de-words URL
igor@38 35 new-words -l de URL
igor@38 36
igor@38 37 HELP
igor@38 38 }
igor@38 39
igor@38 40 if [ "$1" = "-h" ]
igor@38 41 then
igor@38 42 show_usage
igor@38 43 exit 0
igor@38 44 fi
igor@38 45
igor@38 46 NEW_WORDS_PY=/home/igor/hg/new-words/new-words.py
igor@38 47 WORK_DIR=~/.new-words/
igor@38 48 TEMP1=`mktemp /tmp/new-words-temp1.XXXXXXXXXX`
igor@38 49 TEMP2=`mktemp /tmp/new-words-temp2.XXXXXXXXXX`
igor@38 50 export ORIGINAL_TEXT=`mktemp /tmp/new-words-orig.XXXXXXXXXX`
igor@38 51 editor=${EDITOR:-vim}
igor@38 52
igor@38 53 # language detection
igor@38 54
igor@38 55 LANGUAGE=en
igor@38 56 my_name="`echo $0 | sed s@.*/@@ | sed s/-.*// `"
igor@38 57 for arg
igor@38 58 do
igor@38 59 if echo "$arg" | grep -q http://...wikipedia.org/wiki/
igor@38 60 then
igor@38 61 LANGUAGE="`echo $arg | sed s@http://@@ | sed s@.wikipedia.*@@`"
igor@38 62 fi
igor@38 63 done
igor@38 64 [ "${my_name}" = "new" ] || LANGUAGE="$my_name"
igor@38 65
igor@38 66 #----------------------------------------------------
igor@38 67 # command line options processing
igor@38 68
igor@38 69 STAT_ONLY=NO
igor@38 70 NEED_TO_USE_VOCABULARY_WHEN_SORT=NO
igor@38 71 DONT_ADD_MARKS=NO
igor@38 72 NON_INTERACTIVE_MODE=NO
igor@38 73 PART_TO_PROCESS=''
igor@38 74 GROUP_WORDS_BY_THREE=NO
igor@38 75 GROUP_WORDS_BY_TWO=NO
igor@38 76 TAG_NAME=''
igor@38 77 MERGE_THIS_TAGS=''
igor@38 78 TAGS_LIST_ONLY=NO
igor@38 79 MERGE_TAGGED_WORDS=NO
igor@38 80 MERGE_ALL_TAGGED=NO
igor@38 81 DONT_ADD_MARKLINES=NO
igor@38 82 FILTER_WORDS=YES
igor@38 83 SHOW_VOC_STAT=NO
igor@38 84 COMPRESSED_WORDLIST=NO
igor@38 85 OLD_STYLE="NO"
igor@38 86 while getopts Ocl:sSkanNp:t:Tm:Mr:23 opt
igor@38 87 do
igor@38 88 case "$opt" in
igor@38 89 O) OLD_STYLE=YES;;
igor@38 90 c) COMPRESSED_WORDLIST=YES;;
igor@38 91 s) STAT_ONLY=YES;;
igor@38 92 S) SHOW_VOC_STAT=YES;;
igor@38 93 k) NEED_TO_USE_VOCABULARY_WHEN_SORT=YES;;
igor@38 94 l) LANGUAGE="$OPTARG";;
igor@38 95 a) DONT_ADD_MARKS=YES;;
igor@38 96 n) NON_INTERACTIVE_MODE=YES;;
igor@38 97 N) FILTER_WORDS=NO;;
igor@38 98 p) PART_TO_PROCESS="$OPTARG";;
igor@38 99 t) TAG_NAME="$OPTARG";;
igor@38 100 T) TAGS_LIST_ONLY="YES";;
igor@38 101 m) DONT_ADD_MARKLINES="YES"; MERGE_TAGGED_WORDS="YES"; MERGE_THIS_TAGS="$TAG_NAME $OPTARG";;
igor@38 102 M) DONT_ADD_MARKLINES="YES"; MERGE_ALL_TAGGED="YES";;
igor@38 103 r) REMOVE_TAG="YES"; TAG_NAME="$TAG_NAME $OPTARG";;
igor@38 104 2) GROUP_WORDS_BY_TWO=YES;;
igor@38 105 3) GROUP_WORDS_BY_THREE=YES;;
igor@38 106 \?) # unknown flag
igor@38 107 show_usage
igor@38 108 exit 1;;
igor@38 109 esac
igor@38 110 done
igor@38 111 shift `expr $OPTIND - 1`
igor@38 112
igor@38 113 if [ "$1" = "-l" ]
igor@38 114 then
igor@38 115 LANGUAGE="$2"
igor@38 116 shift 2
igor@38 117 fi
igor@38 118
igor@38 119 VOCABULARY=${LANGUAGE}.txt
igor@38 120 NOTES_FILE=notes-${LANGUAGE}.txt
igor@38 121
igor@38 122 if [ "${SHOW_VOC_STAT}" = "YES" ]
igor@38 123 then
igor@38 124 $0 -l "${LANGUAGE}" -N -n ${WORK_DIR}/${VOCABULARY} | head -1 | awk '{print $5}' | tr -d "<>"
igor@38 125 exit 0
igor@38 126 fi
igor@38 127
igor@38 128 #----------------------------------------------------
igor@38 129
igor@38 130 get_words()
igor@38 131 {
igor@38 132 if [ "$OLD_STYLE" = NO ]
igor@38 133 then
igor@38 134 $NEW_WORDS_PY -l "$LANGUAGE" -f get_words "$1"
igor@38 135 else
igor@38 136 get_words_OLD "$@"
igor@38 137 fi
igor@38 138 }
igor@38 139
igor@38 140 get_words_OLD()
igor@38 141 {
igor@38 142 export FILTER_WORDS
igor@38 143 tr ' ' '\n' | sed 's/--/ /g' \
igor@38 144 | sed "s/'/__APOSTROPHE__/g" \
igor@38 145 | perl -MEncode -Mutf8 -n -e '$_ = decode( "utf8", $_);y/*\r,.:#@()+=—<>$;"?!|·[]^%&/ /; binmode STDOUT, ":utf8"; print if /^[[:alpha:] '"'"'_-]*$/'\
igor@38 146 | sed "s/__APOSTROPHE__/'/g" \
igor@38 147 | tr ' ' '\n' \
igor@38 148 | tee "$1" \
igor@38 149 | grep_v_english_perl \
igor@38 150 | sort | uniq -c | awk '{if ($2!="") print;}' | sort -rn
igor@38 151 }
igor@38 152
igor@38 153 add_stat()
igor@38 154 {
igor@38 155 if [ "$DONT_ADD_MARKLINES" = "YES" ]
igor@38 156 then
igor@38 157 cat
igor@38 158 return
igor@38 159 fi
igor@38 160 before="$1"
igor@38 161 after=${before}2
igor@38 162 cat > "$after"
igor@38 163 total="`wc -w $1 | awk '{print $1}'`"
igor@38 164 total_unknown="`cat $after|awk '{s=s+$1}END{print s}'`"
igor@38 165 total_known="`echo $total-$total_unknown|bc`"
igor@38 166 percentage="`echo '100*('$total-$total_unknown')'/$total | bc -l | sed 's/\\.\(.\).*/.\1/'`"
igor@38 167 #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@38 168 sentences="`cat $ORIGINAL_TEXT | perl -e 'local $/; $_=<>; s/[^.]//msg; print length($_);'`"
igor@38 169
igor@38 170
igor@38 171 if [ "$STAT_ONLY" = "YES" ]
igor@38 172 then
igor@38 173 echo "LANG KNOWN% UNKNOWN% KNOWN TOTAL WPS UWPS*10"
igor@38 174 echo "$LANGUAGE $percentage `echo \(100-$percentage\) | bc -l` $total_known $total `echo $total/$sentences|bc` `echo 10*$total_unknown/$sentences|bc` "
igor@38 175 rm $after
igor@38 176 return 0
igor@38 177 else
igor@38 178 groups="`echo $(grep '# groups' $after | awk '{print $3}')`"
igor@38 179 words="`echo $(grep -v '^#' $after | wc -l)`"
igor@38 180 echo "# $LANGUAGE, $percentage, <$total_known/$total>, <$groups/$words>"
igor@38 181 fi
igor@38 182
igor@38 183 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@38 184 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@38 185 my $total=shift(@ARGV);
igor@38 186 my $total_known=shift(@ARGV);
igor@38 187 my $s=0;
igor@38 188 my $mark_line=int($total_known*100/$total/5)*5;
igor@38 189 if ($mark_line>=90) {
igor@38 190 $mark_line=int($total_known*100/$total)+1;
igor@38 191 } else { $mark_line +=5; };
igor@38 192 while(<>)
igor@38 193 {
igor@38 194 next if /^#\s*groups\s*/;
igor@38 195 print;
igor@38 196 /^\s*([0-9]*)\s*/;
igor@38 197 $s+=$1;
igor@38 198 if (($total_known+$s)*100/$total>=$mark_line) {
igor@38 199 print "# $mark_line\n";
igor@38 200 if ($mark_line>=90) { $mark_line+=1; } else { $mark_line +=5; };
igor@38 201 }
igor@38 202 }
igor@38 203 PERL_SCRIPT
igor@38 204 perl $PERL_SCRIPT_TEMP_NAME "$total" "$total_known" "$after"
igor@38 205 rm $PERL_SCRIPT_TEMP_NAME
igor@38 206 rm $after
igor@38 207 }
igor@38 208
igor@38 209 two_and_three_words()
igor@38 210 {
igor@38 211 if [ $GROUP_WORDS_BY_THREE = NO -a $GROUP_WORDS_BY_TWO = NO ]
igor@38 212 then
igor@38 213 cat
igor@38 214 else
igor@38 215 cat
igor@38 216
igor@38 217 export GROUP_WORDS_BY_THREE
igor@38 218 export GROUP_WORDS_BY_TWO
igor@38 219 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-two-and-three-XXXXXXXX`
igor@38 220 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@38 221 #!/usr/bin/perl
igor@38 222 local $/;
igor@38 223 $words=<>;
igor@38 224 $words=~ s@[!?;,:#1-9".]@ @g;
igor@38 225 $words =~ s@\s+@ @g;
igor@38 226 @words = split /\s+/, $words;
igor@38 227 for ($i=0; $i<$#words-3;$i++) {
igor@38 228 my ($a,$b,$c)= ($words[$i],$words[$i+1],$words[$i+2]);
igor@38 229 if ($ENV{GROUP_WORDS_BY_THREE} eq "YES" and ($a && $b && $c)) {
igor@38 230 print "${a}_${b}_${c}\n";
igor@38 231 };
igor@38 232 if ($ENV{GROUP_WORDS_BY_TWO} eq "YES" and ($a && $b)) {
igor@38 233 print "${a}_${b}\n";
igor@38 234 };
igor@38 235 }
igor@38 236 PERL_SCRIPT
igor@38 237 perl $PERL_SCRIPT_TEMP_NAME "$ORIGINAL_TEXT"
igor@38 238 rm $PERL_SCRIPT_TEMP_NAME
igor@38 239 fi
igor@38 240 }
igor@38 241
igor@38 242 grep_v_english()
igor@38 243 {
igor@38 244 [ -e "$VOCABULARY" ] || touch "$VOCABULARY"
igor@38 245 eval $(cat $VOCABULARY | tr -d "'" | xargs -n10 echo | tr ' ' '|' | sed 's/^/egrep -xv "RRRRRRR|/' | sed 's/$/"/' | tr '\n' '|')cat
igor@38 246 }
igor@38 247
igor@38 248 grep_v_english_perl()
igor@38 249 {
igor@38 250 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@38 251 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@38 252 if ($ENV{FILTER_WORDS} eq "NO") {
igor@38 253 while(<>) { print; }
igor@38 254 exit(0);
igor@38 255 }
igor@38 256 $voc_files=$ENV{VOC_FILES};
igor@38 257 $voc_files=~s@^ @@;
igor@38 258 for $voc_file (split /\s+/,$voc_files) {
igor@38 259 if (open(VOC, $voc_file)) {
igor@38 260 while (<VOC>){
igor@38 261 chomp;
igor@38 262 #s/'//g;
igor@38 263 $voc{$_}="1";
igor@38 264 }
igor@38 265 }
igor@38 266 }
igor@38 267 while(<>) {
igor@38 268 chomp;
igor@38 269 if (not defined($voc{$_})) { print "$_\n"; }
igor@38 270 }
igor@38 271 PERL_SCRIPT
igor@38 272 [ -e "$VOCABULARY" ] || touch "$VOCABULARY"
igor@38 273 export VOCABULARY VOC_FILES
igor@38 274 VOC_FILES=$VOCABULARY
igor@38 275 for i in $TAG_NAME
igor@38 276 do
igor@38 277 VOC_FILES="${VOC_FILES} `tag_file_name $i`"
igor@38 278 done
igor@38 279 perl $PERL_SCRIPT_TEMP_NAME
igor@38 280 rm $PERL_SCRIPT_TEMP_NAME
igor@38 281 }
igor@38 282
igor@38 283 group_words()
igor@38 284 {
igor@38 285 if [ "$OLD_STYLE" = NO ]
igor@38 286 then
igor@38 287 $NEW_WORDS_PY -l "$LANGUAGE" -f group_words "$1"
igor@38 288 else
igor@38 289 group_words_OLD "$@"
igor@38 290 fi
igor@38 291 }
igor@38 292
igor@38 293 group_words_OLD()
igor@38 294 {
igor@38 295 #if [ "$LANGUAGE" != "en" ]
igor@38 296 #then
igor@38 297 # cat
igor@38 298 # return
igor@38 299 #fi
igor@38 300 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-group-words-XXXXXXXX`
igor@38 301 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@38 302 #!/usr/bin/perl
igor@38 303
igor@38 304 use Encode;
igor@38 305 use utf8;
igor@38 306 use Lingua::Stem::Snowball qw(stem);
igor@38 307
igor@38 308 eval {
igor@38 309 # http://stackoverflow.com/questions/251694/how-can-i-check-if-i-have-a-perl-module-before-using-it
igor@38 310 require String::Similarity;
igor@38 311 String::Similarity->import();
igor@38 312 };
igor@38 313 unless($@)
igor@38 314 {
igor@38 315 our $HAVE_String_Similarity=1;
igor@38 316 }
igor@38 317
igor@38 318
igor@38 319 sub load_notes_dict()
igor@38 320 {
igor@38 321 my %dict;
igor@38 322 if (open(NOTES, $ENV{NOTES_FILE})) {
igor@38 323 while(<NOTES>) {
igor@38 324 $_ = decode( "utf8", $_);
igor@38 325 chomp;
igor@38 326 s/^\s+//;
igor@38 327 my ($a,$b)=split /\s+/,$_,2;
igor@38 328 $dict{$a}=$b;
igor@38 329 }
igor@38 330 }
igor@38 331 return %dict;
igor@38 332 }
igor@38 333
igor@38 334 sub similar($$){
igor@38 335 my $a=shift;
igor@38 336 my $b=shift;
igor@38 337 if ($HAVE_String_Similarity) {
igor@38 338 return $Similarity{"$a $b"};
igor@38 339 }
igor@38 340 else {
igor@38 341 return 0;
igor@38 342 }
igor@38 343 }
igor@38 344
igor@38 345
igor@38 346 sub normalize_without_linked($)
igor@38 347 {
igor@38 348 if ( $ENV{LANGUAGE} eq "en" ) { return normalize_english(shift); }
igor@38 349 elsif ( $ENV{LANGUAGE} eq "de" ) { return normalize_german(shift); }
igor@38 350 elsif ( $ENV{LANGUAGE} eq "uk" ) { return normalize_ukrainian(shift); }
igor@38 351 elsif ( $ENV{LANGUAGE} eq "io" ) { return normalize_esperanto(shift); }
igor@38 352 else { return shift ; }
igor@38 353 }
igor@38 354
igor@38 355 sub normalize_with_linked($)
igor@38 356 {
igor@38 357 my $word = normalize_without_linked(shift);
igor@38 358 #return $word;
igor@38 359 if ($linked_words{$word}) {
igor@38 360 return $linked_words{$word};
igor@38 361 }
igor@38 362 else {
igor@38 363 return $word;
igor@38 364 }
igor@38 365 }
igor@38 366
igor@38 367 sub normalize($)
igor@38 368 {
igor@38 369 return normalize_with_linked(shift);
igor@38 370 }
igor@38 371
igor@38 372 sub normalize_ukrainian($)
igor@38 373 {
igor@38 374 $_=lc(shift);
igor@38 375 s/[юіоеуаи]$//g;
igor@38 376 return $_;
igor@38 377 }
igor@38 378
igor@38 379 sub normalize_esperanto($)
igor@38 380 {
igor@38 381 $_=lc(shift);
igor@38 382 # verbs
igor@38 383 s/i$//; s/is$//; s/os$//; s/as$//; s/us$//;
igor@38 384
igor@38 385 # nouns
igor@38 386 s/j?n?$//;
igor@38 387
igor@38 388 return $_;
igor@38 389 }
igor@38 390
igor@38 391 sub normalize_german($)
igor@38 392 {
igor@38 393 @stems = stem('de', \@_);
igor@38 394 return $stems[0];
igor@38 395 }
igor@38 396
igor@38 397 sub normalize_german_($)
igor@38 398 {
igor@38 399 $_=lc(shift);
igor@38 400
igor@38 401 s/heit$//; s/keit$//; s/tum$//; s/ung$//; s/nis$//;s/schaft$//; s/ist$//;
igor@38 402 s/en$//; s/er$//;
igor@38 403
igor@38 404 s/lich$//; s/ig$//;
igor@38 405 s/al$//; s/isch$//;
igor@38 406 s/ell$//; s/haft$//;
igor@38 407
igor@38 408 s/bar$//; s/sam$//; s/lich$//;
igor@38 409
igor@38 410 @prefixes=qw(
igor@38 411 ab an auf aus bei dazwischen ein fest heraus her hinaus hin los mit nach voraus vorbei vor weg weiter zurück zusammen zu
igor@38 412 be emp ent er ge miss ver zer durch über um unter wieder);
igor@38 413 @prefixes=();
igor@38 414 for $pref (@prefixes) {
igor@38 415 s/^$pref//;
igor@38 416 }
igor@38 417
igor@38 418
igor@38 419 return $_;
igor@38 420 }
igor@38 421
igor@38 422 sub normalize_english($)
igor@38 423 {
igor@38 424 $_=lc(shift);
igor@38 425
igor@38 426 s/s$//;
igor@38 427
igor@38 428 s/ation$//; s/ness$//; s/ship$//; s/ally$//; s/ance$//;s/ity$//; s/ment$//;
igor@38 429
igor@38 430 s/ed$//;
igor@38 431 s/en$//;
igor@38 432 s/er$//;
igor@38 433 s/est$//;
igor@38 434 s/ing$//;
igor@38 435
igor@38 436 s/ism$//; s/ist$//; s/ful$//; s/able$//; s/ably$//;
igor@38 437 s/ify$//; s/fy$//; s/ly$//;
igor@38 438 s/ise$//; s/ize$//;
igor@38 439
igor@38 440 s/e$//;
igor@38 441 return $_;
igor@38 442 }
igor@38 443
igor@38 444
igor@38 445 sub compare($$)
igor@38 446 {
igor@38 447 my $a=shift;
igor@38 448 my $b=shift;
igor@38 449 $a =~ s/^\s*//;
igor@38 450 $b =~ s/^\s*//;
igor@38 451 my ($a1, $a2)= split /\s+/,$a,2;
igor@38 452 my ($b1, $b2)= split /\s+/,$b,2;
igor@38 453
igor@38 454 my $cmp = $group_weight{normalize($a2)} <=> $group_weight{normalize($b2)};
igor@38 455
igor@38 456 if ($cmp) {
igor@38 457 return $cmp;
igor@38 458 }
igor@38 459 else {
igor@38 460 if (normalize($a2) ne normalize($b2)) {
igor@38 461 return normalize($a2) cmp normalize($b2);
igor@38 462 }
igor@38 463 else {
igor@38 464 return $a1 <=> $b1;
igor@38 465 }
igor@38 466 }
igor@38 467 }
igor@38 468
igor@38 469 sub log_($)
igor@38 470 {
igor@38 471 return;
igor@38 472 open(LOG, ">>", "/tmp/log1");
igor@38 473 print LOG $_[0];
igor@38 474 close(LOG);
igor@38 475 }
igor@38 476
igor@38 477 sub find_linked_words($)
igor@38 478 {
igor@38 479 my %linked_words;
igor@38 480 my $dict = shift;
igor@38 481 log_("1");
igor@38 482 log_(join(" ", keys(%$dict)));
igor@38 483
igor@38 484 for $key (keys(%$dict)) {
igor@38 485 $val = $dict->{$key};
igor@38 486 log_($key."\n");
igor@38 487 if ($val =~ /\@([a-z]*)/) {
igor@38 488 $linked_words{normalize($key)} = normalize($1);
igor@38 489 log_(normalize($key)." = ".normalize($1)."\n");
igor@38 490 }
igor@38 491 }
igor@38 492 return %linked_words;
igor@38 493 }
igor@38 494
igor@38 495 sub lc_length($)
igor@38 496 {
igor@38 497 my $a= shift;
igor@38 498 $a =~ s/[a-z]//g;
igor@38 499 return length($a);
igor@38 500 }
igor@38 501
igor@38 502 our %dict = load_notes_dict();
igor@38 503 our %linked_words = find_linked_words(\%dict);
igor@38 504
igor@38 505 our %Vocabulary;
igor@38 506 open(VOC, $ENV{VOCABULARY})
igor@38 507 or die "Can't open VOCABULARY";
igor@38 508 while (<VOC>){
igor@38 509 chomp;
igor@38 510 #s/'//g;
igor@38 511 $Vocabulary{normalize($_)}="1";
igor@38 512 }
igor@38 513 close(VOC);
igor@38 514
igor@38 515 binmode STDIN,":utf8";
igor@38 516 @lines=<STDIN>;
igor@38 517 for $L (@lines) {
igor@38 518 chomp($L);
igor@38 519 #$L = decode( "utf8", $L);
igor@38 520 $l=$L;
igor@38 521 $l =~ s/^\s*//;
igor@38 522 my ($a, $b)=split(/\s+/,$l,2);
igor@38 523 $group_weight{normalize($b)}+=$a;
igor@38 524 }
igor@38 525 if ($ENV{NEED_TO_USE_VOCABULARY_WHEN_SORT} eq "YES") {
igor@38 526 for $k (keys %group_weight) {
igor@38 527 if (defined($Vocabulary{$k})) {
igor@38 528 $group_weight{$k} *= 2;
igor@38 529 }
igor@38 530 }
igor@38 531 }
igor@38 532 @lines2 = sort { compare($b,$a) } @lines;
igor@38 533 binmode STDOUT, ":utf8";
igor@38 534 print "# groups ".scalar(keys(%group_weight))."\n";
igor@38 535 if ($ENV{COMPRESSED_WORDLIST} eq "YES") {
igor@38 536 my $sum = 0;
igor@38 537 my $min = 9999;
igor@38 538 for $L (@lines2) {
igor@38 539 chomp($L); $l=$L; $l =~ s/^(\s*)//; my $spaces = $1; my ($a, $b)=split(/\s+/,$l,2);
igor@38 540 $group_name = normalize($b);
igor@38 541 if ($group_name ne $prev_group_name and $prev_group_name ne '' ) {
igor@38 542 #print (" "x(7-length($sum))),"$sum $main_word\n";
igor@38 543 print +(" "x(7-length($sum))),"$sum $main_word\n";
igor@38 544 $sum = $a;
igor@38 545 $min = length($b) + 2*lc_length($b);
igor@38 546 $main_word = $b;
igor@38 547 }
igor@38 548 else {
igor@38 549 $sum += $a;
igor@38 550 if ($min > length($b) + 2*lc_length($b)) {
igor@38 551 $min = length($b) + 2*lc_length($b);
igor@38 552 $main_word = $b;
igor@38 553 }
igor@38 554 }
igor@38 555 $prev_group_name = $group_name;
igor@38 556 }
igor@38 557 }
igor@38 558 else {
igor@38 559 for $l (@lines2) {
igor@38 560 print "$l\n";
igor@38 561 }
igor@38 562 }
igor@38 563 PERL_SCRIPT
igor@38 564 export VOCABULARY
igor@38 565 export NEED_TO_USE_VOCABULARY_WHEN_SORT
igor@38 566 export LANGUAGE
igor@38 567 export COMPRESSED_WORDLIST
igor@38 568 [ -e "$NOTES_FILE" ] || touch "$NOTES_FILE"
igor@38 569 export NOTES_FILE
igor@38 570 perl $PERL_SCRIPT_TEMP_NAME
igor@38 571 rm $PERL_SCRIPT_TEMP_NAME
igor@38 572 }
igor@38 573
igor@38 574 text_from_url()
igor@38 575 {
igor@38 576 lynx -dump "$1" | perl -p -e 's@http://[a-zA-Z&_.:/0-9%?=,#+()\[\]~-]*@@'
igor@38 577 }
igor@38 578
igor@38 579 add_marks()
igor@38 580 {
igor@38 581 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@38 582 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@38 583 use Encode;
igor@38 584
igor@38 585 sub load_notes_dict()
igor@38 586 {
igor@38 587 my %dict;
igor@38 588 if (open(NOTES, $ENV{NOTES_FILE})) {
igor@38 589 while(<NOTES>) {
igor@38 590 $_ = decode( "utf8", $_);
igor@38 591 chomp;
igor@38 592 s/^\s+//;
igor@38 593 my ($a,$b)=split /\s+/,$_,2;
igor@38 594 $dict{$a}=$b;
igor@38 595 }
igor@38 596 }
igor@38 597 return %dict;
igor@38 598 }
igor@38 599
igor@38 600 %dict = load_notes_dict();
igor@38 601
igor@38 602 $file = $ARGV[0];
igor@38 603 if (open(F, $file)) {
igor@38 604 @lines=<F>;
igor@38 605 close(F);
igor@38 606 for (@lines) {$_ = decode( "utf8", $_);};
igor@38 607
igor@38 608 if (open(F, ">$file")) {
igor@38 609 binmode F, ":utf8";
igor@38 610 for (@lines) {
igor@38 611 m/\s+\S+\s+(\S+)/;
igor@38 612 $name=$1;
igor@38 613 if (not /^#/ and defined($dict{$name})) {
igor@38 614 chomp;
igor@38 615 $mark=$dict{$name};
igor@38 616 $space=" "x(30-length($_));
igor@38 617 print F "$_$space$mark\n";
igor@38 618 }
igor@38 619 else {
igor@38 620 print F "$_";
igor@38 621 }
igor@38 622 }
igor@38 623 close(F);
igor@38 624 }
igor@38 625 }
igor@38 626 PERL_SCRIPT
igor@38 627 [ -e "$NOTES_FILE" ] || touch "$NOTES_FILE"
igor@38 628 export NOTES_FILE
igor@38 629 perl $PERL_SCRIPT_TEMP_NAME "$1"
igor@38 630 rm $PERL_SCRIPT_TEMP_NAME
igor@38 631 }
igor@38 632
igor@38 633 remove_marks()
igor@38 634 {
igor@38 635 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@38 636 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@38 637 $file = $ARGV[0];
igor@38 638 our %dict;
igor@38 639 if (open(F, $file)) {
igor@38 640 @lines=<F>;
igor@38 641 close(F);
igor@38 642
igor@38 643 if (open(F, ">$file")) {
igor@38 644 for (@lines) {
igor@38 645 chomp;
igor@38 646 if (not /^#/ and m/(\s+)(\S+)(\s+)(\S+)(\s+)(.*)/) {
igor@38 647 my $name=$4;
igor@38 648 my $comment=$6;
igor@38 649 $dict{$name}=$comment;
igor@38 650 print F "$1$2$3$4\n";
igor@38 651 }
igor@38 652 else {
igor@38 653 print F "$_\n";
igor@38 654 }
igor@38 655 }
igor@38 656 }
igor@38 657 }
igor@38 658 if (($ENV{DONT_ADD_MARKS} ne "YES") and open(NOTES, $ENV{NOTES_FILE})) {
igor@38 659 @lines=<NOTES>;
igor@38 660 close(NOTES);
igor@38 661
igor@38 662 if (open(NOTES, ">".$ENV{NOTES_FILE})) {
igor@38 663 for (@lines) {
igor@38 664 chomp;
igor@38 665 s/^\s+//;
igor@38 666 my ($a,$b)=split /\s+/,$_,2;
igor@38 667 if (not defined($dict{$a}) || ($dict{$a} eq $b)) {
igor@38 668 print NOTES "$_\n";
igor@38 669 if (defined($dict{$a})) { unset($dict{$a}); }
igor@38 670 }
igor@38 671 }
igor@38 672 for (keys %dict) {
igor@38 673 $mark=$dict{$_};
igor@38 674 $space=" "x(30-length($_));
igor@38 675 print NOTES "$_$space$mark\n";
igor@38 676 }
igor@38 677 }
igor@38 678 }
igor@38 679 PERL_SCRIPT
igor@38 680 [ -e "$NOTES_FILE" ] || touch "$NOTES_FILE"
igor@38 681 export NOTES_FILE
igor@38 682 export DONT_ADD_MARKS
igor@38 683 perl $PERL_SCRIPT_TEMP_NAME "$1"
igor@38 684 rm $PERL_SCRIPT_TEMP_NAME
igor@38 685 }
igor@38 686
igor@38 687 part()
igor@38 688 {
igor@38 689 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-part-XXXXXXXX`
igor@38 690 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@38 691 #!/usr/bin/perl
igor@38 692
igor@38 693 my @lines=<STDIN>;
igor@38 694 my $lines=$#lines;
igor@38 695 my $interval=$ARGV[0];
igor@38 696 if (not $interval) {
igor@38 697 print @lines;
igor@38 698 }
igor@38 699 else {
igor@38 700 my ($start,$stop,$total);
igor@38 701 if ($interval =~ m@(.*)/(.*)@) {
igor@38 702 $start = $1;
igor@38 703 $total = $2;
igor@38 704 }
igor@38 705 else {
igor@38 706 $start=$interval;
igor@38 707 $total=0;
igor@38 708 }
igor@38 709 if ($start =~ m@(.*)-(.*)@) {
igor@38 710 $start = $1;
igor@38 711 $stop = $2;
igor@38 712 }
igor@38 713 if ($start =~ m@(.*)\+(.*)@) {
igor@38 714 $start = $1;
igor@38 715 $stop = $start+$2;
igor@38 716 }
igor@38 717
igor@38 718 $start=int($lines/$total*$start);
igor@38 719 $stop=int($lines/$total*$stop);
igor@38 720
igor@38 721 for($i=$start;$i<$stop;$i++){
igor@38 722 print $lines[$i];
igor@38 723 }
igor@38 724 }
igor@38 725 PERL_SCRIPT
igor@38 726 perl $PERL_SCRIPT_TEMP_NAME "$1"
igor@38 727 rm $PERL_SCRIPT_TEMP_NAME
igor@38 728 }
igor@38 729
igor@38 730 if [ "$TAGS_LIST_ONLY" = "YES" ]
igor@38 731 then
igor@38 732 cd "${WORK_DIR}"
igor@38 733 echo ${LANGUAGE}_*.txt | tr ' ' '\n' | grep -v '*' | sed 's/[^_]*_//;s/.txt$//'
igor@38 734 exit 0
igor@38 735 fi
igor@38 736
igor@38 737 tag_file_name()
igor@38 738 {
igor@38 739 echo "${LANGUAGE}_${1}.txt"
igor@38 740 }
igor@38 741
igor@38 742 if [ "$REMOVE_TAG" = "YES" ]
igor@38 743 then
igor@38 744 cd "${WORK_DIR}"
igor@38 745 for i in $TAG_NAME
igor@38 746 do
igor@38 747 echo "$TAGNAME" | grep -q '[/*?]' && continue
igor@38 748 f="`tag_file_name $i`"
igor@38 749 if [ -e "$f" ]
igor@38 750 then
igor@38 751 rm -f "$f" && echo Tag "'$i'" removed
igor@38 752 else
igor@38 753 echo Unknown tag "'$i'"
igor@38 754 fi
igor@38 755 done
igor@38 756 exit 0
igor@38 757 fi
igor@38 758
igor@38 759 mkdir -p $WORK_DIR
igor@38 760 oldpwd="$PWD"
igor@38 761 cd $WORK_DIR
igor@38 762 if [ "$MERGE_TAGGED_WORDS" = "YES" ]
igor@38 763 then
igor@38 764 VOC_FILES=''
igor@38 765 for i in $MERGE_THIS_TAGS
igor@38 766 do
igor@38 767 f=`tag_file_name $i`
igor@38 768 [ -e "$f" ] && VOC_FILES="${VOC_FILES} $f"
igor@38 769 done
igor@38 770 if [ -z "$VOC_FILES" ]
igor@38 771 then
igor@38 772 echo Unknown tags: $MERGE_THIS_TAGS > /dev/stderr
igor@38 773 else
igor@38 774 cat $VOC_FILES
igor@38 775 fi
igor@38 776 elif [ "$MERGE_ALL_TAGGED" = "YES" ]
igor@38 777 then
igor@38 778 cat ${LANGUAGE}_*.txt
igor@38 779 elif echo "$1" | grep -q http:
igor@38 780 then
igor@38 781 text_from_url "$1"
igor@38 782 elif [ "$#" != 0 ]
igor@38 783 then
igor@38 784 if echo $1 | grep -q ^/
igor@38 785 then
igor@38 786 cat "$1"
igor@38 787 else
igor@38 788 cat "$oldpwd/$1"
igor@38 789 fi
igor@38 790 else
igor@38 791 cat
igor@38 792 fi \
igor@38 793 | part $PART_TO_PROCESS \
igor@38 794 | tee $ORIGINAL_TEXT \
igor@38 795 | two_and_three_words \
igor@38 796 | get_words ${TEMP1}-full \
igor@38 797 | group_words \
igor@38 798 | add_stat ${TEMP1}-full \
igor@38 799 | tee "$TEMP1" > "$TEMP2"
igor@38 800
igor@38 801 if [ "$STAT_ONLY" = "YES" ]
igor@38 802 then
igor@38 803 cat "$TEMP1"
igor@38 804 elif [ "$NON_INTERACTIVE_MODE" = "YES" ]
igor@38 805 then
igor@38 806 cat "$TEMP1"
igor@38 807 else
igor@38 808 if [ `wc -l "$TEMP2" | awk '{print $1}'` != 0 ]
igor@38 809 then
igor@38 810 [ "$DONT_ADD_MARKS" = "YES" ] || add_marks "$TEMP2"
igor@38 811 if [ "$editor" = vim ]
igor@38 812 then
igor@38 813 vim -c 'set keywordprg='"$LANGUAGE" -c 'set iskeyword=@,48-57,/,.,-,_,+,,,#,$,%,~,=,48-255' "$TEMP2" < /dev/tty > /dev/tty
igor@38 814 else
igor@38 815 $editor "$TEMP2"
igor@38 816 fi
igor@38 817 remove_marks "$TEMP2"
igor@38 818
igor@38 819 vocabulary="$VOCABULARY"
igor@38 820 [ -n "$TAG_NAME" ] && vocabulary="`tag_file_name $TAG_NAME`"
igor@38 821 diff "$TEMP1" "$TEMP2" | awk '{print $3}' | sort -u >> "$vocabulary"
igor@38 822 fi
igor@38 823 fi
igor@38 824
igor@38 825 rm -f "$TEMP1" "$TEMP2" "${TEMP1}-full" "$ORIGINAL_TEXT"