new-words

annotate new-words.sh @ 21:190d4ac6b07c

find changesets by author, revision, files, or words in the commit message
-m and -n switches
author Igor Chubin <igor@chub.in>
date Thu May 06 21:05:20 2010 +0300 (2010-05-06)
parents 23f949c03f95
children 46e987f4636d
rev   line source
igor@0 1 #!/bin/sh
igor@20 2
igor@20 3 show_usage()
igor@20 4 {
igor@20 5 cat <<HELP > /dev/stderr
igor@0 6
igor@16 7 USAGE:
igor@16 8
igor@16 9 new-words [ -l lang ] [ -s ] [ ARG ]
igor@16 10
igor@16 11 SWITCHES:
igor@16 12
igor@20 13 -h print this screen
igor@20 14 -k put higher words that are similar to the known words (only for English)
igor@16 15 -l lang override language settings
igor@21 16 -n non-interactive mode (don't run vi)
igor@21 17 -m don't add marks (and don't save marks added by user)
igor@20 18 -s show the text statistics (percentage of known words and so on) and exit
igor@16 19
igor@20 20 The language of the text can be specified also
igor@20 21 by name of the program new-words (correspondent link must be created before).
igor@20 22 For example, these calls are equivalent:
igor@0 23
igor@20 24 de-words URL
igor@0 25 new-words -l de URL
igor@0 26
igor@20 27 HELP
igor@20 28 }
igor@0 29
igor@20 30 if [ "$1" = "-h" ]
igor@20 31 then
igor@20 32 show_usage
igor@20 33 exit 0
igor@20 34 fi
igor@0 35
igor@0 36 WORK_DIR=~/.new-words/
igor@17 37 TEMP1=`mktemp /tmp/new-words-XXXXXXXXXX-temp1`
igor@17 38 TEMP2=`mktemp /tmp/new-words-XXXXXXXXXX-temp2`
igor@17 39 export ORIGINAL_TEXT=`mktemp /tmp/new-words-XXXXXXXXXX-orig`
igor@0 40 editor=${EDITOR:-vim}
igor@0 41
igor@21 42 #----------------------------------------------------
igor@21 43 # command line options processing
igor@21 44
igor@2 45 LANGUAGE=en
igor@2 46 my_name="`echo $0 | sed s@.*/@@ | sed s/-.*// `"
igor@3 47 if echo "$1" | grep -q http://...wikipedia.org/wiki/
igor@3 48 then
igor@3 49 LANGUAGE="`echo $1 | sed s@http://@@ | sed s@.wikipedia.*@@`"
igor@3 50 fi
igor@2 51 [ "${my_name}" = "new" ] || LANGUAGE="$my_name"
igor@0 52 if [ "$1" = "-l" ]
igor@0 53 then
igor@0 54 LANGUAGE="$2"
igor@0 55 VOCABULARY="$LANGUAGE".txt
igor@0 56 shift 2
igor@0 57 fi
igor@2 58 VOCABULARY=${LANGUAGE}.txt
igor@2 59 NOTES_FILE=notes-${LANGUAGE}.txt
igor@0 60
igor@16 61 STAT_ONLY=NO
igor@16 62 if [ "$1" = "-s" ]
igor@16 63 then
igor@16 64 STAT_ONLY=YES
igor@16 65 shift
igor@16 66 fi
igor@16 67
igor@20 68 NEED_TO_USE_VOCABULARY_WHEN_SORT=NO
igor@20 69 if [ "$1" = "-k" ]
igor@20 70 then
igor@20 71 NEED_TO_USE_VOCABULARY_WHEN_SORT=YES
igor@20 72 shift
igor@20 73 fi
igor@20 74
igor@21 75 DONT_ADD_MARKS=NO
igor@21 76 if [ "$1" = "-m" ]
igor@21 77 then
igor@21 78 DONT_ADD_MARKS=YES
igor@21 79 shift
igor@21 80 fi
igor@21 81
igor@21 82 NON_INTERACTIVE_MODE=NO
igor@21 83 if [ "$1" = "-n" ]
igor@21 84 then
igor@21 85 NON_INTERACTIVE_MODE=YES
igor@21 86 shift
igor@21 87 fi
igor@21 88
igor@21 89 #----------------------------------------------------
igor@21 90
igor@0 91 get_words()
igor@0 92 {
igor@1 93 tr ' ' '\n' | sed 's/--/ /g' \
igor@3 94 | sed "s/'/__APOSTROPHE__/g" \
igor@3 95 | tr '—·-' '-----' \
igor@3 96 | tr '*\r,.:#@()+=—<>$;"?!|·[]^%&' ' ' \
igor@3 97 | tr ' ' '\n' \
igor@3 98 | grep -x '[[:alpha:]'"'"'-]*' \
igor@3 99 | tee "$1" \
igor@3 100 | grep_v_english_perl \
igor@0 101 | sort | uniq -c | awk '{if ($2!="") print;}' | sort -rn
igor@0 102 }
igor@0 103
igor@3 104 add_stat()
igor@3 105 {
igor@3 106 before="$1"
igor@3 107 after=${before}2
igor@3 108 cat > "$after"
igor@3 109 total="`wc -w $1 | awk '{print $1}'`"
igor@3 110 total_unknown="`cat $after|awk '{s=s+$1}END{print s}'`"
igor@3 111 total_known="`echo $total-$total_unknown|bc`"
igor@3 112 percentage="`echo '100*('$total-$total_unknown')'/$total | bc -l | sed 's/\\.\(.\).*/.\1/'`"
igor@18 113 #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@18 114 sentences="`cat $ORIGINAL_TEXT | perl -e 'local $/; $_=<>; s/[^.]//msg; print length($_);'`"
igor@18 115
igor@18 116
igor@16 117 if [ "$STAT_ONLY" = "YES" ]
igor@16 118 then
igor@18 119 echo "LANG KNOWN% UNKNOWN% KNOWN TOTAL WPS UWPS*10"
igor@18 120 echo "$LANGUAGE $percentage `echo \(100-$percentage\) | bc -l` $total_known $total `echo $total/$sentences|bc` `echo 10*$total_unknown/$sentences|bc` "
igor@17 121 rm $after
igor@16 122 return 0
igor@16 123 else
igor@16 124 echo "# $LANGUAGE, $percentage, <$total_known/$total>"
igor@16 125 fi
igor@16 126
igor@14 127 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@14 128 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@14 129 my $total=shift(@ARGV);
igor@14 130 my $total_known=shift(@ARGV);
igor@14 131 my $s=0;
igor@16 132 my $mark_line=int($total_known*100/$total/5)*5;
igor@19 133 if ($mark_line>=90) {
igor@19 134 $mark_line=int($total_known*100/$total)+1;
igor@19 135 } else { $mark_line +=5; };
igor@14 136 while(<>)
igor@14 137 {
igor@14 138 print;
igor@14 139 /^\s*([0-9]*)\s*/;
igor@14 140 $s+=$1;
igor@19 141 if (($total_known+$s)*100/$total>=$mark_line) {
igor@14 142 print "# $mark_line\n";
igor@14 143 if ($mark_line>=90) { $mark_line+=1; } else { $mark_line +=5; };
igor@14 144 }
igor@14 145 }
igor@14 146 PERL_SCRIPT
igor@14 147 perl $PERL_SCRIPT_TEMP_NAME "$total" "$total_known" "$after"
igor@14 148 rm $PERL_SCRIPT_TEMP_NAME
igor@3 149 rm $after
igor@3 150 }
igor@3 151
igor@0 152 grep_v_english()
igor@0 153 {
igor@0 154 [ -e "$VOCABULARY" ] || touch "$VOCABULARY"
igor@0 155 eval $(cat $VOCABULARY | tr -d "'" | xargs -n10 echo | tr ' ' '|' | sed 's/^/egrep -xv "RRRRRRR|/' | sed 's/$/"/' | tr '\n' '|')cat
igor@0 156 }
igor@0 157
igor@0 158 grep_v_english_perl()
igor@0 159 {
igor@0 160 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@0 161 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@0 162 open(VOC, $ENV{VOCABULARY})
igor@0 163 or die "Can't open VOCABULARY";
igor@0 164 while (<VOC>){
igor@0 165 chomp;
igor@3 166 #s/'//g;
igor@0 167 $voc{$_}="1";
igor@0 168 }
igor@0 169 while(<>) {
igor@0 170 chomp;
igor@0 171 if (not defined($voc{$_})) { print "$_\n"; }
igor@0 172 }
igor@0 173 PERL_SCRIPT
igor@0 174 [ -e "$VOCABULARY" ] || touch "$VOCABULARY"
igor@0 175 export VOCABULARY
igor@0 176 perl $PERL_SCRIPT_TEMP_NAME
igor@0 177 rm $PERL_SCRIPT_TEMP_NAME
igor@0 178 }
igor@0 179
igor@11 180 group_words()
igor@11 181 {
igor@11 182 if [ "$LANGUAGE" != "en" ]
igor@11 183 then
igor@11 184 cat
igor@11 185 return
igor@11 186 fi
igor@11 187 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-group-words-XXXXXXXX`
igor@11 188 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@11 189 #!/usr/bin/perl
igor@11 190
igor@12 191 eval {
igor@12 192 # http://stackoverflow.com/questions/251694/how-can-i-check-if-i-have-a-perl-module-before-using-it
igor@12 193 require String::Similarity;
igor@12 194 String::Similarity->import();
igor@12 195 };
igor@12 196 unless($@)
igor@12 197 {
igor@12 198 our $HAVE_String_Similarity=1;
igor@12 199 }
igor@12 200
igor@12 201 sub similar($$){
igor@12 202 my $a=shift;
igor@12 203 my $b=shift;
igor@12 204 if ($HAVE_String_Similarity) {
igor@12 205 return $Similarity{"$a $b"};
igor@12 206 }
igor@12 207 else {
igor@12 208 return 0;
igor@12 209 }
igor@12 210 }
igor@12 211
igor@11 212 sub normalize($)
igor@11 213 {
igor@11 214 $_=lc(shift);
igor@12 215
igor@11 216 s/s$//;
igor@11 217 s/ed$//;
igor@12 218 s/en$//;
igor@12 219 s/er$//;
igor@12 220 s/est$//;
igor@11 221 s/ing$//;
igor@12 222
igor@12 223 s/ism$//; s/ist$//; s/ful$//; s/able$//; s/ably$//;
igor@16 224 s/ation$//; s/ness$//; s/ship$//; s/ally$//;
igor@12 225 s/ment$//; s/ify$//; s/ity$//; s/fy$//; s/ly$//;
igor@12 226 s/ise$//; s/ize$//;
igor@12 227
igor@12 228 s/e$//;
igor@11 229 return $_;
igor@11 230 }
igor@11 231
igor@12 232
igor@11 233 sub compare($$)
igor@11 234 {
igor@11 235 my $a=shift;
igor@11 236 my $b=shift;
igor@11 237 $a =~ s/^\s*//;
igor@11 238 $b =~ s/^\s*//;
igor@11 239 my ($a1, $a2)= split /\s+/,$a,2;
igor@11 240 my ($b1, $b2)= split /\s+/,$b,2;
igor@11 241
igor@11 242 my $cmp = $group_weight{normalize($a2)} <=> $group_weight{normalize($b2)};
igor@12 243
igor@11 244 if ($cmp) {
igor@11 245 return $cmp;
igor@11 246 }
igor@11 247 else {
igor@11 248 if (normalize($a2) ne normalize($b2)) {
igor@11 249 return normalize($a2) cmp normalize($b2);
igor@11 250 }
igor@11 251 else {
igor@11 252 return $a1 <=> $b1;
igor@11 253 }
igor@11 254 }
igor@11 255 }
igor@11 256
igor@20 257 our %Vocabulary;
igor@20 258 open(VOC, $ENV{VOCABULARY})
igor@20 259 or die "Can't open VOCABULARY";
igor@20 260 while (<VOC>){
igor@20 261 chomp;
igor@20 262 #s/'//g;
igor@20 263 $Vocabulary{normalize($_)}="1";
igor@20 264 }
igor@20 265 close(VOC);
igor@20 266
igor@20 267 @lines=<STDIN>;
igor@11 268 for $L (@lines) {
igor@11 269 chomp($L);
igor@11 270 $l=$L;
igor@11 271 $l =~ s/^\s*//;
igor@11 272 my ($a, $b)=split(/\s+/,$l,2);
igor@11 273 $group_weight{normalize($b)}+=$a;
igor@11 274 }
igor@20 275 if ($ENV{NEED_TO_USE_VOCABULARY_WHEN_SORT} eq "YES") {
igor@20 276 for $k (keys %group_weight) {
igor@20 277 if (defined($Vocabulary{$k})) {
igor@20 278 $group_weight{$k} *= 2;
igor@20 279 }
igor@20 280 }
igor@20 281 }
igor@11 282 @lines2 = sort { compare($b,$a) } @lines;
igor@11 283 for $l (@lines2) {
igor@11 284 print "$l\n";
igor@11 285 }
igor@11 286 PERL_SCRIPT
igor@20 287 export VOCABULARY
igor@20 288 export NEED_TO_USE_VOCABULARY_WHEN_SORT
igor@11 289 perl $PERL_SCRIPT_TEMP_NAME
igor@11 290 rm $PERL_SCRIPT_TEMP_NAME
igor@11 291 }
igor@11 292
igor@0 293 text_from_url()
igor@0 294 {
igor@0 295 lynx -dump "$1" | perl -p -e 's@http://[a-zA-Z&_.:/0-9%?=,#+()\[\]~-]*@@'
igor@0 296 }
igor@0 297
igor@2 298 add_marks()
igor@2 299 {
igor@2 300 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@2 301 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@2 302 $file = $ARGV[0];
igor@2 303 our $dict;
igor@2 304 if (open(NOTES, $ENV{NOTES_FILE})) {
igor@2 305 while(<NOTES>) {
igor@2 306 chomp;
igor@2 307 s/^\s+//;
igor@2 308 my ($a,$b)=split /\s+/,$_,2;
igor@2 309 $dict{$a}=$b;
igor@2 310 }
igor@2 311 }
igor@2 312 if (open(F, $file)) {
igor@2 313 @lines=<F>;
igor@2 314 close(F);
igor@2 315
igor@2 316 if (open(F, ">$file")) {
igor@2 317 for (@lines) {
igor@2 318 m/\s+\S+\s+(\S+)/;
igor@2 319 $name=$1;
igor@3 320 if (not /^#/ and defined($dict{$name})) {
igor@2 321 chomp;
igor@2 322 $mark=$dict{$name};
igor@2 323 $space=" "x(30-length($_));
igor@2 324 print F "$_$space$mark\n";
igor@2 325 }
igor@2 326 else {
igor@2 327 print F "$_";
igor@2 328 }
igor@2 329 }
igor@2 330 close(F);
igor@2 331 }
igor@2 332 }
igor@2 333 PERL_SCRIPT
igor@2 334 [ -e "$NOTES_FILE" ] || touch "$NOTES_FILE"
igor@2 335 export NOTES_FILE
igor@2 336 perl $PERL_SCRIPT_TEMP_NAME "$1"
igor@2 337 rm $PERL_SCRIPT_TEMP_NAME
igor@2 338 }
igor@2 339
igor@2 340 remove_marks()
igor@2 341 {
igor@2 342 PERL_SCRIPT_TEMP_NAME=`mktemp /tmp/perl-grep-v-english-XXXXXXXX`
igor@2 343 cat <<'PERL_SCRIPT' > $PERL_SCRIPT_TEMP_NAME
igor@2 344 $file = $ARGV[0];
igor@2 345 our %dict;
igor@2 346 if (open(F, $file)) {
igor@2 347 @lines=<F>;
igor@2 348 close(F);
igor@2 349
igor@2 350 if (open(F, ">$file")) {
igor@2 351 for (@lines) {
igor@2 352 chomp;
igor@3 353 if (not /^#/ and m/(\s+)(\S+)(\s+)(\S+)(\s+)(.*)/) {
igor@2 354 my $name=$4;
igor@2 355 my $comment=$6;
igor@2 356 $dict{$name}=$comment;
igor@2 357 print F "$1$2$3$4\n";
igor@2 358 }
igor@2 359 else {
igor@2 360 print F "$_\n";
igor@2 361 }
igor@2 362 }
igor@2 363 }
igor@2 364 }
igor@21 365 if (($ENV{DONT_ADD_MARKS} ne "YES") and open(NOTES, $ENV{NOTES_FILE})) {
igor@2 366 @lines=<NOTES>;
igor@2 367 close(NOTES);
igor@2 368
igor@2 369 if (open(NOTES, ">".$ENV{NOTES_FILE})) {
igor@2 370 for (@lines) {
igor@2 371 chomp;
igor@2 372 s/^\s+//;
igor@2 373 my ($a,$b)=split /\s+/,$_,2;
igor@2 374 if (not defined($dict{$a}) || ($dict{$a} eq $b)) {
igor@2 375 print NOTES "$_\n";
igor@2 376 if (defined($dict{$a})) { unset($dict{$a}); }
igor@2 377 }
igor@2 378 }
igor@2 379 for (keys %dict) {
igor@2 380 $mark=$dict{$_};
igor@2 381 $space=" "x(30-length($_));
igor@2 382 print NOTES "$_$space$mark\n";
igor@2 383 }
igor@2 384 }
igor@2 385 }
igor@2 386 PERL_SCRIPT
igor@2 387 [ -e "$NOTES_FILE" ] || touch "$NOTES_FILE"
igor@2 388 export NOTES_FILE
igor@21 389 export DONT_ADD_MARKS
igor@2 390 perl $PERL_SCRIPT_TEMP_NAME "$1"
igor@2 391 rm $PERL_SCRIPT_TEMP_NAME
igor@2 392 }
igor@2 393
igor@0 394 mkdir -p $WORK_DIR
igor@8 395 oldpwd="$PWD"
igor@0 396 cd $WORK_DIR
igor@0 397 if echo "$1" | grep -q http:
igor@0 398 then
igor@11 399 text_from_url "$1"
igor@0 400 elif [ "$#" != 0 ]
igor@0 401 then
igor@13 402 if echo $1 | grep -q ^/
igor@13 403 then
igor@13 404 cat "$1"
igor@13 405 else
igor@13 406 cat "$oldpwd/$1"
igor@13 407 fi
igor@0 408 else
igor@11 409 cat
igor@11 410 fi \
igor@11 411 | tee $ORIGINAL_TEXT \
igor@11 412 | get_words ${TEMP1}-full \
igor@11 413 | group_words \
igor@11 414 | add_stat ${TEMP1}-full \
igor@11 415 | tee "$TEMP1" > "$TEMP2"
igor@0 416
igor@16 417 if [ "$STAT_ONLY" = "YES" ]
igor@0 418 then
igor@16 419 cat "$TEMP1"
igor@21 420 elif [ "$NON_INTERACTIVE_MODE" = "YES" ]
igor@21 421 then
igor@21 422 cat "$TEMP1"
igor@0 423 else
igor@21 424 [ "$DONT_ADD_MARKS" = "YES" ] || add_marks "$TEMP2"
igor@16 425 if [ "$editor" = vim ]
igor@16 426 then
igor@16 427 vim -c 'set keywordprg='"$LANGUAGE" -c 'set iskeyword=@,48-57,/,.,-,_,+,,,#,$,%,~,=' "$TEMP2" < /dev/tty > /dev/tty
igor@16 428 else
igor@16 429 echo 2
igor@16 430 $editor "$TEMP2"
igor@16 431 fi
igor@16 432 remove_marks "$TEMP2"
igor@0 433 fi
igor@2 434
igor@0 435 diff "$TEMP1" "$TEMP2" | awk '{print $3}' | sort -u >> "$VOCABULARY"
igor@5 436 rm -f "$TEMP1" "$TEMP2" "${TEMP1}-full" "$ORIGINAL_TEXT"