p
/
tdv
1
1
Fork 0
Code Issues Pull Requests Releases Wiki Activity

gnu-fdl-en-cz.sh: employ Pango formatting 

Also add collation fields.
This commit is contained in:
Přemysl Eric Janouch 2021-10-10 06:11:28 +02:00
parent 16d6eaf012
commit de7089d669
Signed by: p
GPG Key ID: A0420B94F92B9493
1 changed files with 18 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

28
dicts/gnu-fdl-en-cz.sh
View File

@ -2,20 +2,28 @@
# GNU/FDL English-Czech dictionary, see https://www.svobodneslovniky.cz/ # GNU/FDL English-Czech dictionary, see https://www.svobodneslovniky.cz/
curl -Lo- https://www.svobodneslovniky.cz/data/en-cs.txt.gz | \ curl -Lo- https://www.svobodneslovniky.cz/data/en-cs.txt.gz | \
zcat | grep -v ^# | sed 's/\\//g' | perl -CSD -F\\t -le ' zcat | grep -v ^# | sed 's/\\//g' | perl -CSD -F\\t -le '
sub e { shift =~ s/\\/\\\\/gr =~ s/\n/\\n/gr =~ s/\t/\\t/gr } sub tabesc { shift =~ s/\\/\\\\/gr =~ s/\n/\\n/gr =~ s/\t/\\t/gr }
sub w { sub w {
open(my $f, "|-", "tabfile gnu-fdl-$_[0]") or die $!; my ($name, $dict, $collation) = @_;
print $f e($k) . "\t" . e(join("\n", @$v)) open(my $f, "|-", "tabfile", "--pango", "--collation=$collation",
while ($k, $v) = each %{$_[1]}; "--website=https://www.svobodneslovniky.cz",
"gnu-fdl-$name") or die $!;
print $f tabesc($keyword) . "\t" . tabesc(join("\n", @$defs))
while ($keyword, $defs) = each %{$dict};
close($f); close($f);
} }
my ($en, $cz, $notes, $special, $translator) = @F; sub xmlesc { shift =~ s/&/&amp;/gr =~ s/</&lt;/gr =~ s/>/&gt;/gr }
if ($cz) { sub entry {
my ($definition, $notes) = map {xmlesc($_)} @_;
$notes ? "$definition <i>$notes</i>" : $definition;
}
my ($en, $cs, $notes, $special, $translator) = @F;
if ($cs) {
$notes =~ s/\w+:\s?//g; # remove word classes $notes =~ s/\w+:\s?//g; # remove word classes
$notes =~ s/(\w+\.)(?!])/($1)/; # quote "pl." $notes =~ s/(\w+\.)(?!])/($1)/; # quote "pl."
push(@{$encz{$en}}, $notes ? "$cz " . $notes : $cz); push(@{$encs{$en}}, entry($cs, $notes));
push(@{$czen{$cz}}, $notes ? "$en " . $notes : $en); push(@{$csen{$cs}}, entry($en, $notes));
} END { } END {
w("en-cz", \%encz); w("en-cz", \%encs, "en");
w("cz-en", \%czen); w("cz-en", \%csen, "cs");
}' }'