simcop2387/DBD-SQLite-SQLcipher
1
0
Fork 0
mirror of https://github.com/DBD-SQLite/DBD-SQLite synced 2025-06-07 14:19:10 -04:00
Code Issues Releases Wiki Activity

fixed #75 -- lastCharOffset must copy the OLD value of piEndOffset, BEFORE it is recomputed

Browse source 
also fixed 43_fts3.t so that it uses \p{Word} instead of \w, because en-us locale did not handle accented characters
This commit is contained in:
dami 2021-05-09 22:06:28 +02:00
parent be9f64b2c9
commit b5c3f9d528
2 changed files with 23 additions and 26 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
dbdimp_tokenizer.inc
View file

@ -205,22 +205,33 @@ static int perl_tokenizer_Next(
*pnBytes = POPi;
token = POPpx;
if (c->pInput) { /* if working with utf8 data */
#ifdef DEBUG_OFFSETS
warn("INI: token: %s, start=%d, end=%d, nBytes=%d\n", token, *piStartOffset, *piEndOffset, *pnBytes);
#endif
/* recompute *pnBytes in bytes, not in chars */
*pnBytes = strlen(token);
/* recompute start/end offsets in bytes, not in chars */
/* recompute start offset in bytes, not in chars */
hop = *piStartOffset - c->lastCharOffset;
byteOffset = (char*)utf8_hop((U8*)c->lastByteOffset, hop);
hop = *piEndOffset - *piStartOffset;
*piStartOffset = byteOffset - c->pInput;
byteOffset = (char*)utf8_hop((U8*)byteOffset, hop);
*piEndOffset = byteOffset - c->pInput;
/* remember where we are for next round */
c->lastCharOffset = *piEndOffset,
c->lastCharOffset = *piEndOffset;
c->lastByteOffset = byteOffset;
/* recompute end offset in bytes, not in chars */
*piEndOffset = byteOffset - c->pInput;
#ifdef DEBUG_OFFSETS
warn("FIX: token: %s, start=%d, end=%d, nBytes=%d\n", token, *piStartOffset, *piEndOffset, *pnBytes);
#endif
}
/* make sure we have enough storage for copying the token */

20
t/43_fts3.t
View file

@ -42,25 +42,11 @@ BEGIN {
}
}
# Perl may spit a warning on locale
# use Test::NoWarnings;
BEGIN {
# Sadly perl for windows (and probably sqlite, too) may hang
# if the system locale doesn't support european languages.
# en-us should be a safe default. if it doesn't work, use 'C'.
if ( $^O eq 'MSWin32') {
use POSIX 'locale_h';
setlocale(LC_COLLATE, 'en-us');
}
}
use locale;
sub locale_tokenizer { # see also: Search::Tokenizer
sub Unicode_Word_tokenizer { # see also: Search::Tokenizer
return sub {
my $string = shift;
my $regex = qr/\w+/;
my $regex = qr/\p{Word}+/;
my $term_index = 0;
return sub {
@ -85,7 +71,7 @@ for my $use_unicode (0, 1) {
# create fts table
$dbh->do(<<"") or die DBI::errstr;
CREATE VIRTUAL TABLE try_$fts
USING $fts(content, tokenize=perl 'main::locale_tokenizer')
USING $fts(content, tokenize=perl 'main::Unicode_Word_tokenizer')
# populate it
my $insert_sth = $dbh->prepare(<<"") or die DBI::errstr;