https://bugs.gentoo.org/921490 https://www.sqlite.org/forum/forumpost/d16aeb397d https://github.com/kovidgoyal/calibre/commit/0a23fabd5b4bff152047ba0cffbc2b67592645d1 https://github.com/kovidgoyal/calibre/commit/3349979551db05ec0c8754d7ea5a4e57a97f8a0a https://github.com/kovidgoyal/calibre/commit/da82b673f09efbe9cdcc1e0900f0a9f04d8de052 --- a/ext/fts5/fts5_aux.c +++ b/ext/fts5/fts5_aux.c @@ -209,10 +209,18 @@ rc = fts5CInstIterNext(&p->iter); } } if( iPos==p->iRangeEnd ){ + if( p->bOpen ){ + if( p->iter.iStart>=0 && iPos>=p->iter.iStart ){ + fts5HighlightAppend(&rc, p, &p->zIn[p->iOff], iEndOff - p->iOff); + p->iOff = iEndOff; + } + fts5HighlightAppend(&rc, p, p->zClose, -1); + p->bOpen = 0; + } fts5HighlightAppend(&rc, p, &p->zIn[p->iOff], iEndOff - p->iOff); p->iOff = iEndOff; } return rc; ADDED ext/fts5/test/fts5tokenizer2.test Index: ext/fts5/test/fts5tokenizer2.test ================================================================== --- /dev/null +++ b/ext/fts5/test/fts5tokenizer2.test @@ -0,0 +1,89 @@ +# 2023 Nov 03 +# +# The author disclaims copyright to this source code. In place of +# a legal notice, here is a blessing: +# +# May you do good and not evil. +# May you find forgiveness for yourself and forgive others. +# May you share freely, never taking more than you give. +# +#*********************************************************************** +# +# Tests focusing on the built-in fts5 tokenizers. +# + +source [file join [file dirname [info script]] fts5_common.tcl] +set testprefix fts5tokenizer2 + +# If SQLITE_ENABLE_FTS5 is defined, omit this file. +ifcapable !fts5 { + finish_test + return +} + +sqlite3_fts5_create_tokenizer db tst get_tst_tokenizer +proc get_tst_tokenizer {args} { + return "tst_tokenizer" +} +proc tst_tokenizer {flags txt} { + set token "" + set lTok [list] + + foreach c [split $txt {}] { + if {$token==""} { + append token $c + } else { + set t1 [string is upper $token] + set t2 [string is upper $c] + + if {$t1!=$t2} { + lappend lTok $token + set token "" + } + append token $c + } + } + if {$token!=""} { lappend lTok $token } + + set iOff 0 + foreach t $lTok { + set n [string length $t] + sqlite3_fts5_token $t $iOff [expr $iOff+$n] + incr iOff $n + } +} + +do_execsql_test 1.0 { + CREATE VIRTUAL TABLE t1 USING fts5(t, tokenize=tst); +} + +do_execsql_test 1.1 { + INSERT INTO t1 VALUES('AAdontBBmess'); +} + +do_execsql_test 1.2 { + SELECT snippet(t1, 0, '>', '<', '...', 4) FROM t1('BB'); +} {AAdont>BB', '<') FROM t1('BB'); +} {AAdont>BB', '<') FROM t1('AA'); +} {>AA', '<') FROM t1('dont'); +} {AA>dont', '<') FROM t1('mess'); +} {AAdontBB>mess<} + +do_execsql_test 1.7 { + SELECT highlight(t1, 0, '>', '<') FROM t1('BB mess'); +} {AAdont>BBmess<} + + +finish_test