Fix the spellfix1_scriptcode() function to ignore whitespace and punctuation,

and to recognize hebrew and arabic scripts.

FossilOrigin-Name: 7adfa4a5794e47f97491c08abeaaac90e826b331

diff --git a/ext/misc/spellfix.c b/ext/misc/spellfix.c
index 0515b7bc6c9e781ce07c6e1f5975a12baa27c34e..c0da6b339bfcdc6df1fcd4b87b655602bc172595 100644 (file)
--- a/ext/misc/spellfix.c
+++ b/ext/misc/spellfix.c
@@ -1717,17 +1717,23 @@ static void scriptCodeSqlFunc(
 # define SCRIPT_LATIN       0x0001
 # define SCRIPT_CYRILLIC    0x0002
 # define SCRIPT_GREEK       0x0004
+# define SCRIPT_HEBREW      0x0008
+# define SCRIPT_ARABIC      0x0010
 
   while( nIn>0 ){
     c = utf8Read(zIn, nIn, &sz);
     zIn += sz;
     nIn -= sz;
-    if( c<0x02af ){
+    if( c<0x02af && (c>=0x80 || midClass[c&0x7f]<CCLASS_DIGIT) ){
       scriptMask |= SCRIPT_LATIN;
     }else if( c>=0x0400 && c<=0x04ff ){
       scriptMask |= SCRIPT_CYRILLIC;
     }else if( c>=0x0386 && c<=0x03ce ){
       scriptMask |= SCRIPT_GREEK;
+    }else if( c>=0x0590 && c<=0x05ff ){
+      scriptMask |= SCRIPT_HEBREW;
+    }else if( c>=0x0600 && c<=0x06ff ){
+      scriptMask |= SCRIPT_ARABIC;
     }
   }
   switch( scriptMask ){
@@ -1735,6 +1741,8 @@ static void scriptCodeSqlFunc(
     case SCRIPT_LATIN:     res = 215; break;
     case SCRIPT_CYRILLIC:  res = 220; break;
     case SCRIPT_GREEK:     res = 200; break;
+    case SCRIPT_HEBREW:    res = 125; break;
+    case SCRIPT_ARABIC:    res = 160; break;
     default:               res = 998; break;
   }
   sqlite3_result_int(context, res);

diff --git a/manifest b/manifest
index df0b61a31c2987682f73c7e1d94de4d3cc1c54cc..799ac619df5dcf1f68132335d9ad3f954646c7af 100644 (file)
--- a/manifest
+++ b/manifest
@@ -1,5 +1,5 @@
-C Fixes\sfor\sharmless\scompiler\swarnings.
-D 2015-12-17T13:28:50.207
+C Fix\sthe\sspellfix1_scriptcode()\sfunction\sto\signore\swhitespace\sand\spunctuation,\nand\sto\srecognize\shebrew\sand\sarabic\sscripts.
+D 2015-12-17T14:18:21.904
 F Makefile.in 28bcd6149e050dff35d4dcfd97e890cd387a499d
 F Makefile.linux-gcc 7bc79876b875010e8c8f9502eb935ca92aa3c434
 F Makefile.msc 5fff077fcc46de7714ed6eebb6159a4c00eab751
@@ -198,7 +198,7 @@ F ext/misc/regexp.c af92cdaa5058fcec1451e49becc7ba44dba023dc
 F ext/misc/rot13.c 1ac6f95f99b575907b9b09c81a349114cf9be45a
 F ext/misc/series.c b8fb7befd85b3a9b4a10e701b30b2b79ca92b6d4
 F ext/misc/showauth.c 732578f0fe4ce42d577e1c86dc89dd14a006ab52
-F ext/misc/spellfix.c b9065af7ab1f2597b505a8aa9892620866d502fc
+F ext/misc/spellfix.c df6efb90eb668d1860c9b59e4320e985e46dffa7
 F ext/misc/totype.c 4a167594e791abeed95e0a8db028822b5e8fe512
 F ext/misc/vfslog.c fe40fab5c077a40477f7e5eba994309ecac6cc95
 F ext/misc/vtshim.c babb0dc2bf116029e3e7c9a618b8a1377045303e
@@ -1035,6 +1035,7 @@ F test/speed4p.test 0e51908951677de5a969b723e03a27a1c45db38b
 F test/speedtest1.c f8bf04214e7b5f745feea99f7bde68b1c4870666
 F test/spellfix.test 0597065ff57042df1f138e6a2611ae19c2698135
 F test/spellfix2.test dfc8f519a3fc204cb2dfa8b4f29821ae90f6f8c3
+F test/spellfix3.test f7bf7b3482971473d32b6b00f6944c5c066cff97
 F test/sqldiff1.test 8f6bc7c6a5b3585d350d779c6078869ba402f8f5
 F test/sqllimits1.test a74ee2a3740b9f9c2437c246d8fb77354862a142
 F test/sqllog.test a8faa2df39610a037dd372ed872d124260d32953
@@ -1404,8 +1405,7 @@ F tool/vdbe_profile.tcl 246d0da094856d72d2c12efec03250d71639d19f
 F tool/warnings-clang.sh f6aa929dc20ef1f856af04a730772f59283631d4
 F tool/warnings.sh 48bd54594752d5be3337f12c72f28d2080cb630b
 F tool/win/sqlite.vsix deb315d026cc8400325c5863eef847784a219a2f
-P f2fc3aede55c8fa0351697f9f5ce4e544334c9af 1d0e6aa119da8e15d35508f5d75ffc729979da92
-R d6064250d2ceb34ab73cc0f26f2d0c34
-T +closed 1d0e6aa119da8e15d35508f5d75ffc729979da92
+P 85ebd46c701e0a094a4690cd8f1d0cbae9aa257c
+R 70cdc580d7d9b92b032d33f5e0d50f2e
 U drh
-Z 55d4a7733cfc9336492f45b410da6806
+Z 6d4b5ed9cd1870281d7d851922f323d8

diff --git a/manifest.uuid b/manifest.uuid
index d3a2a6c2c5a57b22cd406acd1c053c9a30634059..eebc618153485980aafd38a08d432ec8139d625e 100644 (file)
--- a/manifest.uuid
+++ b/manifest.uuid
@@ -1 +1 @@
-85ebd46c701e0a094a4690cd8f1d0cbae9aa257c
\ No newline at end of file
+7adfa4a5794e47f97491c08abeaaac90e826b331
\ No newline at end of file

diff --git a/test/spellfix3.test b/test/spellfix3.test
new file mode 100644 (file)
index 0000000..ce002ed
--- /dev/null
+++ b/test/spellfix3.test
@@ -0,0 +1,43 @@
+# 2015-12-17
+#
+# 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.
+#
+#***********************************************************************
+#
+
+set testdir [file dirname $argv0]
+source $testdir/tester.tcl
+set testprefix spellfix3
+
+ifcapable !vtab { finish_test ; return }
+
+load_static_extension db spellfix
+
+do_execsql_test 100 {
+  SELECT spellfix1_scriptcode('And God said, “Let there be light”');
+} {215}
+do_execsql_test 110 {
+  SELECT spellfix1_scriptcode('Бог сказал: "Да будет свет"');
+} {220}
+do_execsql_test 120 {
+  SELECT spellfix1_scriptcode('και ειπεν ο θεος γενηθητω φως και εγενετο φως');
+} {200}
+do_execsql_test 130 {
+  SELECT spellfix1_scriptcode('וַיֹּ֥אמֶר אֱלֹהִ֖ים יְהִ֣י א֑וֹר וַֽיְהִי־אֽוֹר׃');
+} {125}
+do_execsql_test 140 {
+  SELECT spellfix1_scriptcode('فِي ذَلِكَ الوَقتِ، قالَ اللهُ: لِيَكُنْ نُورٌ. فَصَارَ نُورٌ.');
+} {160}
+do_execsql_test 200 {
+  SELECT spellfix1_scriptcode('+3.14159');
+} {999}
+do_execsql_test 210 {
+  SELECT spellfix1_scriptcode('And God said: "Да будет свет"');
+} {998}
+
+finish_test