paweld/lazarus
1
0
Fork 0
mirror of https://gitlab.com/freepascal.org/lazarus/lazarus.git synced 2025-04-23 01:39:31 +02:00
Code Issues Packages Projects Releases Wiki Activity

Adds initial support in utf8lowercase for archaic and non-slavic cyrillic

Browse Source 
git-svn-id: trunk@32837 -
This commit is contained in:
sekelsenmat 2011-10-12 06:57:15 +00:00
parent dd582c9852
commit 21f584b765
3 changed files with 37 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
components/lazutils/lazutf8.pas
View File

@ -1205,7 +1205,7 @@ begin
begin
c := InStr^;
case c of
'A'..'Z',#$C3, #$C4, #$C5..#$C8, #$CE, #$D0: Break;
'A'..'Z',#$C3, #$C4, #$C5..#$C8, #$CE, #$D0..#$D2: Break;
// already lower, or otherwhise not affected
else
inc(InStr);
@ -1396,6 +1396,39 @@ begin
end;
end;
end;
// Archaic and non-slavic cyrillic 460-47F = D1A0-D1BF
// These require just adding 1 to get the lowercase
#$D1:
begin
c := OutStr[1];
inc(InStr);
inc(OutStr, 2);
case c of
#$A0..#$BF:
begin
if ord(c) mod 2 = 0 then
OutStr[-1] := chr(ord(c) + 1);
end;
end;
end;
// Archaic and non-slavic cyrillic 480-4BF = D280-D2BF
// These mostly require just adding 1 to get the lowercase
#$D2:
begin
c := OutStr[1];
inc(InStr);
inc(OutStr, 2);
case c of
#$80: OutStr[-1] := chr(ord(c) + 1);
// #$81 is already lowercase
// #$82-#$89 ???
#$8A..#$BF:
begin
if ord(c) mod 2 = 0 then
OutStr[-1] := chr(ord(c) + 1);
end;
end;
end;
else
begin
inc(InStr);

2
test/lazutils/testunicode.lpi
View File

@ -57,7 +57,7 @@
<Linking>
<Debugging>
<GenerateDebugInfo Value="True"/>
<DebugInfoType Value="dsAuto"/>
<DebugInfoType Value="dsDwarf2"/>
</Debugging>
</Linking>
<Other>

4
test/lazutils/testunicode.pas
View File

@ -113,8 +113,8 @@ begin
// Cyrillic
AssertStringOperationUTF8LowerCase('Russian UTF8LowerCase 1', '', 'АБВЕЁЖЗКЛМНОПРДЙГ СУФХЦЧШЩЪЫЬЭЮЯИТ', 'абвеёжзклмнопрдйг суфхцчшщъыьэюяит');
AssertStringOperationUTF8LowerCase('Russian UTF8LowerCase 2', '', 'абвеёжзклмнопрдйг суфхцчшщъыьэюяит', 'абвеёжзклмнопрдйг суфхцчшщъыьэюяит');
//AssertStringOperationUTF8LowerCase('Cyrillic UTF8UpperCase 1', '', 'Ѡѡ Ѣѣ Ѥѥ Ѧѧ Ѩѩ Ѫѫ Ѭѭ Ѯѯ Ѱѱ Ѳѳ Ѵѵ Ѷѷ Ѹѹ Ѻѻ Ѽѽ Ѿѿ Ҁҁ ', 'ѡѡ ѣѣ ѥѥ ѧѧ ѩѩ ѫѫ ѭѭ ѯѯ ѱѱ ѳѳ ѵѵ ѷѷ ѹѹ ѻѻ ѽѽ ѿѿ ҁҁ');
//AssertStringOperationUTF8LowerCase('Cyrillic UTF8UpperCase 2', '', 'Ҋҋ Ҍҍ Ҏҏ Ґґ Ғғ Ҕҕ Җҗ Ҙҙ Ққ Ҝҝ Ҟҟ Ҡҡ Ңң Ҥҥ Ҧҧ Ҩҩ Ҫҫ Ҭҭ Үү Ұұ Ҳҳ Ҵҵ Ҷҷ Ҹҹ Һһ Ҽҽ Ҿҿ', 'ҋҋ ҍҍ ҏҏ ґґ ғғ ҕҕ җҗ ҙҙ ққ ҝҝ ҟҟ ҡҡ ңң ҥҥ ҧҧ ҩҩ ҫҫ ҭҭ үү ұұ ҳҳ ҵҵ ҷҷ ҹҹ һһ ҽҽ ҿҿ');
AssertStringOperationUTF8LowerCase('Cyrillic UTF8UpperCase 1', '', 'Ѡѡ Ѣѣ Ѥѥ Ѧѧ Ѩѩ Ѫѫ Ѭѭ Ѯѯ Ѱѱ Ѳѳ Ѵѵ Ѷѷ Ѹѹ Ѻѻ Ѽѽ Ѿѿ Ҁҁ', 'ѡѡ ѣѣ ѥѥ ѧѧ ѩѩ ѫѫ ѭѭ ѯѯ ѱѱ ѳѳ ѵѵ ѷѷ ѹѹ ѻѻ ѽѽ ѿѿ ҁҁ');
AssertStringOperationUTF8LowerCase('Cyrillic UTF8UpperCase 2', '', 'Ҋҋ Ҍҍ Ҏҏ Ґґ Ғғ Ҕҕ Җҗ Ҙҙ Ққ Ҝҝ Ҟҟ Ҡҡ Ңң Ҥҥ Ҧҧ Ҩҩ Ҫҫ Ҭҭ Үү Ұұ Ҳҳ Ҵҵ Ҷҷ Ҹҹ Һһ Ҽҽ Ҿҿ', 'ҋҋ ҍҍ ҏҏ ґґ ғғ ҕҕ җҗ ҙҙ ққ ҝҝ ҟҟ ҡҡ ңң ҥҥ ҧҧ ҩҩ ҫҫ ҭҭ үү ұұ ҳҳ ҵҵ ҷҷ ҹҹ һһ ҽҽ ҿҿ');
// What shouldnt change
AssertStringOperationUTF8LowerCase('Chinese UTF8LowerCase 1', '', '名字叫嘉英，嘉陵江的嘉，英國的英', '名字叫嘉英，嘉陵江的嘉，英國的英');