[reactos] 01/01: [RTL][ADVAPI32_APITEST] RtlIsTextUnicode: Don't change weight unless lead-byte flag (#5396)

[Katayama Hirofumi MZ]

https://git.reactos.org/?p=reactos.git;a=commitdiff;h=17e9dcc9224b7773148c01...

commit 17e9dcc9224b7773148c015f6b4d916e4478272b
Author:     Katayama Hirofumi MZ [e-mail]
AuthorDate: Wed Jul 5 12:01:42 2023 +0900
Commit:     GitHub [e-mail]
CommitDate: Wed Jul 5 12:01:42 2023 +0900

[RTL][ADVAPI32_APITEST] RtlIsTextUnicode: Don't change weight unless lead-byte flag (#5396)

- Fix the garbled characters (a.k.a. mojibake) in some Japanese text files.
    - weight will remain 3 unless IS_TEXT_UNICODE_DBCS_LEADBYTE flag is set.
    - Strengthen advapi32_apitest IsTextUnicode testcase.
    CORE-19016
---
 modules/rostests/apitests/advapi32/IsTextUnicode.c | 30 ++++++++++++++++++++--
 sdk/lib/rtl/unicode.c                              |  5 ++--
 2 files changed, 30 insertions(+), 5 deletions(-)

diff --git a/modules/rostests/apitests/advapi32/IsTextUnicode.c b/modules/rostests/apitests/advapi32/IsTextUnicode.c
index 404ab69d07d..d3f9f3dfcb5 100644
--- a/modules/rostests/apitests/advapi32/IsTextUnicode.c
+++ b/modules/rostests/apitests/advapi32/IsTextUnicode.c
@@ -4,6 +4,7 @@
  * PURPOSE:         Tests for (Rtl)IsTextUnicode.
  * PROGRAMMERS:     Hermes Belusca-Maito
  *                  Dmitry Chapyshev
+ *                  Katayama Hirofumi MZ
  */

#include "precomp.h"
@@ -128,6 +129,9 @@ START_TEST(IsTextUnicode)
     };

const char japanese_with_lead[] = "ABC" "\x83\x40" "D";
+    const char japanese_sjis[] = "\x93\xFA\x96\x7B\x8C\xEA\x31\x32\x33\x93\xFA\x96\x7B\x8C\xEA";
+    const char japanese_utf8[] = "\xE6\x97\xA5\xE6\x9C\xAC\xE8\xAA\x9E\x31\x32\x33\xE6\x97\xA5"
+                                 "\xE6\x9C\xAC\xE8\xAA\x9E";
     const char simplfied_chinese_with_lead[] = "ABC" "\xC5\xC5" "D";
     const char korean_with_lead[] = "ABC" "\xBF\xAD" "D";
     const char traditional_chinese_with_lead[] = "ABC" "\xB1\xC1" "D";
@@ -149,8 +153,30 @@ START_TEST(IsTextUnicode)
     SetupLocale(932, 932, -1);

Result = IS_TEXT_UNICODE_DBCS_LEADBYTE;
-    ok(!IsTextUnicode(japanese_with_lead, sizeof(japanese_with_lead), &Result), "IsTextUnicode() returned TRUE, expected FALSE\n");
-    ok(Result == IS_TEXT_UNICODE_DBCS_LEADBYTE, "Result returned 0x%x, expected 0x%x\n", Result, IS_TEXT_UNICODE_DBCS_LEADBYTE);
+    ok_int(IsTextUnicode(japanese_with_lead, sizeof(japanese_with_lead), &Result), FALSE);
+    ok_int(Result, IS_TEXT_UNICODE_DBCS_LEADBYTE);
+
+    ok_int(IsTextUnicode(japanese_sjis, sizeof(japanese_sjis) - 1, NULL), FALSE);
+
+    Result = IS_TEXT_UNICODE_STATISTICS | IS_TEXT_UNICODE_REVERSE_STATISTICS;
+    ok_int(IsTextUnicode(japanese_sjis, sizeof(japanese_sjis) - 1, &Result), FALSE);
+    ok_int(Result, 0);
+
+    Result = IS_TEXT_UNICODE_STATISTICS | IS_TEXT_UNICODE_REVERSE_STATISTICS |
+             IS_TEXT_UNICODE_DBCS_LEADBYTE;
+    ok_int(IsTextUnicode(japanese_sjis, sizeof(japanese_sjis) - 1, &Result), FALSE);
+    ok_int(Result, (IS_TEXT_UNICODE_DBCS_LEADBYTE | IS_TEXT_UNICODE_REVERSE_STATISTICS));
+
+    ok_int(IsTextUnicode(japanese_utf8, sizeof(japanese_utf8) - 1, NULL), FALSE);
+
+    Result = IS_TEXT_UNICODE_STATISTICS | IS_TEXT_UNICODE_REVERSE_STATISTICS;
+    ok_int(IsTextUnicode(japanese_utf8, sizeof(japanese_utf8) - 1, &Result), FALSE);
+    ok_int(Result, 0);
+
+    Result = IS_TEXT_UNICODE_STATISTICS | IS_TEXT_UNICODE_REVERSE_STATISTICS |
+             IS_TEXT_UNICODE_DBCS_LEADBYTE;
+    ok_int(IsTextUnicode(japanese_utf8, sizeof(japanese_utf8) - 1, &Result), FALSE);
+    ok_int(Result, (IS_TEXT_UNICODE_DBCS_LEADBYTE | IS_TEXT_UNICODE_STATISTICS));

/* Simplified Chinese */
     SetupLocale(936, 936, -1);
diff --git a/sdk/lib/rtl/unicode.c b/sdk/lib/rtl/unicode.c
index 8c7248db155..c29dfa4f3b8 100644
--- a/sdk/lib/rtl/unicode.c
+++ b/sdk/lib/rtl/unicode.c
@@ -1356,7 +1356,7 @@ RtlIsTextUnicode(CONST VOID* buf, INT len, INT* pf)
         }
     }

-    if (NlsMbCodePageTag)
+    if (NlsMbCodePageTag && pf && (*pf & IS_TEXT_UNICODE_DBCS_LEADBYTE))
     {
         for (i = 0; i < len; i++)
         {
@@ -1378,8 +1378,7 @@ RtlIsTextUnicode(CONST VOID* buf, INT len, INT* pf)
             else
                 weight = 1;

-            if (pf && (*pf & IS_TEXT_UNICODE_DBCS_LEADBYTE))
-                out_flags |= IS_TEXT_UNICODE_DBCS_LEADBYTE;
+            out_flags |= IS_TEXT_UNICODE_DBCS_LEADBYTE;
         }
     }