[PATCH] fs: make nls_cp936.c handle some U00XY characters and U20AC correctly

Twenty characters in cp936 are not correctly handled.  They're all in the
U00 plane.  nls_cp936 converts all U00XY to XY but this is not correct for
some characters.(e.g.  U00B7 -> A1A4, U00A8 -> A1A7).

This problem is fixed by generating u2c_00 based on all c2u_xx and changing
uni2char() to give U00 plane a special handling.  The "€"(U20AC,80 in
cp936) is also be handled properly.

Acked-by: Gang Chen <cgdlut@gmail.com>
Cc: OGAWA Hirofumi <hirofumi@mail.parknet.co.jp>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>

1 files changed, 103 insertions, 10 deletions

diff --git a/fs/nls/nls_cp936.c b/fs/nls/nls_cp936.c
index 046fde8170ea..65e640c61c8b 100644
--- a/fs/nls/nls_cp936.c
+++ b/fs/nls/nls_cp936.c
@@ -4421,6 +4421,73 @@ static wchar_t *page_charset2uni[256] = {
         c2u_F8, c2u_F9, c2u_FA, c2u_FB, c2u_FC, c2u_FD, c2u_FE, NULL,   
 };
 
+static unsigned char u2c_00[512] = {
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x00-0x03 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x04-0x07 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x08-0x0B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x0C-0x0F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x10-0x13 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x14-0x17 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x18-0x1B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x1C-0x1F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x20-0x23 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x24-0x27 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x28-0x2B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x2C-0x2F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x30-0x33 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x34-0x37 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x38-0x3B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x3C-0x3F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x40-0x43 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x44-0x47 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x48-0x4B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x4C-0x4F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x50-0x53 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x54-0x57 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x58-0x5B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x5C-0x5F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x60-0x63 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x64-0x67 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x68-0x6B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x6C-0x6F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x70-0x73 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x74-0x77 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x78-0x7B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x7C-0x7F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x80-0x83 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x84-0x87 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x88-0x8B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x8C-0x8F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x90-0x93 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x94-0x97 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x98-0x9B */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x9C-0x9F */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xA0-0xA3 */
+        0xA1, 0xE8, 0x00, 0x00, 0x00, 0x00, 0xA1, 0xEC, /* 0xA4-0xA7 */
+        0xA1, 0xA7, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xA8-0xAB */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xAC-0xAF */
+        0xA1, 0xE3, 0xA1, 0xC0, 0x00, 0x00, 0x00, 0x00, /* 0xB0-0xB3 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xA1, 0xA4, /* 0xB4-0xB7 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xB8-0xBB */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xBC-0xBF */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xC0-0xC3 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xC4-0xC7 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xC8-0xCB */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xCC-0xCF */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xD0-0xD3 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xA1, 0xC1, /* 0xD4-0xD7 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xD8-0xDB */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xDC-0xDF */
+        0xA8, 0xA4, 0xA8, 0xA2, 0x00, 0x00, 0x00, 0x00, /* 0xE0-0xE3 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xE4-0xE7 */
+        0xA8, 0xA8, 0xA8, 0xA6, 0xA8, 0xBA, 0x00, 0x00, /* 0xE8-0xEB */
+        0xA8, 0xAC, 0xA8, 0xAA, 0x00, 0x00, 0x00, 0x00, /* 0xEC-0xEF */
+        0x00, 0x00, 0x00, 0x00, 0xA8, 0xB0, 0xA8, 0xAE, /* 0xF0-0xF3 */
+        0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0xA1, 0xC2, /* 0xF4-0xF7 */
+        0x00, 0x00, 0xA8, 0xB4, 0xA8, 0xB2, 0x00, 0x00, /* 0xF8-0xFB */
+        0xA8, 0xB9, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0xFC-0xFF */
+};
+
 static unsigned char u2c_01[512] = {
         0xA8, 0xA1, 0xA8, 0xA1, 0x00, 0x00, 0x00, 0x00, /* 0x00-0x03 */
         0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, /* 0x04-0x07 */
@@ -10825,7 +10892,7 @@ static unsigned char u2c_FF[512] = {
 };
 
 static unsigned char *page_uni2charset[256] = {
-        NULL,   u2c_01, u2c_02, u2c_03, u2c_04, NULL,   NULL,   NULL,   
+        u2c_00, u2c_01, u2c_02, u2c_03, u2c_04, NULL,   NULL,   NULL,
         NULL,   NULL,   NULL,   NULL,   NULL,   NULL,   NULL,   NULL,   
         NULL,   NULL,   NULL,   NULL,   NULL,   NULL,   NULL,   NULL,   
         NULL,   NULL,   NULL,   NULL,   NULL,   NULL,   NULL,   NULL,   
@@ -10936,11 +11003,34 @@ static int uni2char(const wchar_t uni,
         unsigned char *uni2charset;
         unsigned char cl = uni&0xFF;
         unsigned char ch = (uni>>8)&0xFF;
-        int n;
+        unsigned char out0,out1;
 
         if (boundlen <= 0)
                 return -ENAMETOOLONG;
 
+        if (uni == 0x20ac) {/* Euro symbol.The only exception with a non-ascii unicode */
+                out[0] = 0x80;
+                return 1;
+        }
+
+        if (ch == 0) { /* handle the U00 plane*/
+                /* if (cl == 0) return -EINVAL;*/ /*U0000 is legal in cp936*/
+                out0 = u2c_00[cl*2];
+                out1 = u2c_00[cl*2+1];
+                if (out0 == 0x00 && out1 == 0x00) {
+                        if (cl<0x80) {
+                                out[0] = cl;
+                                return 1;
+                        }
+                        return -EINVAL;
+                } else {
+                        if (boundlen <= 1)
+                                return -ENAMETOOLONG;
+                        out[0] = out0;
+                        out[1] = out1;
+                        return 2;
+                }
+        }
 
         uni2charset = page_uni2charset[ch];
         if (uni2charset) {
@@ -10950,15 +11040,10 @@ static int uni2char(const wchar_t uni,
                 out[1] = uni2charset[cl*2+1];
                 if (out[0] == 0x00 && out[1] == 0x00)
                         return -EINVAL;
-                n = 2;
-        } else if (ch==0 && cl) {
-                out[0] = cl;
-                n = 1;
+                return 2;
         }
         else
                 return -EINVAL;
-
-        return n;
 }
 
 static int char2uni(const unsigned char *rawstring, int boundlen,
@@ -10972,7 +11057,11 @@ static int char2uni(const unsigned char *rawstring, int boundlen,
                 return -ENAMETOOLONG;
 
         if (boundlen == 1) {
-                *uni = rawstring[0];
+                if (rawstring[0]==0x80) { /* Euro symbol.The only exception with a non-ascii unicode */
+                        *uni = 0x20ac;
+                } else {
+                        *uni = rawstring[0];
+                }
                 return 1;
         }
 
@@ -10986,7 +11075,11 @@ static int char2uni(const unsigned char *rawstring, int boundlen,
                         return -EINVAL;
                 n = 2;
         } else{
-                *uni = ch;
+                if (ch==0x80) {/* Euro symbol.The only exception with a non-ascii unicode */
+                        *uni = 0x20ac;
+                } else {
+                        *uni = ch;
+                }
                 n = 1;
         }
         return n;