xref: /reactos/sdk/lib/inflib/infhostrtl.c (revision c2c66aff)
1 /*
2  * PROJECT:    .inf file parser
3  * LICENSE:    GPL - See COPYING in the top level directory
4  * PROGRAMMER: Royce Mitchell III
5  *             Eric Kohl
6  *             Ge van Geldorp <gvg@reactos.org>
7  */
8 
9 /* INCLUDES *****************************************************************/
10 
11 #include "inflib.h"
12 #include "infhost.h"
13 
14 #define NDEBUG
15 #include <debug.h>
16 
17 NTSTATUS NTAPI
RtlMultiByteToUnicodeN(IN PWCHAR UnicodeString,IN ULONG UnicodeSize,IN PULONG ResultSize,IN PCSTR MbString,IN ULONG MbSize)18 RtlMultiByteToUnicodeN(
19    IN PWCHAR UnicodeString,
20    IN ULONG UnicodeSize,
21    IN PULONG ResultSize,
22    IN PCSTR MbString,
23    IN ULONG MbSize)
24 {
25     ULONG Size = 0;
26     ULONG i;
27     PUCHAR WideString;
28 
29     /* single-byte code page */
30     if (MbSize > (UnicodeSize / sizeof(WCHAR)))
31         Size = UnicodeSize / sizeof(WCHAR);
32     else
33         Size = MbSize;
34 
35     if (ResultSize != NULL)
36         *ResultSize = Size * sizeof(WCHAR);
37 
38     WideString = (PUCHAR)UnicodeString;
39     for (i = 0; i < Size; i++)
40     {
41         WideString[2 * i + 0] = (UCHAR)MbString[i];
42         WideString[2 * i + 1] = 0;
43     }
44 
45     return STATUS_SUCCESS;
46 }
47 
48 
49 BOOLEAN
50 NTAPI
RtlIsTextUnicode(PVOID buf,INT len,INT * pf)51 RtlIsTextUnicode( PVOID buf, INT len, INT *pf )
52 {
53     static const WCHAR std_control_chars[] = {'\r','\n','\t',' ',0x3000,0};
54     static const WCHAR byterev_control_chars[] = {0x0d00,0x0a00,0x0900,0x2000,0};
55     const WCHAR *s = buf;
56     int i;
57     unsigned int flags = MAXULONG, out_flags = 0;
58 
59     if (len < sizeof(WCHAR))
60     {
61         /* FIXME: MSDN documents IS_TEXT_UNICODE_BUFFER_TOO_SMALL but there is no such thing... */
62         if (pf) *pf = 0;
63         return FALSE;
64     }
65     if (pf)
66         flags = (unsigned int)*pf;
67     /*
68      * Apply various tests to the text string. According to the
69      * docs, each test "passed" sets the corresponding flag in
70      * the output flags. But some of the tests are mutually
71      * exclusive, so I don't see how you could pass all tests ...
72      */
73 
74     /* Check for an odd length ... pass if even. */
75     if (len & 1) out_flags |= IS_TEXT_UNICODE_ODD_LENGTH;
76 
77     if (((char *)buf)[len - 1] == 0)
78         len--;  /* Windows seems to do something like that to avoid e.g. false IS_TEXT_UNICODE_NULL_BYTES  */
79 
80     len /= (INT)sizeof(WCHAR);
81     /* Windows only checks the first 256 characters */
82     if (len > 256) len = 256;
83 
84     /* Check for the special byte order unicode marks. */
85     if (*s == 0xFEFF) out_flags |= IS_TEXT_UNICODE_SIGNATURE;
86     if (*s == 0xFFFE) out_flags |= IS_TEXT_UNICODE_REVERSE_SIGNATURE;
87 
88     /* apply some statistical analysis */
89     if (flags & IS_TEXT_UNICODE_STATISTICS)
90     {
91         int stats = 0;
92         /* FIXME: checks only for ASCII characters in the unicode stream */
93         for (i = 0; i < len; i++)
94         {
95             if (s[i] <= 255) stats++;
96         }
97         if (stats > len / 2)
98             out_flags |= IS_TEXT_UNICODE_STATISTICS;
99     }
100 
101     /* Check for unicode NULL chars */
102     if (flags & IS_TEXT_UNICODE_NULL_BYTES)
103     {
104         for (i = 0; i < len; i++)
105         {
106             if (!(s[i] & 0xff) || !(s[i] >> 8))
107             {
108                 out_flags |= IS_TEXT_UNICODE_NULL_BYTES;
109                 break;
110             }
111         }
112     }
113 
114     if (flags & IS_TEXT_UNICODE_CONTROLS)
115     {
116         for (i = 0; i < len; i++)
117         {
118             if (strchrW(std_control_chars, s[i]))
119             {
120                 out_flags |= IS_TEXT_UNICODE_CONTROLS;
121                 break;
122             }
123         }
124     }
125 
126     if (flags & IS_TEXT_UNICODE_REVERSE_CONTROLS)
127     {
128         for (i = 0; i < len; i++)
129         {
130             if (strchrW(byterev_control_chars, s[i]))
131             {
132                 out_flags |= IS_TEXT_UNICODE_REVERSE_CONTROLS;
133                 break;
134             }
135         }
136     }
137 
138     if (pf)
139     {
140         out_flags &= (unsigned int)*pf;
141         *pf = (INT)out_flags;
142     }
143     /* check for flags that indicate it's definitely not valid Unicode */
144     if (out_flags & (IS_TEXT_UNICODE_REVERSE_MASK | IS_TEXT_UNICODE_NOT_UNICODE_MASK)) return FALSE;
145     /* now check for invalid ASCII, and assume Unicode if so */
146     if (out_flags & IS_TEXT_UNICODE_NOT_ASCII_MASK) return TRUE;
147     /* now check for Unicode flags */
148     if (out_flags & IS_TEXT_UNICODE_UNICODE_MASK) return TRUE;
149     /* no flags set */
150     return FALSE;
151 }
152