PostgreSQL Source Code  git master
All Data Structures Namespaces Files Functions Variables Typedefs Enumerations Enumerator Macros Pages
ascii.h
Go to the documentation of this file.
1 /*-----------------------------------------------------------------------
2  * ascii.h
3  *
4  * Portions Copyright (c) 1999-2024, PostgreSQL Global Development Group
5  *
6  * src/include/utils/ascii.h
7  *
8  *-----------------------------------------------------------------------
9  */
10 
11 #ifndef _ASCII_H_
12 #define _ASCII_H_
13 
14 #include "port/simd.h"
15 
16 extern void ascii_safe_strlcpy(char *dest, const char *src, size_t destsiz);
17 
18 /*
19  * Verify a chunk of bytes for valid ASCII.
20  *
21  * Returns false if the input contains any zero bytes or bytes with the
22  * high-bit set. Input len must be a multiple of the chunk size (8 or 16).
23  */
24 static inline bool
25 is_valid_ascii(const unsigned char *s, int len)
26 {
27  const unsigned char *const s_end = s + len;
28  Vector8 chunk;
29  Vector8 highbit_cum = vector8_broadcast(0);
30 #ifdef USE_NO_SIMD
31  Vector8 zero_cum = vector8_broadcast(0x80);
32 #endif
33 
34  Assert(len % sizeof(chunk) == 0);
35 
36  while (s < s_end)
37  {
38  vector8_load(&chunk, s);
39 
40  /* Capture any zero bytes in this chunk. */
41 #ifdef USE_NO_SIMD
42 
43  /*
44  * First, add 0x7f to each byte. This sets the high bit in each byte,
45  * unless it was a zero. If any resulting high bits are zero, the
46  * corresponding high bits in the zero accumulator will be cleared.
47  *
48  * If none of the bytes in the chunk had the high bit set, the max
49  * value each byte can have after the addition is 0x7f + 0x7f = 0xfe,
50  * and we don't need to worry about carrying over to the next byte. If
51  * any input bytes did have the high bit set, it doesn't matter
52  * because we check for those separately.
53  */
54  zero_cum &= (chunk + vector8_broadcast(0x7F));
55 #else
56 
57  /*
58  * Set all bits in each lane of the highbit accumulator where input
59  * bytes are zero.
60  */
61  highbit_cum = vector8_or(highbit_cum,
62  vector8_eq(chunk, vector8_broadcast(0)));
63 #endif
64 
65  /* Capture all set bits in this chunk. */
66  highbit_cum = vector8_or(highbit_cum, chunk);
67 
68  s += sizeof(chunk);
69  }
70 
71  /* Check if any high bits in the high bit accumulator got set. */
72  if (vector8_is_highbit_set(highbit_cum))
73  return false;
74 
75 #ifdef USE_NO_SIMD
76  /* Check if any high bits in the zero accumulator got cleared. */
77  if (zero_cum != vector8_broadcast(0x80))
78  return false;
79 #endif
80 
81  return true;
82 }
83 
84 #endif /* _ASCII_H_ */
void ascii_safe_strlcpy(char *dest, const char *src, size_t destsiz)
Definition: ascii.c:174
static bool is_valid_ascii(const unsigned char *s, int len)
Definition: ascii.h:25
#define Assert(condition)
Definition: c.h:837
uint64 chunk
const void size_t len
static Vector8 vector8_broadcast(const uint8 c)
Definition: simd.h:135
static void vector8_load(Vector8 *v, const uint8 *s)
Definition: simd.h:108
static Vector8 vector8_or(const Vector8 v1, const Vector8 v2)
Definition: simd.h:338
uint64 Vector8
Definition: simd.h:60
static bool vector8_is_highbit_set(const Vector8 v)
Definition: simd.h:271