/src/jq/src/jv_unicode.c

Source
#include <stdio.h>
#include <assert.h>
#include "jv_unicode.h"
#include "jv_utf8_tables.h"

// jvp_utf8_backtrack returns the beginning of the last codepoint in the
// string, assuming that start is the last byte in the string.
// If the last codepoint is incomplete, returns the number of missing bytes via
// *missing_bytes.  If there are no leading bytes or an invalid byte is
// encountered, NULL is returned and *missing_bytes is not altered.
const char* jvp_utf8_backtrack(const char* start, const char* min, int *missing_bytes) {
  assert(min <= start);
  if (min == start) {
    return min;
  }
  int length = 0;
  int seen = 1;
  while ((length = utf8_coding_length[(unsigned char)*start]) == UTF8_CONTINUATION_BYTE) {
    if (start == min) break;
    start--;
    seen++;
  }
  if (length == 0 || length == UTF8_CONTINUATION_BYTE || length - seen < 0) {
    return NULL;
  }
  if (missing_bytes) *missing_bytes = length - seen;
  return start;
}

const char* jvp_utf8_next(const char* in, const char* end, int* codepoint_ret) {
  assert(in <= end);
  if (in == end) {
    return 0;
  }
  int codepoint = -1;
  unsigned char first = (unsigned char)in[0];
  int length = utf8_coding_length[first];
  if ((first & 0x80) == 0) {
    /* Fast-path for ASCII */
    codepoint = first;
    length = 1;
  } else if (length == 0 || length == UTF8_CONTINUATION_BYTE) {
    /* Bad single byte - either an invalid byte or an out-of-place continuation byte */
    length = 1;
  } else if (length > end - in) {
    /* String ends before UTF8 sequence ends */
    length = end - in;
  } else {
    codepoint = ((unsigned)in[0]) & utf8_coding_bits[first];
    for (int i=1; i<length; i++) {
      unsigned ch = (unsigned char)in[i];
      if (utf8_coding_length[ch] != UTF8_CONTINUATION_BYTE){
        /* Invalid UTF8 sequence - not followed by the right number of continuation bytes */
        codepoint = -1;
        length = i;
        break;
      }
      codepoint = (codepoint << 6) | (ch & 0x3f);
    }
    if (codepoint < utf8_first_codepoint[length]) {
      /* Overlong UTF8 sequence */
      codepoint = -1;
    }
    if (0xD800 <= codepoint && codepoint <= 0xDFFF) {
      /* Surrogate codepoints can't be encoded in UTF8 */
      codepoint = -1;
    }
    if (codepoint > 0x10FFFF) {
      /* Outside Unicode range */
      codepoint = -1;
    }
  }
  assert(length > 0);
  *codepoint_ret = codepoint;
  return in + length;
}

int jvp_utf8_is_valid(const char* in, const char* end) {
  int codepoint;
  while ((in = jvp_utf8_next(in, end, &codepoint))) {
    if (codepoint == -1) return 0;
  }
  return 1;
}

/* Assumes startchar is the first byte of a valid character sequence */
int jvp_utf8_decode_length(char startchar) {
  if ((startchar & 0x80) == 0) return 1;         // 0___ ____
  else if ((startchar & 0xE0) == 0xC0) return 2; // 110_ ____
  else if ((startchar & 0xF0) == 0xE0) return 3; // 1110 ____
  else return 4;                                 // 1111 ____
}

int jvp_utf8_encode_length(int codepoint) {
  if (codepoint <= 0x7F) return 1;
  else if (codepoint <= 0x7FF) return 2;
  else if (codepoint <= 0xFFFF) return 3;
  else return 4;
}

int jvp_utf8_encode(int codepoint, char* out) {
  assert(codepoint >= 0 && codepoint <= 0x10FFFF);
  char* start = out;
  if (codepoint <= 0x7F) {
    *out++ = codepoint;
  } else if (codepoint <= 0x7FF) {
    *out++ = 0xC0 + ((codepoint & 0x7C0) >> 6);
    *out++ = 0x80 + ((codepoint & 0x03F));
  } else if(codepoint <= 0xFFFF) {
    *out++ = 0xE0 + ((codepoint & 0xF000) >> 12);
    *out++ = 0x80 + ((codepoint & 0x0FC0) >> 6);
    *out++ = 0x80 + ((codepoint & 0x003F));
  } else {
    *out++ = 0xF0 + ((codepoint & 0x1C0000) >> 18);
    *out++ = 0x80 + ((codepoint & 0x03F000) >> 12);
    *out++ = 0x80 + ((codepoint & 0x000FC0) >> 6);
    *out++ = 0x80 + ((codepoint & 0x00003F));
  }
  assert(out - start == jvp_utf8_encode_length(codepoint));
  return out - start;
}

// characters with White_Space property in:
// https://www.unicode.org/Public/UCD/latest/ucd/PropList.txt
int jvp_codepoint_is_whitespace(int c) {
  return
    (c >= 0x0009 && c <= 0x000D) || // <control-0009>..<control-000D>
    c == 0x0020                  || // SPACE
    c == 0x0085                  || // <control-0085>
    c == 0x00A0                  || // NO-BREAK SPACE
    c == 0x1680                  || // OGHAM SPACE MARK
    (c >= 0x2000 && c <= 0x200A) || // EN QUAD..HAIR SPACE
    c == 0x2028                  || // LINE SEPARATOR
    c == 0x2029                  || // PARAGRAPH SEPARATOR
    c == 0x202F                  || // NARROW NO-BREAK SPACE
    c == 0x205F                  || // MEDIUM MATHEMATICAL SPACE
    c == 0x3000                     // IDEOGRAPHIC SPACE
    ;
}

Line	Count	Source
1		#include <stdio.h>
2		#include <assert.h>
3		#include "jv_unicode.h"
4		#include "jv_utf8_tables.h"
5
6		// jvp_utf8_backtrack returns the beginning of the last codepoint in the
7		// string, assuming that start is the last byte in the string.
8		// If the last codepoint is incomplete, returns the number of missing bytes via
9		// *missing_bytes. If there are no leading bytes or an invalid byte is
10		// encountered, NULL is returned and *missing_bytes is not altered.
11	276	const char* jvp_utf8_backtrack(const char* start, const char* min, int *missing_bytes) {
12	276	assert(min <= start);
13	276	if (min == start) {
14	0	return min;
15	0	}
16	276	int length = 0;
17	276	int seen = 1;
18	276	while ((length = utf8_coding_length[(unsigned char)*start]) == UTF8_CONTINUATION_BYTE) {
19	0	if (start == min) break;
20	0	start--;
21	0	seen++;
22	0	}
23	276	if (length == 0 \|\| length == UTF8_CONTINUATION_BYTE \|\| length - seen < 0) {
24	0	return NULL;
25	0	}
26	276	if (missing_bytes) *missing_bytes = length - seen;
27	276	return start;
28	276	}
29
30	412M	const char* jvp_utf8_next(const char* in, const char* end, int* codepoint_ret) {
31	412M	assert(in <= end);
32	412M	if (in == end) {
33	1.92M	return 0;
34	1.92M	}
35	410M	int codepoint = -1;
36	410M	unsigned char first = (unsigned char)in[0];
37	410M	int length = utf8_coding_length[first];
38	410M	if ((first & 0x80) == 0) {
39		/* Fast-path for ASCII */
40	398M	codepoint = first;
41	398M	length = 1;
42	398M	} else if (length == 0 \|\| length == UTF8_CONTINUATION_BYTE) {
43		/* Bad single byte - either an invalid byte or an out-of-place continuation byte */
44	5.31M	length = 1;
45	6.29M	} else if (length > end - in) {
46		/* String ends before UTF8 sequence ends */
47	338	length = end - in;
48	6.29M	} else {
49	6.29M	codepoint = ((unsigned)in[0]) & utf8_coding_bits[first];
50	10.5M	for (int i=1; i<length; i++) {
51	8.42M	unsigned ch = (unsigned char)in[i];
52	8.42M	if (utf8_coding_length[ch] != UTF8_CONTINUATION_BYTE){
53		/* Invalid UTF8 sequence - not followed by the right number of continuation bytes */
54	4.16M	codepoint = -1;
55	4.16M	length = i;
56	4.16M	break;
57	4.16M	}
58	4.25M	codepoint = (codepoint << 6) \| (ch & 0x3f);
59	4.25M	}
60	6.29M	if (codepoint < utf8_first_codepoint[length]) {
61		/* Overlong UTF8 sequence */
62	4.17M	codepoint = -1;
63	4.17M	}
64	6.29M	if (0xD800 <= codepoint && codepoint <= 0xDFFF) {
65		/* Surrogate codepoints can't be encoded in UTF8 */
66	9	codepoint = -1;
67	9	}
68	6.29M	if (codepoint > 0x10FFFF) {
69		/* Outside Unicode range */
70	1	codepoint = -1;
71	1	}
72	6.29M	}
73	410M	assert(length > 0);
74	410M	*codepoint_ret = codepoint;
75	410M	return in + length;
76	410M	}
77
78	1.92M	int jvp_utf8_is_valid(const char* in, const char* end) {
79	1.92M	int codepoint;
80	105M	while ((in = jvp_utf8_next(in, end, &codepoint))) {
81	103M	if (codepoint == -1) return 0;
82	103M	}
83	1.88M	return 1;
84	1.92M	}
85
86		/* Assumes startchar is the first byte of a valid character sequence */
87	0	int jvp_utf8_decode_length(char startchar) {
88	0	if ((startchar & 0x80) == 0) return 1; // 0___ ____
89	0	else if ((startchar & 0xE0) == 0xC0) return 2; // 110_ ____
90	0	else if ((startchar & 0xF0) == 0xE0) return 3; // 1110 ____
91	0	else return 4; // 1111 ____
92	0	}
93
94	306M	int jvp_utf8_encode_length(int codepoint) {
95	306M	if (codepoint <= 0x7F) return 1;
96	9.64M	else if (codepoint <= 0x7FF) return 2;
97	9.62M	else if (codepoint <= 0xFFFF) return 3;
98	7.69k	else return 4;
99	306M	}
100
101	306M	int jvp_utf8_encode(int codepoint, char* out) {
102	306M	assert(codepoint >= 0 && codepoint <= 0x10FFFF);
103	306M	char* start = out;
104	306M	if (codepoint <= 0x7F) {
105	297M	*out++ = codepoint;
106	297M	} else if (codepoint <= 0x7FF) {
107	23.7k	*out++ = 0xC0 + ((codepoint & 0x7C0) >> 6);
108	23.7k	*out++ = 0x80 + ((codepoint & 0x03F));
109	9.62M	} else if(codepoint <= 0xFFFF) {
110	9.61M	*out++ = 0xE0 + ((codepoint & 0xF000) >> 12);
111	9.61M	*out++ = 0x80 + ((codepoint & 0x0FC0) >> 6);
112	9.61M	*out++ = 0x80 + ((codepoint & 0x003F));
113	9.61M	} else {
114	7.69k	*out++ = 0xF0 + ((codepoint & 0x1C0000) >> 18);
115	7.69k	*out++ = 0x80 + ((codepoint & 0x03F000) >> 12);
116	7.69k	*out++ = 0x80 + ((codepoint & 0x000FC0) >> 6);
117	7.69k	*out++ = 0x80 + ((codepoint & 0x00003F));
118	7.69k	}
119	306M	assert(out - start == jvp_utf8_encode_length(codepoint));
120	306M	return out - start;
121	306M	}
122
123		// characters with White_Space property in:
124		// https://www.unicode.org/Public/UCD/latest/ucd/PropList.txt
125	0	int jvp_codepoint_is_whitespace(int c) {
126	0	return
127	0	(c >= 0x0009 && c <= 0x000D) \|\| // <control-0009>..<control-000D>
128	0	c == 0x0020 \|\| // SPACE
129	0	c == 0x0085 \|\| // <control-0085>
130	0	c == 0x00A0 \|\| // NO-BREAK SPACE
131	0	c == 0x1680 \|\| // OGHAM SPACE MARK
132	0	(c >= 0x2000 && c <= 0x200A) \|\| // EN QUAD..HAIR SPACE
133	0	c == 0x2028 \|\| // LINE SEPARATOR
134	0	c == 0x2029 \|\| // PARAGRAPH SEPARATOR
135	0	c == 0x202F \|\| // NARROW NO-BREAK SPACE
136	0	c == 0x205F \|\| // MEDIUM MATHEMATICAL SPACE
137	0	c == 0x3000 // IDEOGRAPHIC SPACE
138	0	;
139	0	}

Coverage Report

Created: 2026-08-31 06:21