Subversion Repositories HelenOS

Compare Revisions

Ignore whitespace Rev 4262 → Rev 4263

/branches/network/uspace/lib/libc/generic/string.c
38,7 → 38,574
#include <limits.h>
#include <ctype.h>
#include <malloc.h>
#include <errno.h>
#include <align.h>
#include <string.h>
 
/** Byte mask consisting of lowest @n bits (out of 8) */
#define LO_MASK_8(n) ((uint8_t) ((1 << (n)) - 1))
 
/** Byte mask consisting of lowest @n bits (out of 32) */
#define LO_MASK_32(n) ((uint32_t) ((1 << (n)) - 1))
 
/** Byte mask consisting of highest @n bits (out of 8) */
#define HI_MASK_8(n) (~LO_MASK_8(8 - (n)))
 
/** Number of data bits in a UTF-8 continuation byte */
#define CONT_BITS 6
 
/** Decode a single character from a string.
*
* Decode a single character from a string of size @a size. Decoding starts
* at @a offset and this offset is moved to the beginning of the next
* character. In case of decoding error, offset generally advances at least
* by one. However, offset is never moved beyond size.
*
* @param str String (not necessarily NULL-terminated).
* @param offset Byte offset in string where to start decoding.
* @param size Size of the string (in bytes).
*
* @return Value of decoded character, U_SPECIAL on decoding error or
* NULL if attempt to decode beyond @a size.
*
*/
wchar_t str_decode(const char *str, size_t *offset, size_t size)
{
if (*offset + 1 > size)
return 0;
/* First byte read from string */
uint8_t b0 = (uint8_t) str[(*offset)++];
/* Determine code length */
unsigned int b0_bits; /* Data bits in first byte */
unsigned int cbytes; /* Number of continuation bytes */
if ((b0 & 0x80) == 0) {
/* 0xxxxxxx (Plain ASCII) */
b0_bits = 7;
cbytes = 0;
} else if ((b0 & 0xe0) == 0xc0) {
/* 110xxxxx 10xxxxxx */
b0_bits = 5;
cbytes = 1;
} else if ((b0 & 0xf0) == 0xe0) {
/* 1110xxxx 10xxxxxx 10xxxxxx */
b0_bits = 4;
cbytes = 2;
} else if ((b0 & 0xf8) == 0xf0) {
/* 11110xxx 10xxxxxx 10xxxxxx 10xxxxxx */
b0_bits = 3;
cbytes = 3;
} else {
/* 10xxxxxx -- unexpected continuation byte */
return U_SPECIAL;
}
if (*offset + cbytes > size)
return U_SPECIAL;
wchar_t ch = b0 & LO_MASK_8(b0_bits);
/* Decode continuation bytes */
while (cbytes > 0) {
uint8_t b = (uint8_t) str[(*offset)++];
/* Must be 10xxxxxx */
if ((b & 0xc0) != 0x80)
return U_SPECIAL;
/* Shift data bits to ch */
ch = (ch << CONT_BITS) | (wchar_t) (b & LO_MASK_8(CONT_BITS));
cbytes--;
}
return ch;
}
 
/** Encode a single character to string representation.
*
* Encode a single character to string representation (i.e. UTF-8) and store
* it into a buffer at @a offset. Encoding starts at @a offset and this offset
* is moved to the position where the next character can be written to.
*
* @param ch Input character.
* @param str Output buffer.
* @param offset Byte offset where to start writing.
* @param size Size of the output buffer (in bytes).
*
* @return EOK if the character was encoded successfully, EOVERFLOW if there
* was not enough space in the output buffer or EINVAL if the character
* code was invalid.
*/
int chr_encode(const wchar_t ch, char *str, size_t *offset, size_t size)
{
if (*offset >= size)
return EOVERFLOW;
if (!chr_check(ch))
return EINVAL;
/* Unsigned version of ch (bit operations should only be done
on unsigned types). */
uint32_t cc = (uint32_t) ch;
/* Determine how many continuation bytes are needed */
unsigned int b0_bits; /* Data bits in first byte */
unsigned int cbytes; /* Number of continuation bytes */
if ((cc & ~LO_MASK_32(7)) == 0) {
b0_bits = 7;
cbytes = 0;
} else if ((cc & ~LO_MASK_32(11)) == 0) {
b0_bits = 5;
cbytes = 1;
} else if ((cc & ~LO_MASK_32(16)) == 0) {
b0_bits = 4;
cbytes = 2;
} else if ((cc & ~LO_MASK_32(21)) == 0) {
b0_bits = 3;
cbytes = 3;
} else {
/* Codes longer than 21 bits are not supported */
return EINVAL;
}
/* Check for available space in buffer */
if (*offset + cbytes >= size)
return EOVERFLOW;
/* Encode continuation bytes */
unsigned int i;
for (i = cbytes; i > 0; i--) {
str[*offset + i] = 0x80 | (cc & LO_MASK_32(CONT_BITS));
cc = cc >> CONT_BITS;
}
/* Encode first byte */
str[*offset] = (cc & LO_MASK_32(b0_bits)) | HI_MASK_8(8 - b0_bits - 1);
/* Advance offset */
*offset += cbytes + 1;
return EOK;
}
 
/** Get size of string.
*
* Get the number of bytes which are used by the string @a str (excluding the
* NULL-terminator).
*
* @param str String to consider.
*
* @return Number of bytes used by the string
*
*/
size_t str_size(const char *str)
{
size_t size = 0;
while (*str++ != 0)
size++;
return size;
}
 
/** Get size of wide string.
*
* Get the number of bytes which are used by the wide string @a str (excluding the
* NULL-terminator).
*
* @param str Wide string to consider.
*
* @return Number of bytes used by the wide string
*
*/
size_t wstr_size(const wchar_t *str)
{
return (wstr_length(str) * sizeof(wchar_t));
}
 
/** Get size of string with length limit.
*
* Get the number of bytes which are used by up to @a max_len first
* characters in the string @a str. If @a max_len is greater than
* the length of @a str, the entire string is measured (excluding the
* NULL-terminator).
*
* @param str String to consider.
* @param max_len Maximum number of characters to measure.
*
* @return Number of bytes used by the characters.
*
*/
size_t str_lsize(const char *str, count_t max_len)
{
count_t len = 0;
size_t offset = 0;
while (len < max_len) {
if (str_decode(str, &offset, STR_NO_LIMIT) == 0)
break;
len++;
}
return offset;
}
 
/** Get size of wide string with length limit.
*
* Get the number of bytes which are used by up to @a max_len first
* wide characters in the wide string @a str. If @a max_len is greater than
* the length of @a str, the entire wide string is measured (excluding the
* NULL-terminator).
*
* @param str Wide string to consider.
* @param max_len Maximum number of wide characters to measure.
*
* @return Number of bytes used by the wide characters.
*
*/
size_t wstr_lsize(const wchar_t *str, count_t max_len)
{
return (wstr_nlength(str, max_len * sizeof(wchar_t)) * sizeof(wchar_t));
}
 
/** Get number of characters in a string.
*
* @param str NULL-terminated string.
*
* @return Number of characters in string.
*
*/
count_t str_length(const char *str)
{
count_t len = 0;
size_t offset = 0;
while (str_decode(str, &offset, STR_NO_LIMIT) != 0)
len++;
return len;
}
 
/** Get number of characters in a wide string.
*
* @param str NULL-terminated wide string.
*
* @return Number of characters in @a str.
*
*/
count_t wstr_length(const wchar_t *wstr)
{
count_t len = 0;
while (*wstr++ != 0)
len++;
return len;
}
 
/** Get number of characters in a string with size limit.
*
* @param str NULL-terminated string.
* @param size Maximum number of bytes to consider.
*
* @return Number of characters in string.
*
*/
count_t str_nlength(const char *str, size_t size)
{
count_t len = 0;
size_t offset = 0;
while (str_decode(str, &offset, size) != 0)
len++;
return len;
}
 
/** Get number of characters in a string with size limit.
*
* @param str NULL-terminated string.
* @param size Maximum number of bytes to consider.
*
* @return Number of characters in string.
*
*/
count_t wstr_nlength(const wchar_t *str, size_t size)
{
count_t len = 0;
count_t limit = ALIGN_DOWN(size, sizeof(wchar_t));
count_t offset = 0;
while ((offset < limit) && (*str++ != 0)) {
len++;
offset += sizeof(wchar_t);
}
return len;
}
 
/** Check whether character is plain ASCII.
*
* @return True if character is plain ASCII.
*
*/
bool ascii_check(wchar_t ch)
{
if ((ch >= 0) && (ch <= 127))
return true;
return false;
}
 
/** Check whether character is valid
*
* @return True if character is a valid Unicode code point.
*
*/
bool chr_check(wchar_t ch)
{
if ((ch >= 0) && (ch <= 1114111))
return true;
return false;
}
 
/** Compare two NULL terminated strings.
*
* Do a char-by-char comparison of two NULL-terminated strings.
* The strings are considered equal iff they consist of the same
* characters on the minimum of their lengths.
*
* @param s1 First string to compare.
* @param s2 Second string to compare.
*
* @return 0 if the strings are equal, -1 if first is smaller,
* 1 if second smaller.
*
*/
int str_cmp(const char *s1, const char *s2)
{
wchar_t c1 = 0;
wchar_t c2 = 0;
size_t off1 = 0;
size_t off2 = 0;
 
while (true) {
c1 = str_decode(s1, &off1, STR_NO_LIMIT);
c2 = str_decode(s2, &off2, STR_NO_LIMIT);
 
if (c1 < c2)
return -1;
if (c1 > c2)
return 1;
 
if (c1 == 0 || c2 == 0)
break;
}
 
return 0;
}
 
/** Compare two NULL terminated strings with length limit.
*
* Do a char-by-char comparison of two NULL-terminated strings.
* The strings are considered equal iff they consist of the same
* characters on the minimum of their lengths and the length limit.
*
* @param s1 First string to compare.
* @param s2 Second string to compare.
* @param max_len Maximum number of characters to consider.
*
* @return 0 if the strings are equal, -1 if first is smaller,
* 1 if second smaller.
*
*/
int str_lcmp(const char *s1, const char *s2, count_t max_len)
{
wchar_t c1 = 0;
wchar_t c2 = 0;
size_t off1 = 0;
size_t off2 = 0;
count_t len = 0;
 
while (true) {
if (len >= max_len)
break;
 
c1 = str_decode(s1, &off1, STR_NO_LIMIT);
c2 = str_decode(s2, &off2, STR_NO_LIMIT);
 
if (c1 < c2)
return -1;
 
if (c1 > c2)
return 1;
 
if (c1 == 0 || c2 == 0)
break;
 
++len;
}
 
return 0;
 
}
 
/** Copy NULL-terminated string.
*
* Copy source string @a src to destination buffer @a dst.
* No more than @a size bytes are written. NULL-terminator is always
* written after the last succesfully copied character (i.e. if the
* destination buffer is has at least 1 byte, it will be always
* NULL-terminated).
*
* @param src Source string.
* @param dst Destination buffer.
* @param count Size of the destination buffer.
*
*/
void str_ncpy(char *dst, const char *src, size_t size)
{
/* No space for the NULL-terminator in the buffer */
if (size == 0)
return;
wchar_t ch;
size_t str_off = 0;
size_t dst_off = 0;
while ((ch = str_decode(src, &str_off, STR_NO_LIMIT)) != 0) {
if (chr_encode(ch, dst, &dst_off, size) != EOK)
break;
}
if (dst_off >= size)
dst[size - 1] = 0;
else
dst[dst_off] = 0;
}
 
/** Copy NULL-terminated wide string to string
*
* Copy source wide string @a src to destination buffer @a dst.
* No more than @a size bytes are written. NULL-terminator is always
* written after the last succesfully copied character (i.e. if the
* destination buffer is has at least 1 byte, it will be always
* NULL-terminated).
*
* @param src Source wide string.
* @param dst Destination buffer.
* @param count Size of the destination buffer.
*
*/
void wstr_nstr(char *dst, const wchar_t *src, size_t size)
{
/* No space for the NULL-terminator in the buffer */
if (size == 0)
return;
wchar_t ch;
count_t src_idx = 0;
size_t dst_off = 0;
while ((ch = src[src_idx++]) != 0) {
if (chr_encode(ch, dst, &dst_off, size) != EOK)
break;
}
if (dst_off >= size)
dst[size - 1] = 0;
else
dst[dst_off] = 0;
}
 
/** Find first occurence of character in string.
*
* @param str String to search.
* @param ch Character to look for.
*
* @return Pointer to character in @a str or NULL if not found.
*
*/
const char *str_chr(const char *str, wchar_t ch)
{
wchar_t acc;
size_t off = 0;
while ((acc = str_decode(str, &off, STR_NO_LIMIT)) != 0) {
if (acc == ch)
return (str + off);
}
return NULL;
}
 
/** Insert a wide character into a wide string.
*
* Insert a wide character into a wide string at position
* @a pos. The characters after the position are shifted.
*
* @param str String to insert to.
* @param ch Character to insert to.
* @param pos Character index where to insert.
@ @param max_pos Characters in the buffer.
*
* @return True if the insertion was sucessful, false if the position
* is out of bounds.
*
*/
bool wstr_linsert(wchar_t *str, wchar_t ch, count_t pos, count_t max_pos)
{
count_t len = wstr_length(str);
if ((pos > len) || (pos + 1 > max_pos))
return false;
count_t i;
for (i = len; i + 1 > pos; i--)
str[i + 1] = str[i];
str[pos] = ch;
return true;
}
 
/** Remove a wide character from a wide string.
*
* Remove a wide character from a wide string at position
* @a pos. The characters after the position are shifted.
*
* @param str String to remove from.
* @param pos Character index to remove.
*
* @return True if the removal was sucessful, false if the position
* is out of bounds.
*
*/
bool wstr_remove(wchar_t *str, count_t pos)
{
count_t len = wstr_length(str);
if (pos >= len)
return false;
count_t i;
for (i = pos + 1; i <= len; i++)
str[i - 1] = str[i];
return true;
}
 
/** Count the number of characters in the string, not including terminating 0.
*
* @param str String.