/**
* Returns an array of Unicode code points.
*
* @param string $str The UTF-8 encoded string
* @param bool $strict Should the function throw an error if something goes wrong
* @return array of Unicode code points or false on failure
*/
function utf8_to_unicode($str, $strict = false)
{
$rv = array();
$i = 0;
$len = strlen($str);
while ($i < $len) {
$n = $i;
$ord = ord($str[$i++]);
// Move to the next octet at once
if (utf8_is_ascii($ord)) {
$rv[] = $ord;
} elseif (utf8_is_identifier($ord, $seqlen)) {
$cp = utf8_decode_identifier($ord, $seqlen);
$offset = $seqlen - 2;
// Check for exists the last octet of the sequence at once
if ($i + $offset >= $len) {
if ($strict) {
trigger_error(sprintf('utf8_to_unicode: Incomplete ' . 'multi-octet sequence in UTF-8 at octet: %d', $n));
return false;
}
continue;
}
do {
$ord = ord($str[$i]);
if (!utf8_is_trail($ord)) {
if ($strict) {
trigger_error(sprintf('utf8_to_unicode: Unexpected ' . 'value at octet: %d, value: 0x%X', $i, $ord));
return false;
}
continue 2;
// Prepare this octet again
}
$i++;
$cp |= utf8_decode_trail($ord, $offset);
} while ($offset--);
if (!unicode_is_valid($cp)) {
if ($strict) {
$msg = unicode_is_surrogate($cp) ? 'Illegal surrogate in UTF-8' : 'Codepoint out of Unicode range';
trigger_error(sprintf('utf8_to_unicode: %s at octet: %d, ' . 'value: 0x%X', $msg, $n, $cp));
return false;
}
continue;
}
if (utf8_sequence_length($cp) != $seqlen) {
if ($strict) {
trigger_error(sprintf('utf8_to_unicode: Illegal sequence ' . 'identifier in UTF-8 at octet: %d, value: 0x%X', $n, $cp));
return false;
}
continue;
}
$rv[] = $cp;
} elseif ($strict) {
trigger_error(sprintf('utf8_to_unicode: Illegal sequence ' . 'identifier in UTF-8 at octet: %d, value: 0x%X', $n, $ord));
return false;
}
}
return $rv;
}
/**
* @dataProvider providerUtf8IsIdentifier
*/
public function testUtf8IsIdentifier($ord, $rv, $rseqlen)
{
$this->assertEquals(utf8_is_identifier($ord, $seqlen), $rv);
$this->assertEquals($seqlen, $rseqlen);
}
/**
* This function may return boolean false, but may also return a non-boolean 0
* which evaluates to false. Use the === operator for testing the return value
* of this function.
*
* @param string $str The UTF-8 encoded string
* @param int $i The current position of the string
* @param int $rcp If passed, this will be set to the Unicode code point
* for the character
* @return string The character or false on the start of the string
*/
function utf8_get_last_char($str, &$i = null, &$rcp = null)
{
if ($i === null) {
$i = strlen($str) - 1;
}
while (isset($str[$i])) {
$n = $i--;
// Save current position and move to the previous octet
$char = $str[$n++];
// Move to the next octet at once
$ord = ord($char);
if (utf8_is_ascii($ord)) {
$rcp = $ord;
return $char;
}
if (utf8_is_identifier($ord, $seqlen)) {
$cp = utf8_decode_identifier($ord, $seqlen);
$offset = $seqlen - 2;
// Check for exists the last octet of the sequence at once
if (!isset($str[$n + $offset])) {
continue;
}
do {
$ord = ord($str[$n]);
if (!utf8_is_trail($ord)) {
continue 2;
// Skip this sequence
}
$cp |= utf8_decode_trail($ord, $offset);
$char .= $str[$n++];
} while ($offset--);
if (unicode_is_valid($cp) && utf8_sequence_length($cp) == $seqlen) {
$rcp = $cp;
return $char;
}
}
}
return false;
}