SQLite UDFs for LIKE Operator Overloading

sqlite.udf.like.icu.php

<?php

$db->sqliteCreateFunction('like',
	function ($pattern, $data, $escape = null) use ($db)
	{
		static $modifiers = null;

		if (isset($modifiers) !== true)
		{
			$modifiers = ((strncmp($db->query('PRAGMA case_sensitive_like;')->fetchColumn(), '1', 1) === 0) ? '' : 'i') . 'suS';
		}

		if (isset($data) === true)
		{
			if (strpbrk($pattern = preg_quote($pattern, '~'), '%_') !== false)
			{
				$regex = array
				(
					'~%+~S' => '.*',
					'~_~S' => '.',
				);

				if (strlen($escape = preg_quote($escape, '~')) > 0)
				{
					$regex = array
					(
						'~(?<!' . $escape . ')%+~S' => '.*',
						'~(?<!' . $escape . ')_~S' => '.',
						'~(?:' . preg_quote($escape, '~') . ')([%_])~S' => '$1',
					);
				}

				$pattern = preg_replace(array_keys($regex), $regex, $pattern);
			}

			return (preg_match(sprintf('~^%s$~%s', $pattern, $modifiers), $data) > 0);
		}

		return false;
	}
);

sqlite.udf.like.icu.roman.extended.php

<?php

$db->sqliteCreateFunction('like',
	function ($pattern, $data, $escape = null) use ($db)
	{
		static $modifiers = null;

		if (isset($modifiers) !== true)
		{
			$modifiers = ((strncmp($db->query('PRAGMA case_sensitive_like;')->fetchColumn(), '1', 1) === 0) ? '' : 'i') . 'suS';
		}

		if (isset($data) === true)
		{
			foreach (array('pattern', 'data') as $key)
			{
				if (strpos(${$key} = htmlentities(${$key}, ENT_QUOTES, 'UTF-8'), '&') !== false)
				{
					${$key} = html_entity_decode(preg_replace('~&([a-z]{1,2})(?:acute|caron|cedil|circ|grave|lig|orn|ring|slash|tilde|uml);~iS', '$1', ${$key}), ENT_QUOTES, 'UTF-8');
				}

				if (preg_match('~[^[:ascii:]]~S', ${$key}) > 0)
				{
					$latin = array
					(
						'a' => '~[ǎăāǻą]~uS',
						'A' => '~[ǍĂĀǺĄ]~uS',
						'ae' => '~[ǽǣ]~uS',
						'AE' => '~[ǼǢ]~uS',
						'b' => '~[ɓ]~uS',
						'B' => '~[Ɓ]~uS',
						'c' => '~[ćċĉč]~uS',
						'C' => '~[ĆĊĈČ]~uS',
						'd' => '~[ďḍđɗð]~uS',
						'D' => '~[ĎḌĐƊÐ]~uS',
						'e' => '~[ėěĕēęẹǝəɛ]~uS',
						'E' => '~[ĖĚĔĒĘẸƎƏƐ]~uS',
						'g' => '~[ġĝǧğģɣ]~uS',
						'G' => '~[ĠĜǦĞĢƔ]~uS',
						'h' => '~[ĥḥħ]~uS',
						'H' => '~[ĤḤĦ]~uS',
						'I' => '~[IǏĬĪĨĮỊİ]~uS',
						'i' => '~[ıǐĭīĩįịİ]~uS',
						'ij' => '~[ij]~uS',
						'IJ' => '~[IJ]~uS',
						'j' => '~[ĵ]~uS',
						'J' => '~[Ĵ]~uS',
						'k' => '~[ķƙĸ]~uS',
						'K' => '~[ĶƘĸ]~uS',
						'l' => '~[ĺļłľŀ]~uS',
						'L' => '~[ĹĻŁĽĿ]~uS',
						'n' => '~[ʼnń̈ňņŋ]~uS',
						'N' => '~[ʼnŃ̈ŇŅŊ]~uS',
						'o' => '~[ǒŏōőǫọǿơ]~uS',
						'O' => '~[ǑŎŌŐǪỌǾƠ]~uS',
						'r' => '~[ŕřŗ]~uS',
						'R' => '~[ŔŘŖ]~uS',
						'S' => '~[SŚŜŞȘṢ]~uS',
						's' => '~[ſśŝşșṣ]~uS',
						't' => '~[ťţṭŧ]~uS',
						'T' => '~[ŤŢṬŦ]~uS',
						'u' => '~[ǔŭūũűůųụư]~uS',
						'U' => '~[ǓŬŪŨŰŮŲỤƯ]~uS',
						'w' => '~[ẃẁŵẅƿ]~uS',
						'W' => '~[ẂẀŴẄǷ]~uS',
						'y' => '~[ỳŷȳỹƴ]~uS',
						'Y' => '~[ỲŶȲỸƳ]~uS',
						'z' => '~[źżžẓ]~uS',
						'Z' => '~[ŹŻŽẒ]~uS',
					);

					${$key} = preg_replace($latin, array_keys($latin), ${$key});
				}
			}

			if (strpbrk($pattern = preg_quote($pattern, '~'), '%_') !== false)
			{
				$regex = array
				(
					'~%+~S' => '.*',
					'~_~S' => '.',
				);

				if (strlen($escape = preg_quote($escape, '~')) > 0)
				{
					$regex = array
					(
						'~(?<!' . $escape . ')%+~S' => '.*',
						'~(?<!' . $escape . ')_~S' => '.',
						'~(?:' . preg_quote($escape, '~') . ')([%_])~S' => '$1',
					);
				}

				$pattern = preg_replace(array_keys($regex), $regex, $pattern);
			}

			return (preg_match(sprintf('~^%s$~%s', $pattern, $modifiers), $data) > 0);
		}

		return false;
	}
);

sqlite.udf.like.icu.roman.php

<?php

$db->sqliteCreateFunction('like',
	function ($pattern, $data, $escape = null) use ($db)
	{
		static $modifiers = null;

		if (isset($modifiers) !== true)
		{
			$modifiers = ((strncmp($db->query('PRAGMA case_sensitive_like;')->fetchColumn(), '1', 1) === 0) ? '' : 'i') . 'suS';
		}

		if (isset($data) === true)
		{
			foreach (array('pattern', 'data') as $key)
			{
				if (strpos(${$key} = htmlentities(${$key}, ENT_QUOTES, 'UTF-8'), '&') !== false)
				{
					${$key} = html_entity_decode(preg_replace('~&([a-z]{1,2})(?:acute|caron|cedil|circ|grave|lig|orn|ring|slash|tilde|uml);~iS', '$1', ${$key}), ENT_QUOTES, 'UTF-8');
				}
			}

			if (strpbrk($pattern = preg_quote($pattern, '~'), '%_') !== false)
			{
				$regex = array
				(
					'~%+~S' => '.*',
					'~_~S' => '.',
				);

				if (strlen($escape = preg_quote($escape, '~')) > 0)
				{
					$regex = array
					(
						'~(?<!' . $escape . ')%+~S' => '.*',
						'~(?<!' . $escape . ')_~S' => '.',
						'~(?:' . preg_quote($escape, '~') . ')([%_])~S' => '$1',
					);
				}

				$pattern = preg_replace(array_keys($regex), $regex, $pattern);
			}

			return (preg_match(sprintf('~^%s$~%s', $pattern, $modifiers), $data) > 0);
		}

		return false;
	}
);