WIP

lib/PluralRulesTokenizer.php

@@ -0,0 +1,137 @@
+<?php
+
+/*
+ * This file is part of the ICanBoogie package.
+ *
+ * (c) Olivier Laviale <olivier.laviale@gmail.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace ICanBoogie\CLDR;
+
+class PluralRulesTokenizer
+{
+	private $data;
+
+	public function __construct(array $data)
+	{
+		$this->data = $data;
+	}
+
+	static public function examples_for($locale)
+	{
+
+	}
+
+	static public function tokenize_witness($data)
+	{
+		$testData = [];
+		foreach ($data as $l => $lData) {
+			$testData[$l] = array();
+			$keys = array_keys($lData);
+			foreach ($keys as $key) {
+				if (!preg_match('/^pluralRule-count-(.+)$/', $key, $m)) {
+					throw new \Exception("Invalid node '$key' in ".$dstFile);
+				}
+				$rule = $m[1];
+				$testData[$l][$rule] = array();
+				$vOriginal = $lData[$key];
+				$examples = explode('@', $vOriginal);
+				$v = trim(array_shift($examples));
+
+//				var_dump($examples); exit;
+
+				foreach ($examples as $example) {
+
+//					var_dump($examples); exit;
+
+					list($exampleNumberType, $exampleValues) = explode(' ', $example, 2);
+					switch ($exampleNumberType) {
+						case 'integer':
+						case 'decimal':
+							$exampleValues = preg_replace('/, …$/', '', $exampleValues);
+							$exampleValuesParsed = array();
+							foreach (explode(', ', trim($exampleValues)) as $ev) {
+								if (preg_match('/^[+\-]?\d+$/', $ev)) {
+									$exampleValuesParsed[] = $ev;
+									$exampleValuesParsed[] = intval($ev);
+								} elseif (preg_match('/^[+\-]?\d+\.\d+$/', $ev)) {
+									$exampleValuesParsed[] = $ev;
+								} elseif (preg_match('/^([+\-]?\d+)~([+\-]?\d+)$/', $ev, $m)) {
+									$exampleValuesParsed[] = $m[1];
+									$exampleValuesParsed[] = intval($m[1]);
+									$exampleValuesParsed[] = $m[2];
+									$exampleValuesParsed[] = intval($m[2]);
+								} elseif (preg_match('/^([+\-]?\d+(\.\d+)?)~([+\-]?\d+(\.\d+)?)$/', $ev, $m)) {
+									$exampleValuesParsed[] = $m[1];
+									$exampleValuesParsed[] = $m[3];
+								} elseif ($ev !== '…') {
+									throw new Exception("Invalid node '$key' in $dstFile: $vOriginal");
+								}
+							}
+							$testData[$l][$rule] = $exampleValuesParsed;
+							break;
+						default:
+							throw new Exception("Invalid node '$key' in $dstFile: $vOriginal");
+					}
+				}
+				if ($rule === 'other') {
+					if (strlen($v) > 0) {
+						throw new Exception("Invalid node '$key' in $dstFile: $vOriginal");
+					}
+				} else {
+					$v = str_replace(' = ', ' == ', $v);
+					$map = array('==' => 'true', '!=' => 'false');
+					foreach (array('^', ' and ', ' or ') as $pre) {
+						while (preg_match(
+							'/'.$pre.'(([nivfwft]( % \\d+)?) (==|!=) ((\\d+)(((\\.\\.)|,)+(\\d+))+))/',
+							$v,
+							$m
+						)) {
+							$found = $m[1];
+							$leftyPart = $m[2]; // eg 'n % 10'
+							$operator = $m[4]; // eg '=='
+							$ranges = explode(',', $m[5]);
+							foreach (array_keys($ranges) as $j) {
+								if (preg_match('/^(\\d+)\\.\\.(\\d+)$/', $ranges[$j], $m)) {
+									$ranges[$j] = "array({$m[1]}, {$m[2]})";
+								}
+							}
+							$v = str_replace($found, "static::inRange($leftyPart, {$map[$operator]}, ".implode(', ', $ranges).')', $v);
+						}
+					}
+					if (strpos($v, '..') !== false) {
+						throw new Exception("Invalid node '$key' in $dstFile: $vOriginal");
+					}
+					foreach (array(
+						         'n' => '%1$s', // absolute value of the source number (integer and decimals).
+						         'i' => '%2$s', // integer digits of n
+						         'v' => '%3$s', // number of visible fraction digits in n, with trailing zeros.
+						         'w' => '%4$s', // number of visible fraction digits in n, without trailing zeros.
+						         'f' => '%5$s', // visible fractional digits in n, with trailing zeros.
+						         't' => '%6$s', // visible fractional digits in n, without trailing zeros.
+					         ) as $from => $to) {
+						$v = preg_replace('/^'.$from.' /', "$to ", $v);
+						$v = preg_replace("/^$from /", "$to ", $v);
+						$v = str_replace(" $from ", " $to ", $v);
+						$v = str_replace("($from, ", "($to, ", $v);
+						$v = str_replace("($from ", "($to ", $v);
+						$v = str_replace(" $from,", " $to,", $v);
+					}
+					$v = str_replace(' % ', ' %% ', $v);
+					$lData[$rule] = $v;
+				}
+				unset($lData[$key]);
+			}
+			$data[$l] = $lData;
+		}
+
+		var_dump($data['ar'], $testData['ar']);
+
+		var_export($testData['ar']);
+
+//		saveJsonFile($testData, TESTS_DIR.DIRECTORY_SEPARATOR.basename($dstFile));
+	}
+}

lib/Plurals.php

@@ -0,0 +1,31 @@
+<?php
+
+/*
+ * This file is part of the ICanBoogie package.
+ *
+ * (c) Olivier Laviale <olivier.laviale@gmail.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace ICanBoogie\CLDR;
+
+use ICanBoogie\CLDR\Plurals\PluralSamples;
+
+class Plurals extends \ArrayObject
+{
+
+	public function examples_for($locale)
+	{
+		$examples = [];
+
+		foreach ($this[$locale] as $count => $rules)
+		{
+			$count = substr($count, strlen('pluralRule-count-'));
+			$examples[$count] = PluralSamples::from($rules);
+		}
+
+		var_dump($examples); exit;
+	}
+}

tests/PluralRulesTokenizerTe_st.php

@@ -0,0 +1,105 @@
+<?php
+
+/*
+ * This file is part of the ICanBoogie package.
+ *
+ * (c) Olivier Laviale <olivier.laviale@gmail.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace ICanBoogie\CLDR;
+
+class PluralRulesTokenizerTest extends \PHPUnit_Framework_TestCase
+{
+	public function test_tokenize()
+	{
+		$data = get_repository()->supplemental['plurals'];
+
+		$plurals = new Plurals($data);
+
+		$plurals->examples_for('fr');
+
+//		PluralRulesTokenizer::tokenize($data);
+	}
+
+	public function provide_examples()
+	{
+		return [
+
+			[ 'ar', [
+
+				'zero' => [
+
+					0 => '0.0',
+					1 => '0.00',
+					2 => '0.000',
+					3 => '0.0000',
+
+				],
+
+				'one' => [
+
+					0 => '1.0',
+					1 => '1.00',
+					2 => '1.000',
+					3 => '1.0000',
+				],
+
+				'two' => [
+
+					0 => '2.0',
+					1 => '2.00',
+					2 => '2.000',
+					3 => '2.0000',
+				],
+
+				'few' => [
+
+					0 => '3.0',
+					1 => '4.0',
+					2 => '5.0',
+					3 => '6.0',
+					4 => '7.0',
+					5 => '8.0',
+					6 => '9.0',
+					7 => '10.0',
+					8 => '103.0',
+					9 => '1003.0',
+
+				],
+
+				'many' => [
+
+					0 => '11.0',
+					1 => '12.0',
+					2 => '13.0',
+					3 => '14.0',
+					4 => '15.0',
+					5 => '16.0',
+					6 => '17.0',
+					7 => '18.0',
+					8 => '111.0',
+					9 => '1011.0',
+
+				],
+
+				'other' => [
+
+					0 => '0.1',
+					1 => '0.9',
+					2 => '1.1',
+					3 => '1.7',
+					4 => '10.1',
+					5 => '100.0',
+					6 => '1000.0',
+					7 => '10000.0',
+					8 => '100000.0',
+					9 => '1000000.0',
+
+				],
+			] ]
+		];
+	}
+}