Added support for specifying the list of strings as JSON

Author: JoshyPHP

README.md

@@ -37,6 +37,13 @@ $ ./build-regexp.phar --infile strings.txt
 (?:one|t(?:hree|wo))
 ```
 
+Alternatively, the list of strings can be passed as a JSON array:
+```
+$ echo '["foo","bar"]' > strings.json
+$ ./build-regexp.phar --infile strings.json --infile-format json
+(?:bar|foo)
+```
+
 By default, the result is output in the terminal directly. Alternatively, it can be saved to a file specified via the `outfile` option. In the following example, we save the result to a `out.txt` file before checking its content:
 ```
 $ ./build-regexp.phar --outfile out.txt foo bar baz

composer.json

@@ -20,7 +20,7 @@
 	},
 	"autoload-dev": {
 		"psr-4": {
-			"s9e\\RegexpBuilder\\Tests\\Command\\": "tests"
+			"s9e\\RegexpBuilder\\Command\\Tests\\": "tests"
 		}
 	},
 	"scripts": {

src/Build.php

@@ -16,6 +16,7 @@
 use Symfony\Component\Console\Output\OutputInterface;
 use UnhandledMatchError;
 use s9e\RegexpBuilder\Builder;
+use s9e\RegexpBuilder\Command\InputFormatter\InputFormatterInterface;
 
 class Build extends Command
 {
@@ -28,7 +29,8 @@ protected function configure(): void
 			null,
 			InputOption::VALUE_REQUIRED,
 			'Regexp preset: "java", "javascript", "pcre", "raw", or "re2"',
-			'pcre'
+			'pcre',
+			['java', 'javascript', 'pcre', 'raw', 're2']
 		);
 		$this->addOption(
 			'flags',
@@ -49,6 +51,14 @@ protected function configure(): void
 			InputOption::VALUE_REQUIRED,
 			'Input file'
 		);
+		$this->addOption(
+			'infile-format',
+			null,
+			InputOption::VALUE_REQUIRED,
+			'Format of the input file: "json" or "lsv" (line-separated values)',
+			'lsv',
+			['json', 'lsv']
+		);
 		$this->addOption(
 			'outfile',
 			null,
@@ -147,7 +157,18 @@ protected function getStrings(InputInterface $input): array
 
 		$text = ($filepath === '-') ? $this->readStdin($input) : $this->readFile($filepath);
 
-		return preg_split('(\\r?\\n)', $text, -1, PREG_SPLIT_NO_EMPTY);
+		return $this->getInputFormatter($input)->format($text);
+	}
+
+	protected function getInputFormatter(InputInterface $input): InputFormatterInterface
+	{
+		$className = __NAMESPACE__ . '\\InputFormatter\\' . ucfirst(strtolower($input->getOption('infile-format')));
+		if (!class_exists($className))
+		{
+			throw new RuntimeException('Unsupported infile-format');
+		}
+
+		return new $className;
 	}
 
 	protected function readFile(string $filepath): string

src/InputFormatter/InputFormatterInterface.php

@@ -0,0 +1,13 @@
+<?php declare(strict_types=1);
+
+/**
+* @package   s9e\RegexpBuilderCommand
+* @copyright Copyright (c) 2022 The s9e authors
+* @license   http://www.opensource.org/licenses/mit-license.php The MIT License
+*/
+namespace s9e\RegexpBuilder\Command\InputFormatter;
+
+interface InputFormatterInterface
+{
+	public function format(string $text): array;
+}

src/InputFormatter/Json.php

@@ -0,0 +1,35 @@
+<?php declare(strict_types=1);
+
+/**
+* @package   s9e\RegexpBuilderCommand
+* @copyright Copyright (c) 2022 The s9e authors
+* @license   http://www.opensource.org/licenses/mit-license.php The MIT License
+*/
+namespace s9e\RegexpBuilder\Command\InputFormatter;
+
+use RuntimeException;
+
+class Json implements InputFormatterInterface
+{
+	public function format(string $text): array
+	{
+		$values = json_decode($text);
+		if ($values === null)
+		{
+			throw new RuntimeException('Input is not valid JSON');
+		}
+		if (!is_array($values))
+		{
+			throw new RuntimeException('Input is not a JSON array');
+		}
+		foreach ($values as $value)
+		{
+			if (!is_string($value))
+			{
+				throw new RuntimeException('Input contains non-string values');
+			}
+		}
+
+		return $values;
+	}
+}

src/InputFormatter/Lsv.php

@@ -0,0 +1,16 @@
+<?php declare(strict_types=1);
+
+/**
+* @package   s9e\RegexpBuilderCommand
+* @copyright Copyright (c) 2022 The s9e authors
+* @license   http://www.opensource.org/licenses/mit-license.php The MIT License
+*/
+namespace s9e\RegexpBuilder\Command\InputFormatter;
+
+class Lsv implements InputFormatterInterface
+{
+	public function format(string $text): array
+	{
+		return preg_split('(\\r?\\n)', $text, -1, PREG_SPLIT_NO_EMPTY);
+	}
+}

tests/BuildTest.php

@@ -14,7 +14,6 @@
 */
 class BuildTest extends TestCase
 {
-
 	public function testWriteFile()
 	{
 		vfsStream::setup('root');
@@ -167,6 +166,28 @@ function (CommandTester $commandTester): void
 					file_put_contents(vfsStream::url('root/strings.txt'), "a\nb");
 				}
 			],
+			[
+				[
+					'--infile'        => vfsStream::url('root/strings.txt'),
+					'--infile-format' => 'lsv'
+				],
+				'[ab]',
+				function (CommandTester $commandTester): void
+				{
+					file_put_contents(vfsStream::url('root/strings.txt'), "a\nb");
+				}
+			],
+			[
+				[
+					'--infile'        => vfsStream::url('root/strings.json'),
+					'--infile-format' => 'json'
+				],
+				'[abd]',
+				function (CommandTester $commandTester): void
+				{
+					file_put_contents(vfsStream::url('root/strings.json'), '["a","b","d"]');
+				}
+			],
 			[
 				[
 					'--infile' => '-'
@@ -272,6 +293,28 @@ function ()
 					chmod($path, 0);
 				}
 			],
+			[
+				[
+					'--infile'        => vfsStream::url('root/invalid.json'),
+					'--infile-format' => 'json'
+				],
+				new RuntimeException('Input is not valid JSON'),
+				function (): void
+				{
+					file_put_contents(vfsStream::url('root/invalid.json'), 'invalid');
+				}
+			],
+			[
+				[
+					'--infile'        => vfsStream::url('root/file.txt'),
+					'--infile-format' => 'idk'
+				],
+				new RuntimeException('Unsupported infile-format'),
+				function (): void
+				{
+					file_put_contents(vfsStream::url('root/file.txt'), '');
+				}
+			],
 		];
 	}
 }

tests/InputFormatter/JsonTest.php

@@ -0,0 +1,52 @@
+<?php declare(strict_types=1);
+
+namespace s9e\RegexpBuilder\Command\Tests;
+
+use PHPUnit\Framework\TestCase;
+use s9e\RegexpBuilder\Command\InputFormatter\Json;
+
+/**
+* @covers s9e\RegexpBuilder\Command\InputFormatter\Json
+*/
+class JsonTest extends TestCase
+{
+	public function test()
+	{
+		$this->assertEquals(
+			['foo', 'bar', 'baz'],
+			(new Json)->format('["foo","bar","baz"]')
+		);
+	}
+
+	public function testUnicode()
+	{
+		$this->assertEquals(
+			["\xF0\x9F\x98\x80"],
+			(new Json)->format('["\\uD83D\\uDE00"]')
+		);
+	}
+
+	public function testInvalidJson()
+	{
+		$this->expectException('RuntimeException');
+		$this->expectExceptionMessage('Input is not valid JSON');
+
+		(new Json)->format('invalid');
+	}
+
+	public function testInvalidJsonType()
+	{
+		$this->expectException('RuntimeException');
+		$this->expectExceptionMessage('Input is not a JSON array');
+
+		(new Json)->format('{"foo":"bar"}');
+	}
+
+	public function testInvalidJsonValue()
+	{
+		$this->expectException('RuntimeException');
+		$this->expectExceptionMessage('Input contains non-string values');
+
+		(new Json)->format('["foo",1]');
+	}
+}

tests/InputFormatter/LsvTest.php

@@ -0,0 +1,20 @@
+<?php declare(strict_types=1);
+
+namespace s9e\RegexpBuilder\Command\Tests;
+
+use PHPUnit\Framework\TestCase;
+use s9e\RegexpBuilder\Command\InputFormatter\Lsv;
+
+/**
+* @covers s9e\RegexpBuilder\Command\InputFormatter\Lsv
+*/
+class LsvTest extends TestCase
+{
+	public function test()
+	{
+		$this->assertEquals(
+			['foo', 'bar', 'baz'],
+			(new Lsv)->format("foo\nbar\r\nbaz")
+		);
+	}
+}