From 235b150e0921292b4b98df3059bb5ffc3220631c Mon Sep 17 00:00:00 2001
From: Bogdan Condorachi <bogdan.condorachi@gmail.com>
Date: Wed, 15 May 2024 22:38:27 +0300
Subject: [PATCH 1/5] Add python support

---
 src/Highlighter.php                           |  2 +
 .../Python/Patterns/PyArgumentPattern.php     | 24 ++++++
 .../Python/Patterns/PyBooleanPattern.php      | 24 ++++++
 .../Python/Patterns/PyBuiltinPattern.php      | 38 ++++++++++
 .../Python/Patterns/PyClassNamePattern.php    | 27 +++++++
 .../Python/Patterns/PyCommentPattern.php      | 24 ++++++
 .../Python/Patterns/PyDecoratorPattern.php    | 27 +++++++
 .../Python/Patterns/PyFunctionPattern.php     | 26 +++++++
 .../Python/Patterns/PyKeywordPattern.php      | 33 +++++++++
 .../Python/Patterns/PyOperatorPattern.php     | 24 ++++++
 .../PyTripleDoubleQuoteStringPattern.php      | 24 ++++++
 .../PyTripleSingleQuoteStringPattern.php      | 24 ++++++
 src/Languages/Python/PythonLanguage.php       | 74 +++++++++++++++++++
 13 files changed, 371 insertions(+)
 create mode 100644 src/Languages/Python/Patterns/PyArgumentPattern.php
 create mode 100644 src/Languages/Python/Patterns/PyBooleanPattern.php
 create mode 100644 src/Languages/Python/Patterns/PyBuiltinPattern.php
 create mode 100644 src/Languages/Python/Patterns/PyClassNamePattern.php
 create mode 100644 src/Languages/Python/Patterns/PyCommentPattern.php
 create mode 100644 src/Languages/Python/Patterns/PyDecoratorPattern.php
 create mode 100644 src/Languages/Python/Patterns/PyFunctionPattern.php
 create mode 100644 src/Languages/Python/Patterns/PyKeywordPattern.php
 create mode 100644 src/Languages/Python/Patterns/PyOperatorPattern.php
 create mode 100644 src/Languages/Python/Patterns/PyTripleDoubleQuoteStringPattern.php
 create mode 100644 src/Languages/Python/Patterns/PyTripleSingleQuoteStringPattern.php
 create mode 100644 src/Languages/Python/PythonLanguage.php

diff --git a/src/Highlighter.php b/src/Highlighter.php
index 1c76e06..1e24796 100644
--- a/src/Highlighter.php
+++ b/src/Highlighter.php
@@ -17,6 +17,7 @@
 use Tempest\Highlight\Languages\JavaScript\JavaScriptLanguage;
 use Tempest\Highlight\Languages\Json\JsonLanguage;
 use Tempest\Highlight\Languages\Php\PhpLanguage;
+use Tempest\Highlight\Languages\Python\PythonLanguage;
 use Tempest\Highlight\Languages\Sql\SqlLanguage;
 use Tempest\Highlight\Languages\Text\TextLanguage;
 use Tempest\Highlight\Languages\Twig\TwigLanguage;
@@ -49,6 +50,7 @@ public function __construct(
             ->addLanguage(new JavaScriptLanguage())
             ->addLanguage(new JsonLanguage())
             ->addLanguage(new PhpLanguage())
+            ->addLanguage(new PythonLanguage())
             ->addLanguage(new SqlLanguage())
             ->addLanguage(new XmlLanguage())
             ->addLanguage(new YamlLanguage())
diff --git a/src/Languages/Python/Patterns/PyArgumentPattern.php b/src/Languages/Python/Patterns/PyArgumentPattern.php
new file mode 100644
index 0000000..fd5e1e4
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyArgumentPattern.php
@@ -0,0 +1,24 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+final readonly class PyArgumentPattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return '(?<match>\w+)(s*=)';
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::VARIABLE;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyBooleanPattern.php b/src/Languages/Python/Patterns/PyBooleanPattern.php
new file mode 100644
index 0000000..dd7f83d
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyBooleanPattern.php
@@ -0,0 +1,24 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+final readonly class PyBooleanPattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return '\b(?<match>(?:False|None|True))\b';
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::TYPE;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyBuiltinPattern.php b/src/Languages/Python/Patterns/PyBuiltinPattern.php
new file mode 100644
index 0000000..60275bb
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyBuiltinPattern.php
@@ -0,0 +1,38 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+final readonly class PyBuiltinPattern implements Pattern
+{
+    use IsPattern;
+
+    public function __construct(private array $builtinFunctions = [
+        '__import__', 'abs', 'aiter', 'all', 'any', 'anext', 'ascii', 'bin', 'bool', 
+        'breakpoint', 'bytearray', 'bytes', 'callable', 'chr', 'classmethod', 'compile', 
+        'complex', 'delattr', 'dict', 'dir', 'divmod', 'enumerate', 'eval', 'exec', 
+        'filter', 'float', 'format', 'frozenset', 'getattr', 'globals', 'hasattr', 
+        'hash', 'help', 'hex', 'id', 'input', 'int', 'isinstance', 'issubclass', 'iter', 
+        'len', 'list', 'locals', 'map', 'max', 'memoryview', 'min', 'next', 'object', 
+        'oct', 'open', 'ord', 'pow', 'print', 'property', 'range', 'repr', 'reversed', 
+        'round', 'set', 'setattr', 'slice', 'sorted', 'staticmethod', 'str', 'sum', 
+        'super', 'tuple', 'type', 'vars', 'zip',
+    ]) {}
+
+    public function getPattern(): string
+    {
+        $builtinFunctions = implode('|', $this->builtinFunctions);
+
+        return "\b(?<match>(?:{$builtinFunctions}))\b";
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::TYPE;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyClassNamePattern.php b/src/Languages/Python/Patterns/PyClassNamePattern.php
new file mode 100644
index 0000000..b667051
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyClassNamePattern.php
@@ -0,0 +1,27 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\PatternTest;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+#[PatternTest(input: 'class MyClass:', output: 'MyClass')]
+#[PatternTest(input: 'class HisClass(MyClass):', output: 'HisClass')]
+final readonly class PyClassNamePattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return '\bclass\s+(?<match>\w*)(?=[\s*\:(])';
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::PROPERTY;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyCommentPattern.php b/src/Languages/Python/Patterns/PyCommentPattern.php
new file mode 100644
index 0000000..bb96a8d
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyCommentPattern.php
@@ -0,0 +1,24 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+final readonly class PyCommentPattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return '(^|[^\\\\])(?<match>#.*)';
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::COMMENT;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyDecoratorPattern.php b/src/Languages/Python/Patterns/PyDecoratorPattern.php
new file mode 100644
index 0000000..7989504
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyDecoratorPattern.php
@@ -0,0 +1,27 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\PatternTest;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+#[PatternTest(input: '@decorator', output: '@decorator')]
+#[PatternTest(input: '@decorator.chained', output: '@decorator.chained')]
+final readonly class PyDecoratorPattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return '(^|\n)\s*(?<match>@\s*\w*(?:\.\w+)*)';
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::PROPERTY;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyFunctionPattern.php b/src/Languages/Python/Patterns/PyFunctionPattern.php
new file mode 100644
index 0000000..cc92274
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyFunctionPattern.php
@@ -0,0 +1,26 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\PatternTest;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+#[PatternTest(input: 'def fibonacci(n)', output: 'fibonacci')]
+final readonly class PyFunctionPattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return '\bdef\s+(?<match>\w*)(?=\s*\()';
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::PROPERTY;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyKeywordPattern.php b/src/Languages/Python/Patterns/PyKeywordPattern.php
new file mode 100644
index 0000000..c5a1554
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyKeywordPattern.php
@@ -0,0 +1,33 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+final readonly class PyKeywordPattern implements Pattern
+{
+    use IsPattern;
+
+    public function __construct(private array $keywords = [
+        'and', 'as', 'assert', 'async', 'await', 'break', 'class', 'continue',
+        'def', 'del', 'elif', 'else', 'except', 'finally', 'for', 'from', 'global',
+        'if', 'import', 'in', 'is', 'lambda', 'nonlocal', 'not', 'or', 'pass',
+        'raise', 'return', 'try', 'while', 'with', 'yield',
+    ]) {}
+
+    public function getPattern(): string
+    {
+        $keywords = implode('|', $this->keywords);
+
+        return "\b(?<match>(?:_(?=\s*:){$keywords}))\b";
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::KEYWORD;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyOperatorPattern.php b/src/Languages/Python/Patterns/PyOperatorPattern.php
new file mode 100644
index 0000000..9920a9d
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyOperatorPattern.php
@@ -0,0 +1,24 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+final readonly class PyOperatorPattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return "(?<match>([-+&%=]=?|!=|:=|>>=|<<=|\|=|\^=|\*\*?=?|\/\/?=?|<[<=]?|>[=>]?|[\|^~]))";
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::OPERATOR;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyTripleDoubleQuoteStringPattern.php b/src/Languages/Python/Patterns/PyTripleDoubleQuoteStringPattern.php
new file mode 100644
index 0000000..7f249b7
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyTripleDoubleQuoteStringPattern.php
@@ -0,0 +1,24 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+final readonly class PyTripleDoubleQuoteStringPattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return '/(?<match>"""(.|\n)*?""")/m';
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::VALUE;
+    }
+}
diff --git a/src/Languages/Python/Patterns/PyTripleSingleQuoteStringPattern.php b/src/Languages/Python/Patterns/PyTripleSingleQuoteStringPattern.php
new file mode 100644
index 0000000..7f12024
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyTripleSingleQuoteStringPattern.php
@@ -0,0 +1,24 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+final readonly class PyTripleSingleQuoteStringPattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return '/(?<match>\'\'\'(.|\n)*?\'\'\')/m';
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::VALUE;
+    }
+}
diff --git a/src/Languages/Python/PythonLanguage.php b/src/Languages/Python/PythonLanguage.php
new file mode 100644
index 0000000..6d13f1b
--- /dev/null
+++ b/src/Languages/Python/PythonLanguage.php
@@ -0,0 +1,74 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python;
+
+use Tempest\Highlight\Languages\Base\BaseLanguage;
+use Tempest\Highlight\Languages\Php\Patterns\DoubleQuoteValuePattern;
+use Tempest\Highlight\Languages\Php\Patterns\SingleQuoteValuePattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyArgumentPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyBooleanPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyBuiltinPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyClassNamePattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyCommentPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyDecoratorPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyFunctionPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyKeywordPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyOperatorPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyTripleDoubleQuoteStringPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyTripleSingleQuoteStringPattern;
+
+class PythonLanguage extends BaseLanguage
+{
+    public function getName(): string
+    {
+        return 'python';
+    }
+
+    public function getAliases(): array
+    {
+        return ['py'];
+    }
+
+    public function getInjections(): array
+    {
+        return [
+            ...parent::getInjections(),
+        ];
+    }
+
+    public function getPatterns(): array
+    {
+        return [
+            ...parent::getPatterns(),
+
+            // KEYWORDS
+            new PyKeywordPattern(),
+
+            // PROPERTY
+            new PyClassNamePattern(),
+            new PyDecoratorPattern(),
+            new PyFunctionPattern(),
+
+            // TYPES
+            new PyBooleanPattern(),
+            new PyBuiltinPattern(),
+
+            // COMMENTS
+            new PyCommentPattern(),
+
+            // OPERATORS
+            new PyOperatorPattern(),
+
+            // VALUES
+            new SingleQuoteValuePattern(),
+            new DoubleQuoteValuePattern(),
+            new PyTripleDoubleQuoteStringPattern(),
+            new PyTripleSingleQuoteStringPattern(),
+
+            // VARIABLES
+            new PyArgumentPattern(),
+        ];
+    }
+}

From 8a258aaec70e8ed2601a27fb35ae4db2a348f355 Mon Sep 17 00:00:00 2001
From: Bogdan Condorachi <bogdan.condorachi@gmail.com>
Date: Thu, 16 May 2024 00:09:29 +0300
Subject: [PATCH 2/5] Fix styling

---
 .../Python/Patterns/PyBuiltinPattern.php      | 20 ++++++++++---------
 .../Python/Patterns/PyKeywordPattern.php      |  4 +++-
 2 files changed, 14 insertions(+), 10 deletions(-)

diff --git a/src/Languages/Python/Patterns/PyBuiltinPattern.php b/src/Languages/Python/Patterns/PyBuiltinPattern.php
index 60275bb..061aa2c 100644
--- a/src/Languages/Python/Patterns/PyBuiltinPattern.php
+++ b/src/Languages/Python/Patterns/PyBuiltinPattern.php
@@ -13,16 +13,18 @@
     use IsPattern;
 
     public function __construct(private array $builtinFunctions = [
-        '__import__', 'abs', 'aiter', 'all', 'any', 'anext', 'ascii', 'bin', 'bool', 
-        'breakpoint', 'bytearray', 'bytes', 'callable', 'chr', 'classmethod', 'compile', 
-        'complex', 'delattr', 'dict', 'dir', 'divmod', 'enumerate', 'eval', 'exec', 
-        'filter', 'float', 'format', 'frozenset', 'getattr', 'globals', 'hasattr', 
-        'hash', 'help', 'hex', 'id', 'input', 'int', 'isinstance', 'issubclass', 'iter', 
-        'len', 'list', 'locals', 'map', 'max', 'memoryview', 'min', 'next', 'object', 
-        'oct', 'open', 'ord', 'pow', 'print', 'property', 'range', 'repr', 'reversed', 
-        'round', 'set', 'setattr', 'slice', 'sorted', 'staticmethod', 'str', 'sum', 
+        '__import__', 'abs', 'aiter', 'all', 'any', 'anext', 'ascii', 'bin', 'bool',
+        'breakpoint', 'bytearray', 'bytes', 'callable', 'chr', 'classmethod', 'compile',
+        'complex', 'delattr', 'dict', 'dir', 'divmod', 'enumerate', 'eval', 'exec',
+        'filter', 'float', 'format', 'frozenset', 'getattr', 'globals', 'hasattr',
+        'hash', 'help', 'hex', 'id', 'input', 'int', 'isinstance', 'issubclass', 'iter',
+        'len', 'list', 'locals', 'map', 'max', 'memoryview', 'min', 'next', 'object',
+        'oct', 'open', 'ord', 'pow', 'print', 'property', 'range', 'repr', 'reversed',
+        'round', 'set', 'setattr', 'slice', 'sorted', 'staticmethod', 'str', 'sum',
         'super', 'tuple', 'type', 'vars', 'zip',
-    ]) {}
+    ])
+    {
+    }
 
     public function getPattern(): string
     {
diff --git a/src/Languages/Python/Patterns/PyKeywordPattern.php b/src/Languages/Python/Patterns/PyKeywordPattern.php
index c5a1554..eb13e4c 100644
--- a/src/Languages/Python/Patterns/PyKeywordPattern.php
+++ b/src/Languages/Python/Patterns/PyKeywordPattern.php
@@ -17,7 +17,9 @@ public function __construct(private array $keywords = [
         'def', 'del', 'elif', 'else', 'except', 'finally', 'for', 'from', 'global',
         'if', 'import', 'in', 'is', 'lambda', 'nonlocal', 'not', 'or', 'pass',
         'raise', 'return', 'try', 'while', 'with', 'yield',
-    ]) {}
+    ])
+    {
+    }
 
     public function getPattern(): string
     {

From 6de46b156c078302978ad4f35d2761843345b5fd Mon Sep 17 00:00:00 2001
From: Bogdan Condorachi <bogdan.condorachi@gmail.com>
Date: Fri, 17 May 2024 01:04:19 +0300
Subject: [PATCH 3/5] Improve python argument match

---
 src/Languages/Python/Patterns/PyArgumentPattern.php | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/Languages/Python/Patterns/PyArgumentPattern.php b/src/Languages/Python/Patterns/PyArgumentPattern.php
index fd5e1e4..05f74ac 100644
--- a/src/Languages/Python/Patterns/PyArgumentPattern.php
+++ b/src/Languages/Python/Patterns/PyArgumentPattern.php
@@ -14,7 +14,7 @@
 
     public function getPattern(): string
     {
-        return '(?<match>\w+)(s*=)';
+        return '(?<=,|\()\s*(?<match>\w+)s*=';
     }
 
     public function getTokenType(): TokenTypeEnum

From 116da0e9a99568f55c23d4ba9a2b5b8214e2d0cd Mon Sep 17 00:00:00 2001
From: Bogdan Condorachi <bogdan.condorachi@gmail.com>
Date: Fri, 17 May 2024 01:05:27 +0300
Subject: [PATCH 4/5] Add python unit test

---
 tests/Languages/Python/PythonLanguageTest.php | 58 +++++++++++++++++++
 1 file changed, 58 insertions(+)
 create mode 100644 tests/Languages/Python/PythonLanguageTest.php

diff --git a/tests/Languages/Python/PythonLanguageTest.php b/tests/Languages/Python/PythonLanguageTest.php
new file mode 100644
index 0000000..61ceda5
--- /dev/null
+++ b/tests/Languages/Python/PythonLanguageTest.php
@@ -0,0 +1,58 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Tests\Languages\Python;
+
+use PHPUnit\Framework\Attributes\DataProvider;
+use PHPUnit\Framework\TestCase;
+use Tempest\Highlight\Highlighter;
+
+class PythonLanguageTest extends TestCase
+{
+    #[DataProvider('data')]
+    public function test_highlight(string $content, string $expected): void
+    {
+        $highlighter = new Highlighter();
+
+        $this->assertSame(
+            $expected,
+            $highlighter->parse($content, 'python'),
+        );
+
+        $this->assertSame(
+            $expected,
+            $highlighter->parse($content, 'py'),
+        );
+    }
+
+    public static function data(): array
+    {
+        return [
+            [<<<TXT
+def fib(n):    # write Fibonacci series up to n
+    """Print a Fibonacci series up to n."""
+    a, b = 0, 1
+    while a < n:
+        print(a, end=' ')
+        a, b = b, a+b
+    print()
+
+# Now call the function we just defined:
+fib(2000)
+TXT,
+            <<<TXT
+<span class="hl-keyword">def</span> <span class="hl-property">fib</span>(n):    <span class="hl-comment"># write Fibonacci series up to n</span>
+    <span class="hl-value">&quot;&quot;&quot;Print a Fibonacci series up to n.&quot;&quot;&quot;</span>
+    a, b <span class="hl-operator">=</span> 0, 1
+    <span class="hl-keyword">while</span> a <span class="hl-operator">&lt;</span> n:
+        <span class="hl-type">print</span>(a, <span class="hl-variable">end</span><span class="hl-operator">=</span>'<span class="hl-value"> </span>')
+        a, b <span class="hl-operator">=</span> b, a<span class="hl-operator">+</span>b
+    <span class="hl-type">print</span>()
+
+<span class="hl-comment"># Now call the function we just defined:</span>
+fib(2000)
+TXT],
+        ];
+    }
+}

From dbde1c2ed6e9b63e43fe2837684ba59241d13fce Mon Sep 17 00:00:00 2001
From: Bogdan Condorachi <bogdan.condorachi@gmail.com>
Date: Thu, 23 May 2024 14:07:35 +0300
Subject: [PATCH 5/5] Add python #123 support

---
 .../Python/Patterns/PyBooleanPattern.php      |  2 +-
 .../Python/Patterns/PyNumberPattern.php       | 24 +++++++++++++++++++
 src/Languages/Python/PythonLanguage.php       |  8 ++++++-
 tests/Languages/Python/PythonLanguageTest.php |  4 ++--
 4 files changed, 34 insertions(+), 4 deletions(-)
 create mode 100644 src/Languages/Python/Patterns/PyNumberPattern.php

diff --git a/src/Languages/Python/Patterns/PyBooleanPattern.php b/src/Languages/Python/Patterns/PyBooleanPattern.php
index dd7f83d..e42d3af 100644
--- a/src/Languages/Python/Patterns/PyBooleanPattern.php
+++ b/src/Languages/Python/Patterns/PyBooleanPattern.php
@@ -19,6 +19,6 @@ public function getPattern(): string
 
     public function getTokenType(): TokenTypeEnum
     {
-        return TokenTypeEnum::TYPE;
+        return TokenTypeEnum::BOOLEAN;
     }
 }
diff --git a/src/Languages/Python/Patterns/PyNumberPattern.php b/src/Languages/Python/Patterns/PyNumberPattern.php
new file mode 100644
index 0000000..b8e14aa
--- /dev/null
+++ b/src/Languages/Python/Patterns/PyNumberPattern.php
@@ -0,0 +1,24 @@
+<?php
+
+declare(strict_types=1);
+
+namespace Tempest\Highlight\Languages\Python\Patterns;
+
+use Tempest\Highlight\IsPattern;
+use Tempest\Highlight\Pattern;
+use Tempest\Highlight\Tokens\TokenTypeEnum;
+
+final readonly class PyNumberPattern implements Pattern
+{
+    use IsPattern;
+
+    public function getPattern(): string
+    {
+        return '(?<match>\b0(?:[bB](?:_?[01])+|[oO](?:_?[0-7])+|[xX](?:_?[a-fA-F0-9])+)\b|(?:\b\d+(?:_\d+)*(?:\.(?:\d+(?:_\d+)*)?)?|\B\.\d+(?:_\d+)*)(?:[eE][+-]?\d+(?:_\d+)*)?j?(?!\w))';
+    }
+
+    public function getTokenType(): TokenTypeEnum
+    {
+        return TokenTypeEnum::NUMBER;
+    }
+}
diff --git a/src/Languages/Python/PythonLanguage.php b/src/Languages/Python/PythonLanguage.php
index 6d13f1b..9b0ced0 100644
--- a/src/Languages/Python/PythonLanguage.php
+++ b/src/Languages/Python/PythonLanguage.php
@@ -15,6 +15,7 @@
 use Tempest\Highlight\Languages\Python\Patterns\PyDecoratorPattern;
 use Tempest\Highlight\Languages\Python\Patterns\PyFunctionPattern;
 use Tempest\Highlight\Languages\Python\Patterns\PyKeywordPattern;
+use Tempest\Highlight\Languages\Python\Patterns\PyNumberPattern;
 use Tempest\Highlight\Languages\Python\Patterns\PyOperatorPattern;
 use Tempest\Highlight\Languages\Python\Patterns\PyTripleDoubleQuoteStringPattern;
 use Tempest\Highlight\Languages\Python\Patterns\PyTripleSingleQuoteStringPattern;
@@ -52,12 +53,17 @@ public function getPatterns(): array
             new PyFunctionPattern(),
 
             // TYPES
-            new PyBooleanPattern(),
             new PyBuiltinPattern(),
 
             // COMMENTS
             new PyCommentPattern(),
 
+            // NUMBERS
+            new PyNumberPattern(),
+
+            // BOOLEANS
+            new PyBooleanPattern(),
+
             // OPERATORS
             new PyOperatorPattern(),
 
diff --git a/tests/Languages/Python/PythonLanguageTest.php b/tests/Languages/Python/PythonLanguageTest.php
index 61ceda5..72e861a 100644
--- a/tests/Languages/Python/PythonLanguageTest.php
+++ b/tests/Languages/Python/PythonLanguageTest.php
@@ -44,14 +44,14 @@ public static function data(): array
             <<<TXT
 <span class="hl-keyword">def</span> <span class="hl-property">fib</span>(n):    <span class="hl-comment"># write Fibonacci series up to n</span>
     <span class="hl-value">&quot;&quot;&quot;Print a Fibonacci series up to n.&quot;&quot;&quot;</span>
-    a, b <span class="hl-operator">=</span> 0, 1
+    a, b <span class="hl-operator">=</span> <span class="hl-number">0</span>, <span class="hl-number">1</span>
     <span class="hl-keyword">while</span> a <span class="hl-operator">&lt;</span> n:
         <span class="hl-type">print</span>(a, <span class="hl-variable">end</span><span class="hl-operator">=</span>'<span class="hl-value"> </span>')
         a, b <span class="hl-operator">=</span> b, a<span class="hl-operator">+</span>b
     <span class="hl-type">print</span>()
 
 <span class="hl-comment"># Now call the function we just defined:</span>
-fib(2000)
+fib(<span class="hl-number">2000</span>)
 TXT],
         ];
     }