[PHP 8.0] Add match expressions (#672)

RFC: https://wiki.php.net/rfc/match_expression_v2 Upstream implementation: php/php-src#5371 Closes #671.
nikic · Jul 15, 2020 · 69c5d48 · 69c5d48
1 parent 6ec527b
commit 69c5d48
Show file tree

Hide file tree

Showing 18 changed files with 2,581 additions and 2,048 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,7 +1,10 @@
 Version 4.6.1-dev
 -----------------
 
-Nothing yet.
+### Added
+
+* [PHP 8.0] Added support for match expressions. These are represented using a new `Expr\Match_`
+  containing `MatchArm`s.
 
 Version 4.6.0 (2020-07-02)
 --------------------------

diff --git a/grammar/php5.y b/grammar/php5.y
@@ -28,6 +28,7 @@ reserved_non_modifiers:
     | T_FUNCTION | T_CONST | T_RETURN | T_PRINT | T_YIELD | T_LIST | T_SWITCH | T_ENDSWITCH | T_CASE | T_DEFAULT
     | T_BREAK | T_ARRAY | T_CALLABLE | T_EXTENDS | T_IMPLEMENTS | T_NAMESPACE | T_TRAIT | T_INTERFACE | T_CLASS
     | T_CLASS_C | T_TRAIT_C | T_FUNC_C | T_METHOD_C | T_LINE | T_FILE | T_DIR | T_NS_C | T_HALT_COMPILER | T_FN
+    | T_MATCH
 ;
 
 semi_reserved:

diff --git a/grammar/php7.y b/grammar/php7.y
@@ -28,6 +28,7 @@ reserved_non_modifiers:
     | T_FUNCTION | T_CONST | T_RETURN | T_PRINT | T_YIELD | T_LIST | T_SWITCH | T_ENDSWITCH | T_CASE | T_DEFAULT
     | T_BREAK | T_ARRAY | T_CALLABLE | T_EXTENDS | T_IMPLEMENTS | T_NAMESPACE | T_TRAIT | T_INTERFACE | T_CLASS
     | T_CLASS_C | T_TRAIT_C | T_FUNC_C | T_METHOD_C | T_LINE | T_FILE | T_DIR | T_NS_C | T_HALT_COMPILER | T_FN
+    | T_MATCH
 ;
 
 semi_reserved:
@@ -399,6 +400,25 @@ case_separator:
     | ';'
 ;
 
+match:
+      T_MATCH '(' expr ')' '{' match_arm_list '}'           { $$ = Expr\Match_[$3, $6]; }
+;
+
+match_arm_list:
+      /* empty */                                           { $$ = []; }
+    | non_empty_match_arm_list optional_comma               { $$ = $1; }
+;
+
+non_empty_match_arm_list:
+      match_arm                                             { init($1); }
+    | non_empty_match_arm_list ',' match_arm                { push($1, $3); }
+;
+
+match_arm:
+      expr_list_allow_comma T_DOUBLE_ARROW expr             { $$ = Node\MatchArm[$1, $3]; }
+    | T_DEFAULT optional_comma T_DOUBLE_ARROW expr          { $$ = Node\MatchArm[null, $4]; }
+;
+
 while_statement:
       statement                                             { $$ = toArray($1); }
     | ':' inner_statement_list T_ENDWHILE ';'               { $$ = $2; }
@@ -666,6 +686,7 @@ expr:
     | variable '=' expr                                     { $$ = Expr\Assign[$1, $3]; }
     | variable '=' '&' variable                             { $$ = Expr\AssignRef[$1, $4]; }
     | new_expr                                              { $$ = $1; }
+    | match                                                 { $$ = $1; }
     | T_CLONE expr                                          { $$ = Expr\Clone_[$2]; }
     | variable T_PLUS_EQUAL expr                            { $$ = Expr\AssignOp\Plus      [$1, $3]; }
     | variable T_MINUS_EQUAL expr                           { $$ = Expr\AssignOp\Minus     [$1, $3]; }
@@ -967,14 +988,14 @@ new_variable:
 
 member_name:
       identifier_ex                                         { $$ = $1; }
-    | '{' expr '}'	                                        { $$ = $2; }
-    | simple_variable	                                    { $$ = Expr\Variable[$1]; }
+    | '{' expr '}'                                          { $$ = $2; }
+    | simple_variable                                       { $$ = Expr\Variable[$1]; }
 ;
 
 property_name:
       identifier                                            { $$ = $1; }
-    | '{' expr '}'	                                        { $$ = $2; }
-    | simple_variable	                                    { $$ = Expr\Variable[$1]; }
+    | '{' expr '}'                                          { $$ = $2; }
+    | simple_variable                                       { $$ = Expr\Variable[$1]; }
     | error                                                 { $$ = Expr\Error[]; $this->errorState = 2; }
 ;
 

diff --git a/grammar/tokens.y b/grammar/tokens.y
@@ -57,6 +57,7 @@
 %token T_ENDDECLARE
 %token T_AS
 %token T_SWITCH
+%token T_MATCH
 %token T_ENDSWITCH
 %token T_CASE
 %token T_DEFAULT

diff --git a/lib/PhpParser/Lexer/Emulative.php b/lib/PhpParser/Lexer/Emulative.php
@@ -7,6 +7,7 @@
 use PhpParser\Lexer;
 use PhpParser\Lexer\TokenEmulator\CoaleseEqualTokenEmulator;
 use PhpParser\Lexer\TokenEmulator\FnTokenEmulator;
+use PhpParser\Lexer\TokenEmulator\MatchTokenEmulator;
 use PhpParser\Lexer\TokenEmulator\NumericLiteralSeparatorEmulator;
 use PhpParser\Lexer\TokenEmulator\TokenEmulatorInterface;
 use PhpParser\Parser\Tokens;
@@ -15,9 +16,11 @@ class Emulative extends Lexer
 {
     const PHP_7_3 = '7.3.0dev';
     const PHP_7_4 = '7.4.0dev';
+    const PHP_8_0 = '8.0.0dev';
 
     const T_COALESCE_EQUAL = 1007;
     const T_FN = 1008;
+    const T_MATCH = 1009;
 
     const FLEXIBLE_DOC_STRING_REGEX = <<<'REGEX'
 /<<<[ \t]*(['"]?)([a-zA-Z_\x80-\xff][a-zA-Z0-9_\x80-\xff]*)\1\r?\n
@@ -39,11 +42,13 @@ public function __construct(array $options = [])
         parent::__construct($options);
 
         $this->tokenEmulators[] = new FnTokenEmulator();
+        $this->tokenEmulators[] = new MatchTokenEmulator();
         $this->tokenEmulators[] = new CoaleseEqualTokenEmulator();
         $this->tokenEmulators[] = new NumericLiteralSeparatorEmulator();
 
         $this->tokenMap[self::T_COALESCE_EQUAL] = Tokens::T_COALESCE_EQUAL;
         $this->tokenMap[self::T_FN] = Tokens::T_FN;
+        $this->tokenMap[self::T_MATCH] = Tokens::T_MATCH;
     }
 
     public function startLexing(string $code, ErrorHandler $errorHandler = null) {

diff --git a/lib/PhpParser/Lexer/TokenEmulator/FnTokenEmulator.php b/lib/PhpParser/Lexer/TokenEmulator/FnTokenEmulator.php
@@ -18,10 +18,8 @@ public function isEmulationNeeded(string $code) : bool
 
     public function emulate(string $code, array $tokens): array
     {
-        // We need to manually iterate and manage a count because we'll change
-        // the tokens array on the way
         foreach ($tokens as $i => $token) {
-            if ($token[0] === T_STRING && $token[1] === 'fn') {
+            if ($token[0] === T_STRING && strtolower($token[1]) === 'fn') {
                 $previousNonSpaceToken = $this->getPreviousNonSpaceToken($tokens, $i);
                 if ($previousNonSpaceToken !== null && $previousNonSpaceToken[0] === T_OBJECT_OPERATOR) {
                     continue;

diff --git a/lib/PhpParser/Lexer/TokenEmulator/MatchTokenEmulator.php b/lib/PhpParser/Lexer/TokenEmulator/MatchTokenEmulator.php
@@ -0,0 +1,51 @@
+<?php declare(strict_types=1);
+
+namespace PhpParser\Lexer\TokenEmulator;
+
+use PhpParser\Lexer\Emulative;
+
+final class MatchTokenEmulator implements TokenEmulatorInterface
+{
+    public function isEmulationNeeded(string $code) : bool
+    {
+        // skip version where this is supported
+        if (version_compare(\PHP_VERSION, Emulative::PHP_8_0, '>=')) {
+            return false;
+        }
+
+        return strpos($code, 'match') !== false;
+    }
+
+    public function emulate(string $code, array $tokens): array
+    {
+        foreach ($tokens as $i => $token) {
+            if ($token[0] === T_STRING && strtolower($token[1]) === 'match') {
+                $previousNonSpaceToken = $this->getPreviousNonSpaceToken($tokens, $i);
+                if ($previousNonSpaceToken !== null && $previousNonSpaceToken[0] === T_OBJECT_OPERATOR) {
+                    continue;
+                }
+
+                $tokens[$i][0] = Emulative::T_MATCH;
+            }
+        }
+
+        return $tokens;
+    }
+
+    /**
+     * @param mixed[] $tokens
+     * @return mixed[]|null
+     */
+    private function getPreviousNonSpaceToken(array $tokens, int $start)
+    {
+        for ($i = $start - 1; $i >= 0; --$i) {
+            if ($tokens[$i][0] === T_WHITESPACE) {
+                continue;
+            }
+
+            return $tokens[$i];
+        }
+
+        return null;
+    }
+}
diff --git a/lib/PhpParser/Node/Expr/Match_.php b/lib/PhpParser/Node/Expr/Match_.php
@@ -0,0 +1,31 @@
+<?php declare(strict_types=1);
+
+namespace PhpParser\Node\Expr;
+
+use PhpParser\Node;
+use PhpParser\Node\MatchArm;
+
+class Match_ extends Node\Expr
+{
+    /** @var Node\Expr */
+    public $cond;
+    /** @var MatchArm[] */
+    public $arms;
+
+    /**
+     * @param MatchArm[] $arms
+     */
+    public function __construct(Node\Expr $cond, array $arms = [], array $attributes = []) {
+        $this->attributes = $attributes;
+        $this->cond = $cond;
+        $this->arms = $arms;
+    }
+
+    public function getSubNodeNames() : array {
+        return ['cond', 'arms'];
+    }
+
+    public function getType() : string {
+        return 'Expr_Match';
+    }
+}
diff --git a/lib/PhpParser/Node/MatchArm.php b/lib/PhpParser/Node/MatchArm.php
@@ -0,0 +1,31 @@
+<?php declare(strict_types=1);
+
+namespace PhpParser\Node;
+
+use PhpParser\Node;
+use PhpParser\NodeAbstract;
+
+class MatchArm extends NodeAbstract
+{
+    /** @var null|Node\Expr[] */
+    public $conds;
+    /** @var Node\Expr */
+    public $body;
+
+    /**
+     * @param null|Node\Expr[] $conds
+     */
+    public function __construct($conds, Node\Expr $body, array $attributes = []) {
+        $this->conds = $conds;
+        $this->body = $body;
+        $this->attributes = $attributes;
+    }
+
+    public function getSubNodeNames() : array {
+        return ['conds', 'body'];
+    }
+
+    public function getType() : string {
+        return 'MatchArm';
+    }
+}