From aa58a696655593b0a012166cfbcf34fafac6e945 Mon Sep 17 00:00:00 2001
From: Nicolas Grekas <nicolas.grekas@gmail.com>
Date: Thu, 20 Jan 2022 11:57:37 +0100
Subject: [PATCH] [Console] fix parsing escaped chars in StringInput

---
 Input/StringInput.php           | 23 +++++++++++++++++++----
 Tests/Input/StringInputTest.php |  1 +
 2 files changed, 20 insertions(+), 4 deletions(-)

diff --git a/Input/StringInput.php b/Input/StringInput.php
index eb5c07fdd..76f1d5030 100644
--- a/Input/StringInput.php
+++ b/Input/StringInput.php
@@ -24,7 +24,7 @@
  */
 class StringInput extends ArgvInput
 {
-    public const REGEX_STRING = '([^\s]+?)(?:\s|(?<!\\\\)"|(?<!\\\\)\'|$)';
+    public const REGEX_STRING = '([^\s\\\\]+?)';
     public const REGEX_QUOTED_STRING = '(?:"([^"\\\\]*(?:\\\\.[^"\\\\]*)*)"|\'([^\'\\\\]*(?:\\\\.[^\'\\\\]*)*)\')';
 
     /**
@@ -47,14 +47,25 @@ private function tokenize(string $input): array
         $tokens = [];
         $length = \strlen($input);
         $cursor = 0;
+        $token = null;
         while ($cursor < $length) {
+            if ('\\' === $input[$cursor]) {
+                $token .= $input[++$cursor] ?? '';
+                ++$cursor;
+                continue;
+            }
+
             if (preg_match('/\s+/A', $input, $match, 0, $cursor)) {
+                if (null !== $token) {
+                    $tokens[] = $token;
+                    $token = null;
+                }
             } elseif (preg_match('/([^="\'\s]+?)(=?)('.self::REGEX_QUOTED_STRING.'+)/A', $input, $match, 0, $cursor)) {
-                $tokens[] = $match[1].$match[2].stripcslashes(str_replace(['"\'', '\'"', '\'\'', '""'], '', substr($match[3], 1, -1)));
+                $token .= $match[1].$match[2].stripcslashes(str_replace(['"\'', '\'"', '\'\'', '""'], '', substr($match[3], 1, -1)));
             } elseif (preg_match('/'.self::REGEX_QUOTED_STRING.'/A', $input, $match, 0, $cursor)) {
-                $tokens[] = stripcslashes(substr($match[0], 1, -1));
+                $token .= stripcslashes(substr($match[0], 1, -1));
             } elseif (preg_match('/'.self::REGEX_STRING.'/A', $input, $match, 0, $cursor)) {
-                $tokens[] = stripcslashes($match[1]);
+                $token .= $match[1];
             } else {
                 // should never happen
                 throw new InvalidArgumentException(sprintf('Unable to parse input near "... %s ...".', substr($input, $cursor, 10)));
@@ -63,6 +74,10 @@ private function tokenize(string $input): array
             $cursor += \strlen($match[0]);
         }
 
+        if (null !== $token) {
+            $tokens[] = $token;
+        }
+
         return $tokens;
     }
 }
diff --git a/Tests/Input/StringInputTest.php b/Tests/Input/StringInputTest.php
index f781b7ccf..2bd40dec9 100644
--- a/Tests/Input/StringInputTest.php
+++ b/Tests/Input/StringInputTest.php
@@ -71,6 +71,7 @@ public function getTokenizeData()
             ["--long-option='foo bar''another'", ['--long-option=foo baranother'], '->tokenize() parses long options with a value'],
             ["--long-option='foo bar'\"another\"", ['--long-option=foo baranother'], '->tokenize() parses long options with a value'],
             ['foo -a -ffoo --long bar', ['foo', '-a', '-ffoo', '--long', 'bar'], '->tokenize() parses when several arguments and options'],
+            ["--arg=\\\"'Jenny'\''s'\\\"", ["--arg=\"Jenny's\""], '->tokenize() parses quoted quotes'],
         ];
     }