From 3de01fa77c80d9b98c66881b4311764e9b4c3269 Mon Sep 17 00:00:00 2001
From: Arnesh Banerjee <linkrinku13@gmail.com>
Date: Thu, 30 Jul 2026 14:26:23 +0530
Subject: [PATCH] gh-154711: Strip comments after escaped quotes in f-string
 and t-string debug text

When a replacement field ended with a string that had an escaped quote and
was followed by a comment, the comment leaked into the output. It showed up
in debug f-string results and in t-string metadata.

The comment detection loop in _PyLexer_set_ftstring_expr skips escaped
characters, but the comment stripping loop did not. So an escaped quote left
the scanner thinking it was still inside a string, and the comment was never
removed.

This makes the stripping loop skip escaped characters the same way the
detection loop does, so the comment is removed as expected. Tests are added
for f-strings and t-strings.
---
 Lib/test/test_fstring.py                            |  5 +++++
 Lib/test/test_tstring.py                            | 12 ++++++++++++
 .../2026-07-30-14-40-00.gh-issue-154711.nUxYXc.rst  |  3 +++
 Parser/lexer/string.c                               | 13 +++++++++++++
 4 files changed, 33 insertions(+)
 create mode 100644 Misc/NEWS.d/next/Core_and_Builtins/2026-07-30-14-40-00.gh-issue-154711.nUxYXc.rst

diff --git a/Lib/test/test_fstring.py b/Lib/test/test_fstring.py
index 45890f8fb35dcb7..60675099f90d5b5 100644
--- a/Lib/test/test_fstring.py
+++ b/Lib/test/test_fstring.py
@@ -1678,6 +1678,11 @@ def __repr__(self):
         self.assertEqual(f'{" # nooo "=}', '" # nooo "=\' # nooo \'')
         self.assertEqual(f'{" \" # nooo \" "=}', '" \\" # nooo \\" "=\' " # nooo " \'')
 
+        # A real comment after a string that ends with an escaped quote must
+        # still be stripped (gh-154711).
+        self.assertEqual(f"{'\'' = # comment
+}", "'\\'' = \n" + repr("'"))
+
         self.assertEqual(f'{ # some comment goes here
   """hello"""=}',  ' \n  """hello"""=\'hello\'')
         self.assertEqual(f'{"""# this is not a comment
diff --git a/Lib/test/test_tstring.py b/Lib/test/test_tstring.py
index 74653c77c55de17..9e8d5fdefd14265 100644
--- a/Lib/test/test_tstring.py
+++ b/Lib/test/test_tstring.py
@@ -140,6 +140,18 @@ def test_debug_specifier(self):
         )
         self.assertEqual(fstring(t), "Value: value = 42")
 
+    def test_comment_after_escaped_quote(self):
+        # A comment after a string that ends with an escaped quote must be
+        # stripped from the interpolation expression (gh-154711).
+        t = t"{'a\'b' # comment
+}"
+        self.assertEqual(t.interpolations[0].expression, "'a\\'b'")
+
+        t = t"{'a\'b' = # comment
+}"
+        self.assertEqual(t.interpolations[0].expression, "'a\\'b'")
+        self.assertEqual(t.strings, ("'a\\'b' = \n", ""))
+
     def test_raw_tstrings(self):
         path = r"C:\Users"
         t = rt"{path}\Documents"
diff --git a/Misc/NEWS.d/next/Core_and_Builtins/2026-07-30-14-40-00.gh-issue-154711.nUxYXc.rst b/Misc/NEWS.d/next/Core_and_Builtins/2026-07-30-14-40-00.gh-issue-154711.nUxYXc.rst
new file mode 100644
index 000000000000000..e5f90663ce7583a
--- /dev/null
+++ b/Misc/NEWS.d/next/Core_and_Builtins/2026-07-30-14-40-00.gh-issue-154711.nUxYXc.rst
@@ -0,0 +1,3 @@
+Fix a bug where a comment placed after a string that ends with an escaped
+quote could leak into the output of a debug f-string and into t-string
+metadata. Now the comment is stripped as expected.
diff --git a/Parser/lexer/string.c b/Parser/lexer/string.c
index e546954ba5c0d82..006764201f62305 100644
--- a/Parser/lexer/string.c
+++ b/Parser/lexer/string.c
@@ -74,6 +74,19 @@ _PyLexer_set_ftstring_expr(struct tok_state* tok, struct token *token, char c) {
         while (i < tok_mode->last_expr_size - tok_mode->last_expr_end) {
             char ch = tok_mode->last_expr_buffer[i];
 
+            // Copy escaped characters as-is. This keeps an escaped quote from
+            // flipping the in_string state, which would otherwise stop a real
+            // comment from being detected (see the detection loop above).
+            if (ch == '\\') {
+                result[j++] = ch;
+                i++;
+                if (i < tok_mode->last_expr_size - tok_mode->last_expr_end) {
+                    result[j++] = tok_mode->last_expr_buffer[i];
+                    i++;
+                }
+                continue;
+            }
+
             // Handle string quotes
             if (ch == '"' || ch == '\'') {
                 // See comment above to understand this part