From 9f449d94e03c7c7287dac9b1489f637e16e754ac Mon Sep 17 00:00:00 2001
From: Pablo Galindo Salgado <pablogsal@gmail.com>
Date: Sat, 11 Jul 2026 16:53:21 +0100
Subject: [PATCH] gh-153568: Skip newline translation for source without
 carriage returns

The tokenizer copied every input byte by byte to normalize newlines,
but source without a carriage return needs no translation and can be
copied verbatim.
---
 ...7-11-16-53-21.gh-issue-153568.newlines.rst |  2 ++
 Parser/tokenizer/decoder.c                    | 21 ++++++++++++-------
 2 files changed, 16 insertions(+), 7 deletions(-)
 create mode 100644 Misc/NEWS.d/next/Core_and_Builtins/2026-07-11-16-53-21.gh-issue-153568.newlines.rst

diff --git a/Misc/NEWS.d/next/Core_and_Builtins/2026-07-11-16-53-21.gh-issue-153568.newlines.rst b/Misc/NEWS.d/next/Core_and_Builtins/2026-07-11-16-53-21.gh-issue-153568.newlines.rst
new file mode 100644
index 000000000000000..deaa82c96fd3ac4
--- /dev/null
+++ b/Misc/NEWS.d/next/Core_and_Builtins/2026-07-11-16-53-21.gh-issue-153568.newlines.rst
@@ -0,0 +1,2 @@
+Speed up the tokenizer by copying source text verbatim when it contains no
+carriage returns.
diff --git a/Parser/tokenizer/decoder.c b/Parser/tokenizer/decoder.c
index 5b588572f1fb7e8..ab8ae393d03032d 100644
--- a/Parser/tokenizer/decoder.c
+++ b/Parser/tokenizer/decoder.c
@@ -87,15 +87,22 @@ _PyTok_NormalizeNewlines(const char *data, Py_ssize_t len, int preserve_crlf,
         return NULL;
     }
     Py_ssize_t write = 0;
-    for (Py_ssize_t read = 0; read < len; read++) {
-        char c = data[read];
-        if (!preserve_crlf && c == '\r') {
-            if (read + 1 < len && data[read + 1] == '\n') {
-                read++;
+    if (memchr(data, '\r', len) == NULL) {
+        // No carriage returns: nothing to translate, copy verbatim.
+        memcpy(result, data, len);
+        write = len;
+    }
+    else {
+        for (Py_ssize_t read = 0; read < len; read++) {
+            char c = data[read];
+            if (!preserve_crlf && c == '\r') {
+                if (read + 1 < len && data[read + 1] == '\n') {
+                    read++;
+                }
+                c = '\n';
             }
-            c = '\n';
+            result[write++] = c;
         }
-        result[write++] = c;
     }
     int implicit = add_final_newline && write > 0 && result[write - 1] != '\n';
     if (implicit) {