From 86fcb2a1ad0d99d4985c216a9149c795021b397d Mon Sep 17 00:00:00 2001
From: jenkins <jenkins@local>
Date: Tue, 28 Apr 2026 18:43:54 +0530
Subject: [PATCH] Add support for C-style block comments (/* */) in tokenizer

---
 .gitignore           |  6 +++++
 Parser/lexer/lexer.c | 62 ++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 68 insertions(+)

diff --git a/.gitignore b/.gitignore
index 3bf5187d531c23..ad40f75b72b62b 100644
--- a/.gitignore
+++ b/.gitignore
@@ -179,3 +179,9 @@ CLAUDE.local.md
 #### main branch only stuff below this line, things to backport go above. ####
 # main branch only: ABI files are not checked/maintained.
 Doc/data/python*.abi
+
+
+
+
+
+
diff --git a/Parser/lexer/lexer.c b/Parser/lexer/lexer.c
index 7f25afec302c22..9cb7d73efa744d 100644
--- a/Parser/lexer/lexer.c
+++ b/Parser/lexer/lexer.c
@@ -542,11 +542,39 @@ tok_get_normal_mode(struct tok_state *tok, tokenizer_mode* current_tok, struct t
             else if (c == EOF && PyErr_Occurred()) {
                 return MAKE_TOKEN(ERRORTOKEN);
             }
+            else if (c == '/') {
+                int c2 = tok_nextc(tok);
+                if (c2 == '*') {
+                    /* Consume block comment as part of indentation/whitespace */
+                    int prev = 0;
+                    int ch;
+                    while (1) {
+                        ch = tok_nextc(tok);
+                        if (ch == EOF) {
+                            tok->done = E_EOFS;
+                            return MAKE_TOKEN(ERRORTOKEN);
+                        }
+                        if (ch == '\n') {
+                            tok->lineno++;
+                        }
+                        if (prev == '*' && ch == '/') {
+                            break;
+                        }
+                        prev = ch;
+                    }
+                    /* Continue looking for more whitespace/comments */
+                    continue;
+                } else {
+                    tok_backup(tok, c2);
+                    break;
+                }
+            }
             else {
                 break;
             }
         }
         tok_backup(tok, c);
+
         if (c == '#' || c == '\n' || c == '\r') {
             /* Lines with only whitespace and/or comments
                shouldn't affect the indentation and are
@@ -726,6 +754,40 @@ tok_get_normal_mode(struct tok_state *tok, tokenizer_mode* current_tok, struct t
         }
     }
 
+    /* Skip C-style block comment (slash-star ... star-slash) */
+    if (c == '/') {
+        int c2 = tok_nextc(tok);
+        if (c2 == '*') {
+            /* Consume everything until the closing star-slash sequence */
+            int prev = 0;
+            int ch;
+            while (1) {
+                ch = tok_nextc(tok);
+                if (ch == EOF) {
+                    tok->done = E_EOFS;
+                    _PyTokenizer_syntaxerror(tok,
+                        "unterminated C-style block comment (/* ... */) "
+                        "— missing closing */");
+                    return MAKE_TOKEN(ERRORTOKEN);
+                }
+                if (ch == '\n') {
+                    tok->lineno++;
+                    tok->atbol = 1;
+                }
+                if (prev == '*' && ch == '/') {
+                    break;  /* found closing star-slash */
+                }
+                prev = ch;
+            }
+            /* Restart token loop: treat block comment as whitespace */
+            goto again;
+        }
+        else {
+            /* Plain '/' operator: put c2 back, fall through to operator handling */
+            tok_backup(tok, c2);
+        }
+    }
+
     if (tok->done == E_INTERACT_STOP) {
         return MAKE_TOKEN(ENDMARKER);
     }