gh-156047: Prioritize keyword typos on error line

Author: m-aciek

Lib/test/test_traceback.py

@@ -1827,6 +1827,8 @@ class TestKeywordTypoSuggestions(unittest.TestCase):
         ("function f():", "def"),
         ("func f():", "def"),
         ("void f():", "def"),
+        # gh-156047: Earlier names must not exhaust the candidate budget.
+        ("a=b=c=d=e=f=g=h=i=0\ndef fn():\n  retrun True", "return"),
     ]
 
     def test_keyword_suggestions_from_file(self):
@@ -1847,6 +1849,18 @@ def test_keyword_suggestions_from_command_string(self):
                 stderr_text = stderr.decode('utf-8')
                 self.assertIn(f"Did you mean '{expected_kw}'", stderr_text)
 
+    def test_keyword_suggestion_before_reported_error_line(self):
+        source = "def fn():\n  retrun True\n# reported error line\n"
+        exc = SyntaxError(
+            "invalid syntax",
+            ("<string>", 3, 1, "# reported error line\n", 3, 2),
+        )
+        exc._metadata = (0, 0, source)
+
+        result = ''.join(traceback.format_exception_only(exc))
+
+        self.assertIn("Did you mean 'return'", result)
+
     def test_no_keyword_suggestion_for_comma_errors(self):
         # When the parser identifies a missing comma, don't suggest
         # bogus keyword replacements like 'print' -> 'not'

Lib/traceback.py

@@ -1430,7 +1430,6 @@ def _find_keyword_typos(self):
         line, offset, source = self._exc_metadata
         end_line = int(self.lineno) if self.lineno is not None else 0
         lines = None
-        from_filename = False
 
         if source is None:
             if self.filename:
@@ -1439,8 +1438,6 @@ def _find_keyword_typos(self):
                         lines = f.read().splitlines()
                 except Exception:
                     line, end_line, offset = 0,1,0
-                else:
-                    from_filename = True
             lines = lines if lines is not None else self.text.splitlines()
         else:
             lines = source.splitlines()
@@ -1462,17 +1459,19 @@ def _find_keyword_typos(self):
             return  # Original code compiles or is incomplete - can't validate fixes
 
         error_lines = error_code.splitlines()
-        tokens = tokenize.generate_tokens(io.StringIO(error_code).readline)
+        tokens = [
+            token
+            for token in tokenize.generate_tokens(io.StringIO(error_code).readline)
+            if token.type == tokenize.NAME
+        ]
+        # Look for typos on the reported error line first. If the parser only
+        # fails after an earlier typo, the remaining lines are still searched.
+        the_end = end_line if line == 0 else end_line + 1
+        tokens.sort(key=lambda token: token.start[0] + line != the_end)
         tokens_left_to_process = 10
         import difflib
         for token in tokens:
             start, end = token.start, token.end
-            if token.type != tokenize.NAME:
-                continue
-            # Only consider NAME tokens on the same line as the error
-            the_end = end_line if line == 0 else end_line + 1
-            if from_filename and token.start[0]+line != the_end:
-                continue
             wrong_name = token.string
             if wrong_name in keyword.kwlist:
                 continue