pythongh-100445: Improve error message for unterminated strings with escapes

hauntsaninja · hauntsaninja · commit bf712e8013b0 · 2022-12-22T18:50:14.000-06:00
diff --git a/Lib/test/test_syntax.py b/Lib/test/test_syntax.py
@@ -2156,8 +2156,14 @@ def test_error_parenthesis(self):
 
     def test_error_string_literal(self):
 
-        self._check_error("'blech", "unterminated string literal")
-        self._check_error('"blech', "unterminated string literal")
+        self._check_error("'blech", r"unterminated string literal \(.*\)$")
+        self._check_error('"blech', r"unterminated string literal \(.*\)$")
+        self._check_error(
+            r'"blech\"', r"unterminated string literal \(.*\); perhaps you escaped the end quote"
+        )
+        self._check_error(
+            r'r"blech\"', r"unterminated string literal \(.*\); perhaps you escaped the end quote"
+        )
         self._check_error("'''blech", "unterminated triple-quoted string literal")
         self._check_error('"""blech', "unterminated triple-quoted string literal")
 
diff --git a/Parser/tokenizer.c b/Parser/tokenizer.c
@@ -2009,6 +2009,7 @@ tok_get(struct tok_state *tok, struct token *token)
         int quote = c;
         int quote_size = 1;             /* 1 or 3 */
         int end_quote_size = 0;
+        int has_escaped_quote = 0;
 
         /* Nodes of type STRING, especially multi line strings
            must be handled differently in order to get both
@@ -2056,8 +2057,13 @@ tok_get(struct tok_state *tok, struct token *token)
                     return MAKE_TOKEN(ERRORTOKEN);
                 }
                 else {
-                    syntaxerror(tok, "unterminated string literal (detected at"
-                                     " line %d)", start);
+                    if (has_escaped_quote) {
+                        syntaxerror(tok, "unterminated string literal (detected at"
+                                        " line %d); perhaps you escaped the end quote?", start);
+                    } else {
+                        syntaxerror(tok, "unterminated string literal (detected at"
+                                        " line %d)", start);
+                    }
                     if (c != '\n') {
                         tok->done = E_EOLS;
                     }
@@ -2070,7 +2076,11 @@ tok_get(struct tok_state *tok, struct token *token)
             else {
                 end_quote_size = 0;
                 if (c == '\\') {
-                    tok_nextc(tok);  /* skip escaped char */
+                    // skip escaped char, but record whether the escaped char
+                    // was a quote
+                    if (tok_nextc(tok) == quote) {
+                        has_escaped_quote = 1;
+                    }
                 }
             }
         }
