[ruby/prism] Update more escape error messages to match CRuby

https://github.com/ruby/prism/commit/ab43b3ab66

3 files changed, 34 insertions, 14 deletions

diff --git a/prism/prism.c b/prism/prism.c
index be6b52f5b1..75f1432161 100644
--- a/prism/prism.c
+++ b/prism/prism.c
@@ -9658,7 +9658,8 @@ escape_read(pm_parser_t *parser, pm_buffer_t *buffer, pm_buffer_t *regular_expre
                         pm_parser_err(parser, unicode_start, unicode_start + hexadecimal_length, PM_ERR_ESCAPE_INVALID_UNICODE_LONG);
                     } else if (hexadecimal_length == 0) {
                         // there are not hexadecimal characters
-                        pm_parser_err(parser, unicode_start, unicode_start + hexadecimal_length, PM_ERR_ESCAPE_INVALID_UNICODE);
+                        pm_parser_err(parser, parser->current.end, parser->current.end, PM_ERR_ESCAPE_INVALID_UNICODE);
+                        pm_parser_err(parser, parser->current.end, parser->current.end, PM_ERR_ESCAPE_INVALID_UNICODE_TERM);
                         return;
                     }
 
@@ -9707,10 +9708,6 @@ escape_read(pm_parser_t *parser, pm_buffer_t *buffer, pm_buffer_t *regular_expre
                 }
             }
 
-            if (flags & (PM_ESCAPE_FLAG_CONTROL | PM_ESCAPE_FLAG_META)) {
-                pm_parser_err(parser, start, parser->current.end, PM_ERR_INVALID_ESCAPE_CHARACTER);
-            }
-
             return;
         }
         case 'c': {
@@ -9733,6 +9730,12 @@ escape_read(pm_parser_t *parser, pm_buffer_t *buffer, pm_buffer_t *regular_expre
                         return;
                     }
                     parser->current.end++;
+
+                    if (match(parser, 'u') || match(parser, 'U')) {
+                        pm_parser_err(parser, parser->current.start, parser->current.end, PM_ERR_INVALID_ESCAPE_CHARACTER);
+                        return;
+                    }
+
                     escape_read(parser, buffer, regular_expression_buffer, flags | PM_ESCAPE_FLAG_CONTROL);
                     return;
                 case ' ':
@@ -9760,7 +9763,8 @@ escape_read(pm_parser_t *parser, pm_buffer_t *buffer, pm_buffer_t *regular_expre
         case 'C': {
             parser->current.end++;
             if (peek(parser) != '-') {
-                pm_parser_err_current(parser, PM_ERR_ESCAPE_INVALID_CONTROL);
+                size_t width = parser->encoding->char_width(parser->current.end, parser->end - parser->current.end);
+                pm_parser_err(parser, parser->current.start, parser->current.end + width, PM_ERR_ESCAPE_INVALID_CONTROL);
                 return;
             }
 
@@ -9783,6 +9787,12 @@ escape_read(pm_parser_t *parser, pm_buffer_t *buffer, pm_buffer_t *regular_expre
                         return;
                     }
                     parser->current.end++;
+
+                    if (match(parser, 'u') || match(parser, 'U')) {
+                        pm_parser_err(parser, parser->current.start, parser->current.end, PM_ERR_INVALID_ESCAPE_CHARACTER);
+                        return;
+                    }
+
                     escape_read(parser, buffer, regular_expression_buffer, flags | PM_ESCAPE_FLAG_CONTROL);
                     return;
                 case ' ':
@@ -9797,7 +9807,8 @@ escape_read(pm_parser_t *parser, pm_buffer_t *buffer, pm_buffer_t *regular_expre
                     return;
                 default: {
                     if (!char_is_ascii_printable(peeked)) {
-                        pm_parser_err_current(parser, PM_ERR_ESCAPE_INVALID_CONTROL);
+                        size_t width = parser->encoding->char_width(parser->current.end, parser->end - parser->current.end);
+                        pm_parser_err(parser, parser->current.start, parser->current.end + width, PM_ERR_ESCAPE_INVALID_CONTROL);
                         return;
                     }
 
@@ -9810,7 +9821,8 @@ escape_read(pm_parser_t *parser, pm_buffer_t *buffer, pm_buffer_t *regular_expre
         case 'M': {
             parser->current.end++;
             if (peek(parser) != '-') {
-                pm_parser_err_current(parser, PM_ERR_ESCAPE_INVALID_META);
+                size_t width = parser->encoding->char_width(parser->current.end, parser->end - parser->current.end);
+                pm_parser_err(parser, parser->current.start, parser->current.end + width, PM_ERR_ESCAPE_INVALID_META);
                 return;
             }
 
@@ -9828,6 +9840,12 @@ escape_read(pm_parser_t *parser, pm_buffer_t *buffer, pm_buffer_t *regular_expre
                         return;
                     }
                     parser->current.end++;
+
+                    if (match(parser, 'u') || match(parser, 'U')) {
+                        pm_parser_err(parser, parser->current.start, parser->current.end, PM_ERR_INVALID_ESCAPE_CHARACTER);
+                        return;
+                    }
+
                     escape_read(parser, buffer, regular_expression_buffer, flags | PM_ESCAPE_FLAG_META);
                     return;
                 case ' ':
@@ -9842,7 +9860,8 @@ escape_read(pm_parser_t *parser, pm_buffer_t *buffer, pm_buffer_t *regular_expre
                     return;
                 default:
                     if (!char_is_ascii_printable(peeked)) {
-                        pm_parser_err_current(parser, PM_ERR_ESCAPE_INVALID_META);
+                        size_t width = parser->encoding->char_width(parser->current.end, parser->end - parser->current.end);
+                        pm_parser_err(parser, parser->current.start, parser->current.end + width, PM_ERR_ESCAPE_INVALID_META);
                         return;
                     }
 
diff --git a/prism/templates/src/diagnostic.c.erb b/prism/templates/src/diagnostic.c.erb
index eab3dd26bf..d9e195e08f 100644
--- a/prism/templates/src/diagnostic.c.erb
+++ b/prism/templates/src/diagnostic.c.erb
@@ -167,7 +167,7 @@ static const pm_diagnostic_data_t diagnostic_messages[PM_DIAGNOSTIC_ID_MAX] = {
     [PM_ERR_ESCAPE_INVALID_UNICODE_CM_FLAGS]    = { "invalid Unicode escape sequence; Unicode cannot be combined with control or meta flags", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_ESCAPE_INVALID_UNICODE_LITERAL]     = { "invalid Unicode escape sequence; Multiple codepoints at single character literal are disallowed", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_ESCAPE_INVALID_UNICODE_LONG]        = { "invalid Unicode escape sequence; maximum length is 6 digits", PM_ERROR_LEVEL_SYNTAX },
-    [PM_ERR_ESCAPE_INVALID_UNICODE_TERM]        = { "invalid Unicode escape sequence; needs closing `}`", PM_ERROR_LEVEL_SYNTAX },
+    [PM_ERR_ESCAPE_INVALID_UNICODE_TERM]        = { "unterminated Unicode escape", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_EXPECT_ARGUMENT]                    = { "expected an argument", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_EXPECT_EOL_AFTER_STATEMENT]         = { "unexpected %s, expecting end-of-input", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_EXPECT_EXPRESSION_AFTER_AMPAMPEQ]   = { "expected an expression after `&&=`", PM_ERROR_LEVEL_SYNTAX },
@@ -330,12 +330,12 @@ static const pm_diagnostic_data_t diagnostic_messages[PM_DIAGNOSTIC_ID_MAX] = {
     [PM_ERR_STATEMENT_PREEXE_BEGIN]             = { "unexpected a `BEGIN` at a non-statement position", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_STATEMENT_UNDEF]                    = { "unexpected an `undef` at a non-statement position", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_STRING_CONCATENATION]               = { "expected a string for concatenation", PM_ERROR_LEVEL_SYNTAX },
-    [PM_ERR_STRING_INTERPOLATED_TERM]           = { "expected a closing delimiter for the interpolated string", PM_ERROR_LEVEL_SYNTAX },
+    [PM_ERR_STRING_INTERPOLATED_TERM]           = { "unterminated string; expected a closing delimiter for the interpolated string", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_STRING_LITERAL_EOF]                 = { "unterminated string meets end of file", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_STRING_LITERAL_TERM]                = { "unexpected %s, expected a string literal terminator", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_SYMBOL_INVALID]                     = { "invalid symbol", PM_ERROR_LEVEL_SYNTAX }, // TODO expected symbol? prism.c ~9719
     [PM_ERR_SYMBOL_TERM_DYNAMIC]                = { "unterminated quoted string; expected a closing delimiter for the dynamic symbol", PM_ERROR_LEVEL_SYNTAX },
-    [PM_ERR_SYMBOL_TERM_INTERPOLATED]           = { "expected a closing delimiter for the interpolated symbol", PM_ERROR_LEVEL_SYNTAX },
+    [PM_ERR_SYMBOL_TERM_INTERPOLATED]           = { "unterminated symbol; expected a closing delimiter for the interpolated symbol", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_TERNARY_COLON]                      = { "expected a `:` after the true expression of a ternary operator", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_TERNARY_EXPRESSION_FALSE]           = { "expected an expression after `:` in the ternary operator", PM_ERROR_LEVEL_SYNTAX },
     [PM_ERR_TERNARY_EXPRESSION_TRUE]            = { "expected an expression after `?` in the ternary operator", PM_ERROR_LEVEL_SYNTAX },
diff --git a/test/prism/errors_test.rb b/test/prism/errors_test.rb
index 1357c97cf9..3670b90dd7 100644
--- a/test/prism/errors_test.rb
+++ b/test/prism/errors_test.rb
@@ -216,7 +216,7 @@ module Prism
 
     def test_unterminated_interpolated_symbol
       assert_error_messages ":\"#", [
-        "expected a closing delimiter for the interpolated symbol"
+        "unterminated symbol; expected a closing delimiter for the interpolated symbol"
       ]
     end
 
@@ -715,12 +715,13 @@ module Prism
 
       assert_errors expected, '"\u{000z}"', [
         ["invalid Unicode escape sequence", 7..7],
+        ["unterminated Unicode escape", 7..7]
       ]
     end
 
     def test_unterminated_unicode_brackets_should_be_a_syntax_error
       assert_errors expression('?\\u{3'), '?\\u{3', [
-        ["invalid Unicode escape sequence; needs closing `}`", 1..5],
+        ["unterminated Unicode escape", 1..5],
       ]
     end