[ruby/prism] Handle invalid constants in the constant pool

It's possible for us to parse a constant that is invalid in the
current encoding. To fix this, we wrap the intern function in an
rb_protect to ensure we always put a valid symbol into the AST.

https://github.com/ruby/prism/commit/e78398554d

2 files changed, 16 insertions, 3 deletions

diff --git a/prism/prism.c b/prism/prism.c
index 42d9712063..de32ccae63 100644
--- a/prism/prism.c
+++ b/prism/prism.c
@@ -8633,13 +8633,16 @@ parser_lex(pm_parser_t *parser) {
                     if (!lex_state_beg_p(parser) && match(parser, '=')) {
                         lex_state_set(parser, PM_LEX_STATE_BEG);
                         LEX(PM_TOKEN_PERCENT_EQUAL);
-                    }
-                    else if(
+                    } else if (
                         lex_state_beg_p(parser) ||
                         (lex_state_p(parser, PM_LEX_STATE_FITEM) && (peek(parser) == 's')) ||
                         lex_state_spcarg_p(parser, space_seen)
                     ) {
                         if (!parser->encoding.alnum_char(parser->current.end, parser->end - parser->current.end)) {
+                            if (*parser->current.end >= 0x80) {
+                                pm_parser_err_current(parser, PM_ERR_INVALID_PERCENT);
+                            }
+
                             lex_mode_push_string(parser, true, false, lex_mode_incrementor(*parser->current.end), lex_mode_terminator(*parser->current.end));
 
                             size_t eol_length = match_eol(parser);
diff --git a/prism/templates/ext/prism/api_node.c.erb b/prism/templates/ext/prism/api_node.c.erb
index 6da5832185..1ffbf6c9f8 100644
--- a/prism/templates/ext/prism/api_node.c.erb
+++ b/prism/templates/ext/prism/api_node.c.erb
@@ -83,7 +83,17 @@ pm_ast_new(pm_parser_t *parser, pm_node_t *node, rb_encoding *encoding) {
 
     for (uint32_t index = 0; index < parser->constant_pool.size; index++) {
         pm_constant_t *constant = &parser->constant_pool.constants[index];
-        constants[index] = rb_intern3((const char *) constant->start, constant->length, encoding);
+        int state = 0;
+
+        VALUE string = rb_enc_str_new((const char *) constant->start, constant->length, encoding);
+        ID value = rb_protect(rb_intern_str, string, &state);
+
+        if (state != 0) {
+            value = rb_intern_const("?");
+            rb_set_errinfo(Qnil);
+        }
+
+        constants[index] = value;
     }
 
     pm_node_stack_node_t *node_stack = NULL;