From 2522c1ed611b0237b51a7e7c3fa1bc009f7d7fdd Mon Sep 17 00:00:00 2001
From: Oleh Prypin <oleh@pryp.in>
Date: Fri, 12 Nov 2021 15:12:39 +0100
Subject: [PATCH] Show proper syntax errors in some edge cases in the parser

* A backslash in a regex at the end of the source code was a crash.
* A regex capture `$123` with too big of an integer was a crash.
---
 spec/compiler/lexer/lexer_spec.cr     | 1 +
 spec/compiler/parser/parser_spec.cr   | 3 +++
 src/compiler/crystal/syntax/lexer.cr  | 1 +
 src/compiler/crystal/syntax/parser.cr | 4 +++-
 4 files changed, 8 insertions(+), 1 deletion(-)

diff --git a/spec/compiler/lexer/lexer_spec.cr b/spec/compiler/lexer/lexer_spec.cr
index 30910c406354..d0db0cf68c89 100644
--- a/spec/compiler/lexer/lexer_spec.cr
+++ b/spec/compiler/lexer/lexer_spec.cr
@@ -402,6 +402,7 @@ describe "Lexer" do
   end
 
   assert_syntax_error "/foo", "Unterminated regular expression"
+  assert_syntax_error "/\\", "Unterminated regular expression"
   assert_syntax_error ":\"foo", "unterminated quoted symbol"
 
   it "lexes utf-8 char" do
diff --git a/spec/compiler/parser/parser_spec.cr b/spec/compiler/parser/parser_spec.cr
index 8e4216e52ccb..219d08d4e111 100644
--- a/spec/compiler/parser/parser_spec.cr
+++ b/spec/compiler/parser/parser_spec.cr
@@ -1121,6 +1121,9 @@ module Crystal
     it_parses "foo $1", Call.new(nil, "foo", Call.new(Global.new("$~"), "[]", 1.int32))
     it_parses "$~ = 1", Assign.new("$~".var, 1.int32)
 
+    assert_syntax_error "$2147483648"
+    assert_syntax_error "$99999999999999999999999?", "Regex capture index 99999999999999999999999 doesn't fit in an Int32"
+
     it_parses "foo /a/", Call.new(nil, "foo", regex("a"))
     it_parses "foo(/a/)", Call.new(nil, "foo", regex("a"))
     it_parses "foo(//)", Call.new(nil, "foo", regex(""))
diff --git a/src/compiler/crystal/syntax/lexer.cr b/src/compiler/crystal/syntax/lexer.cr
index 5efa3490cca8..93551535850c 100644
--- a/src/compiler/crystal/syntax/lexer.cr
+++ b/src/compiler/crystal/syntax/lexer.cr
@@ -2019,6 +2019,7 @@ module Crystal
         if delimiter_state.allow_escapes
           if delimiter_state.kind == :regex
             char = next_char
+            raise_unterminated_quoted delimiter_state if char == '\0'
             next_char
             @token.type = :STRING
             if char == '/' || char.ascii_whitespace?
diff --git a/src/compiler/crystal/syntax/parser.cr b/src/compiler/crystal/syntax/parser.cr
index f63e324f4383..b0e590b51ba4 100644
--- a/src/compiler/crystal/syntax/parser.cr
+++ b/src/compiler/crystal/syntax/parser.cr
@@ -997,7 +997,9 @@ module Crystal
           method = "[]"
         end
         location = @token.location
-        node_and_next_token Call.new(Global.new("$~").at(location), method, NumberLiteral.new(value.to_i))
+        index = value.to_i?
+        raise "Regex capture index #{value} doesn't fit in an Int32" unless index
+        node_and_next_token Call.new(Global.new("$~").at(location), method, NumberLiteral.new(index))
       when :__LINE__
         node_and_next_token MagicConstant.expand_line_node(@token.location)
       when :__END_LINE__