From a4ea4737ac13ec0e0a34a26624330010f51b27be Mon Sep 17 00:00:00 2001
From: Amaan Qureshi <amaanq12@gmail.com>
Date: Fri, 25 Aug 2023 18:28:27 -0400
Subject: [PATCH] fix: do not increment `current_included_range_index` past
 `included_range_count` in `__do_advance`

---
 cli/src/tests/mod.rs                          |  1 +
 cli/src/tests/parser_hang_test.rs             | 61 +++++++++++++++++++
 lib/src/lexer.c                               |  4 +-
 .../get_col_should_hang_not_crash/corpus.txt  |  0
 .../get_col_should_hang_not_crash/grammar.js  | 13 ++++
 .../get_col_should_hang_not_crash/scanner.c   | 17 ++++++
 6 files changed, 95 insertions(+), 1 deletion(-)
 create mode 100644 cli/src/tests/parser_hang_test.rs
 create mode 100644 test/fixtures/test_grammars/get_col_should_hang_not_crash/corpus.txt
 create mode 100644 test/fixtures/test_grammars/get_col_should_hang_not_crash/grammar.js
 create mode 100644 test/fixtures/test_grammars/get_col_should_hang_not_crash/scanner.c

diff --git a/cli/src/tests/mod.rs b/cli/src/tests/mod.rs
index e579209f..308fc2c5 100644
--- a/cli/src/tests/mod.rs
+++ b/cli/src/tests/mod.rs
@@ -5,6 +5,7 @@ mod helpers;
 mod highlight_test;
 mod language_test;
 mod node_test;
+mod parser_hang_test;
 mod parser_test;
 mod pathological_test;
 mod query_test;
diff --git a/cli/src/tests/parser_hang_test.rs b/cli/src/tests/parser_hang_test.rs
new file mode 100644
index 00000000..cfa34957
--- /dev/null
+++ b/cli/src/tests/parser_hang_test.rs
@@ -0,0 +1,61 @@
+use pretty_assertions::assert_eq;
+use tree_sitter::Parser;
+
+use crate::{
+    generate::generate_parser_for_grammar,
+    tests::helpers::fixtures::{fixtures_dir, get_test_language},
+};
+
+#[test]
+fn test_grammar_that_should_hang_and_not_segfault() {
+    use std::sync::mpsc;
+
+    let (tx, rx) = mpsc::channel();
+
+    std::thread::spawn(move || {
+        let (parser_name, parser_code) = generate_parser_for_grammar(
+            r#"
+            {
+                "name": "get_col_should_hang_not_crash",
+                "rules": {
+                    "source_file": {
+                        "type": "SEQ",
+                        "members": [ { "type": "SYMBOL", "name": "test" } ]
+                    }
+                },
+                "extras": [ { "type": "PATTERN", "value": "\\s" } ],
+                "externals": [ { "type": "SYMBOL", "name": "test" } ]
+            }
+            "#,
+        )
+        .unwrap();
+
+        let mut parser = Parser::new();
+        parser
+            .set_language(get_test_language(
+                &parser_name,
+                &parser_code,
+                Some(
+                    fixtures_dir()
+                        .join("test_grammars")
+                        .join("get_col_should_hang_not_crash")
+                        .as_path(),
+                ),
+            ))
+            .unwrap();
+
+        let code_that_should_hang = "\nHello";
+
+        parser.parse(code_that_should_hang, None).unwrap();
+
+        // Won't be reached
+        let _ = tx.send(());
+    });
+
+    // Ok signifies that it did not hang
+    // RecvTimeoutError::Disconnected signifies that the parser thread exited unexpectedly (crashed)
+    assert_eq!(
+        rx.recv_timeout(std::time::Duration::from_secs(5)),
+        Err(mpsc::RecvTimeoutError::Timeout)
+    );
+}
diff --git a/lib/src/lexer.c b/lib/src/lexer.c
index e32158b2..d108c04e 100644
--- a/lib/src/lexer.c
+++ b/lib/src/lexer.c
@@ -172,7 +172,9 @@ static void ts_lexer__do_advance(Lexer *self, bool skip) {
     self->current_position.bytes >= current_range->end_byte ||
     current_range->end_byte == current_range->start_byte
   ) {
-    self->current_included_range_index++;
+    if (self->current_included_range_index < self->included_range_count) {
+      self->current_included_range_index++;
+    }
     if (self->current_included_range_index < self->included_range_count) {
       current_range++;
       self->current_position = (Length) {
diff --git a/test/fixtures/test_grammars/get_col_should_hang_not_crash/corpus.txt b/test/fixtures/test_grammars/get_col_should_hang_not_crash/corpus.txt
new file mode 100644
index 00000000..e69de29b
diff --git a/test/fixtures/test_grammars/get_col_should_hang_not_crash/grammar.js b/test/fixtures/test_grammars/get_col_should_hang_not_crash/grammar.js
new file mode 100644
index 00000000..83d57d2c
--- /dev/null
+++ b/test/fixtures/test_grammars/get_col_should_hang_not_crash/grammar.js
@@ -0,0 +1,13 @@
+module.exports = grammar({
+  name: 'get_col_should_hang_not_crash',
+
+  externals: $ => [
+    $.test,
+  ],
+
+  rules: {
+    source_file: $ => seq(
+      $.test
+    ),
+  },
+});
diff --git a/test/fixtures/test_grammars/get_col_should_hang_not_crash/scanner.c b/test/fixtures/test_grammars/get_col_should_hang_not_crash/scanner.c
new file mode 100644
index 00000000..d21ec6d4
--- /dev/null
+++ b/test/fixtures/test_grammars/get_col_should_hang_not_crash/scanner.c
@@ -0,0 +1,17 @@
+#include <tree_sitter/parser.h>
+
+unsigned tree_sitter_get_col_should_hang_not_crash_external_scanner_serialize() { return 0; }
+
+void tree_sitter_get_col_should_hang_not_crash_external_scanner_deserialize() {}
+
+void *tree_sitter_get_col_should_hang_not_crash_external_scanner_create() { return NULL; }
+
+void tree_sitter_get_col_should_hang_not_crash_external_scanner_destroy() {}
+
+bool tree_sitter_get_col_should_hang_not_crash_external_scanner_scan(void *payload, TSLexer *lexer,
+                                            const bool *valid_symbols) {
+    while (true) {
+        lexer->advance(lexer, false);
+        lexer->get_column(lexer);
+    }
+}