From 42dcadfccfc3ad3aa3c9e08cd574fb60ea9911b4 Mon Sep 17 00:00:00 2001
From: Boshen <1430279+Boshen@users.noreply.github.com>
Date: Wed, 18 Sep 2024 05:41:44 +0000
Subject: [PATCH] fix(parser): hashbang comment should not keep the end newline
 char (#5844)

Previously it included a newline in the value

```
  "hashbang": {
    "type": "Hashbang",
    "start": 0,
    "end": 16,
    "value": "/usr/bin/node\n"
  },
```

This change will also make the lexer emit a `\n` token, which will make comment position detection correct.
---
 crates/oxc_codegen/src/gen.rs                 |  1 +
 crates/oxc_codegen/src/lib.rs                 |  5 +++++
 crates/oxc_parser/src/lexer/comment.rs        |  3 ++-
 crates/oxc_parser/src/lexer/trivia_builder.rs |  2 +-
 crates/oxc_parser/src/lib.rs                  |  9 +++++++++
 crates/oxc_prettier/src/format/mod.rs         | 17 +++++++++++------
 6 files changed, 29 insertions(+), 8 deletions(-)

diff --git a/crates/oxc_codegen/src/gen.rs b/crates/oxc_codegen/src/gen.rs
index 8c309e532..8f16f6497 100644
--- a/crates/oxc_codegen/src/gen.rs
+++ b/crates/oxc_codegen/src/gen.rs
@@ -52,6 +52,7 @@ impl<'a> Gen for Hashbang<'a> {
     fn gen(&self, p: &mut Codegen, _ctx: Context) {
         p.print_str("#!");
         p.print_str(self.value.as_str());
+        p.print_hard_newline();
     }
 }
 
diff --git a/crates/oxc_codegen/src/lib.rs b/crates/oxc_codegen/src/lib.rs
index eff84bf94..f529d44fc 100644
--- a/crates/oxc_codegen/src/lib.rs
+++ b/crates/oxc_codegen/src/lib.rs
@@ -281,6 +281,11 @@ impl<'a> Codegen<'a> {
         }
     }
 
+    #[inline]
+    fn print_hard_newline(&mut self) {
+        self.print_char(b'\n');
+    }
+
     #[inline]
     fn print_semicolon(&mut self) {
         self.print_char(b';');
diff --git a/crates/oxc_parser/src/lexer/comment.rs b/crates/oxc_parser/src/lexer/comment.rs
index 4d813d054..f366fe8d5 100644
--- a/crates/oxc_parser/src/lexer/comment.rs
+++ b/crates/oxc_parser/src/lexer/comment.rs
@@ -176,10 +176,11 @@ impl<'a> Lexer<'a> {
 
     /// Section 12.5 Hashbang Comments
     pub(super) fn read_hashbang_comment(&mut self) -> Kind {
-        while let Some(c) = self.next_char().as_ref() {
+        while let Some(c) = self.peek_char().as_ref() {
             if is_line_terminator(*c) {
                 break;
             }
+            self.consume_char();
         }
         self.token.is_on_new_line = true;
         Kind::HashbangComment
diff --git a/crates/oxc_parser/src/lexer/trivia_builder.rs b/crates/oxc_parser/src/lexer/trivia_builder.rs
index 5bc783614..1033b8d70 100644
--- a/crates/oxc_parser/src/lexer/trivia_builder.rs
+++ b/crates/oxc_parser/src/lexer/trivia_builder.rs
@@ -178,7 +178,7 @@ token
             kind: CommentKind::Block,
             position: CommentPosition::Leading,
             attached_to: 36,
-            preceded_by_newline: false, // hashbang comment always end in newline
+            preceded_by_newline: true,
             followed_by_newline: true,
         }];
         assert_eq!(comments, expected);
diff --git a/crates/oxc_parser/src/lib.rs b/crates/oxc_parser/src/lib.rs
index d433dae17..8dd4cf90e 100644
--- a/crates/oxc_parser/src/lib.rs
+++ b/crates/oxc_parser/src/lib.rs
@@ -523,6 +523,15 @@ mod test {
         }
     }
 
+    #[test]
+    fn hashbang() {
+        let allocator = Allocator::default();
+        let source_type = SourceType::default();
+        let source = "#!/usr/bin/node\n;";
+        let ret = Parser::new(&allocator, source, source_type).parse();
+        assert_eq!(ret.program.hashbang.unwrap().value.as_str(), "/usr/bin/node");
+    }
+
     #[test]
     fn unambiguous() {
         let allocator = Allocator::default();
diff --git a/crates/oxc_prettier/src/format/mod.rs b/crates/oxc_prettier/src/format/mod.rs
index 222398c71..51b7839ff 100644
--- a/crates/oxc_prettier/src/format/mod.rs
+++ b/crates/oxc_prettier/src/format/mod.rs
@@ -30,7 +30,7 @@ use cow_utils::CowUtils;
 use oxc_allocator::{Box, Vec};
 use oxc_ast::{ast::*, AstKind};
 use oxc_span::GetSpan;
-use oxc_syntax::identifier::is_identifier_name;
+use oxc_syntax::identifier::{is_identifier_name, is_line_terminator};
 
 use self::{array::Array, object::ObjectLike, template_literal::TemplateLiteralPrinter};
 use crate::{
@@ -60,10 +60,6 @@ impl<'a> Format<'a> for Program<'a> {
         let mut parts = p.vec();
         if let Some(hashbang) = &self.hashbang {
             parts.push(hashbang.format(p));
-            let c = p.source_text[..hashbang.span.end as usize].chars().last().unwrap();
-            if p.is_next_line_empty_after_index(hashbang.span.end - c.len_utf8() as u32) {
-                parts.extend(hardline!());
-            }
         }
         if let Some(doc) = block::print_block_body(
             p,
@@ -81,7 +77,16 @@ impl<'a> Format<'a> for Program<'a> {
 
 impl<'a> Format<'a> for Hashbang<'a> {
     fn format(&self, p: &mut Prettier<'a>) -> Doc<'a> {
-        Doc::Str(self.span.source_text(p.source_text))
+        let mut parts = p.vec();
+        parts.push(ss!(self.span.source_text(p.source_text)));
+        parts.extend(hardline!());
+        // Preserve original newline
+        if let Some(c) = p.source_text[self.span.end as usize..].chars().nth(1) {
+            if is_line_terminator(c) {
+                parts.extend(hardline!());
+            }
+        }
+        Doc::Array(parts)
     }
 }