[gcc r14-8612] gccrs: Handle newlines during string parsing while lexing

[gcc r14-8612] gccrs: Handle newlines during string parsing while lexing

[Arthur Cohen]

https://gcc.gnu.org/g:8b7d6adbbb886039f8c6e14a581dcbc9c63ce380

commit r14-8612-g8b7d6adbbb886039f8c6e14a581dcbc9c63ce380
Author: Nirmal Patel <nirmal@nirmal.dev>
Date:   Wed Dec 27 17:49:58 2023 -0500

    gccrs: Handle newlines during string parsing while lexing
    
    If newline strings are encountered while lexing, the lexer now handles
    newline characters by incrementing current line number. This provides
    correct line number when displaying errors. If the lexer encounters end
    of file before string end, then it will use the start of the string as
    the location to an report error.
    
    gcc/rust/ChangeLog:
            * lex/rust-lex.cc (Lexer::parse_byte_string): Handle newline
            while parsing byte strings
            (Lexer::parse_string): Handle newline while parsing strings
    
    Signed-off-by: Nirmal Patel <nirmal@nirmal.dev>

Diff:
---
 gcc/rust/lex/rust-lex.cc                         | 40 +++++++++++++++++-------
 gcc/testsuite/rust/compile/issue-2187.rs         | 11 +++++++
 gcc/testsuite/rust/execute/torture/issue-2187.rs | 23 ++++++++++++++
 3 files changed, 63 insertions(+), 11 deletions(-)

diff --git a/gcc/rust/lex/rust-lex.cc b/gcc/rust/lex/rust-lex.cc
index 5bff2d9125c4..6fc336d97e0a 100644
--- a/gcc/rust/lex/rust-lex.cc
+++ b/gcc/rust/lex/rust-lex.cc
@@ -1766,13 +1766,15 @@ Lexer::parse_byte_string (location_t loc)
   std::string str;
   str.reserve (16); // some sensible default
 
-  int length = 1;
   current_char = peek_input ();
 
+  const location_t string_begin_locus = get_current_location ();
+
   while (current_char != '"' && !current_char.is_eof ())
     {
       if (current_char == '\\')
 	{
+	  int length = 1;
 	  auto escape_length_pair = parse_escape ('"');
 	  char output_char = std::get<0> (escape_length_pair);
 
@@ -1784,18 +1786,25 @@ Lexer::parse_byte_string (location_t loc)
 	  if (output_char != 0 || !std::get<2> (escape_length_pair))
 	    str += output_char;
 
+	  current_column += length;
+
 	  continue;
 	}
 
-      length++;
+      current_column++;
+      if (current_char.value == '\n')
+	{
+	  current_line++;
+	  current_column = 1;
+	  // tell line_table that new line starts
+	  start_line (current_line, max_column_hint);
+	}
 
       str += current_char;
       skip_input ();
       current_char = peek_input ();
     }
 
-  current_column += length;
-
   if (current_char == '"')
     {
       current_column++;
@@ -1805,7 +1814,7 @@ Lexer::parse_byte_string (location_t loc)
     }
   else if (current_char.is_eof ())
     {
-      rust_error_at (get_current_location (), "unended byte string literal");
+      rust_error_at (string_begin_locus, "unended byte string literal");
       return Token::make (END_OF_FILE, get_current_location ());
     }
   else
@@ -1996,14 +2005,17 @@ Lexer::parse_string (location_t loc)
   std::string str;
   str.reserve (16); // some sensible default
 
-  int length = 1;
   current_char = peek_input ();
 
+  const location_t string_begin_locus = get_current_location ();
+
   // FIXME: This fails if the input ends. How do we check for EOF?
   while (current_char.value != '"' && !current_char.is_eof ())
     {
       if (current_char.value == '\\')
 	{
+	  int length = 1;
+
 	  // parse escape
 	  auto utf8_escape_pair = parse_utf8_escape ();
 	  current_char = std::get<0> (utf8_escape_pair);
@@ -2016,21 +2028,28 @@ Lexer::parse_string (location_t loc)
 	  if (current_char != Codepoint (0) || !std::get<2> (utf8_escape_pair))
 	    str += current_char.as_string ();
 
+	  current_column += length;
+
 	  // FIXME: should remove this but can't.
 	  // `parse_utf8_escape` does not update `current_char` correctly.
 	  current_char = peek_input ();
 	  continue;
 	}
 
-      length++;
+      current_column++;
+      if (current_char.value == '\n')
+	{
+	  current_line++;
+	  current_column = 1;
+	  // tell line_table that new line starts
+	  start_line (current_line, max_column_hint);
+	}
 
       str += current_char;
       skip_input ();
       current_char = peek_input ();
     }
 
-  current_column += length;
-
   if (current_char.value == '"')
     {
       current_column++;
@@ -2040,7 +2059,7 @@ Lexer::parse_string (location_t loc)
     }
   else if (current_char.is_eof ())
     {
-      rust_error_at (get_current_location (), "unended string literal");
+      rust_error_at (string_begin_locus, "unended string literal");
       return Token::make (END_OF_FILE, get_current_location ());
     }
   else
@@ -2049,7 +2068,6 @@ Lexer::parse_string (location_t loc)
     }
 
   str.shrink_to_fit ();
-  loc += length - 1;
 
   return Token::make_string (loc, std::move (str));
 }
diff --git a/gcc/testsuite/rust/compile/issue-2187.rs b/gcc/testsuite/rust/compile/issue-2187.rs
new file mode 100644
index 000000000000..deef417a668b
--- /dev/null
+++ b/gcc/testsuite/rust/compile/issue-2187.rs
@@ -0,0 +1,11 @@
+const A: &'static u8 = b"
+";
+const B: &'static str = b"
+";
+const C: &'static u8 = "
+";
+const D: &'static str = "
+";
+ERROR_TIME
+// { dg-error "unrecognised token" "" { target *-*-* } .-1 }
+// { dg-error "failed to parse item in crate" "" { target *-*-* } .-2 }
diff --git a/gcc/testsuite/rust/execute/torture/issue-2187.rs b/gcc/testsuite/rust/execute/torture/issue-2187.rs
new file mode 100644
index 000000000000..b53125750411
--- /dev/null
+++ b/gcc/testsuite/rust/execute/torture/issue-2187.rs
@@ -0,0 +1,23 @@
+/* { dg-output "L1\n\L2\nL3\nL4" } */
+extern "C" {
+    fn printf(s: *const i8, ...);
+}
+
+fn main() -> i32 {
+    let A = b"L1
+L2\0";
+    let B = "L3
+L4\0";
+
+    unsafe {
+        let a = "%s\n\0";
+        let b = a as *const str;
+        let c = b as *const i8;
+
+        printf(c, A);
+        printf(c, B);
+    }
+
+    0
+}
+