[PATCH] libcpp: Fix up raw string literal parsing error-recovery [PR96323]

[PATCH] libcpp: Fix up raw string literal parsing error-recovery [PR96323]

[Jakub Jelinek]

Hi!

For (invalid) newline inside of the raw string literal delimiter, doing
continue means we skip the needed processing of newlines.  Instead of
duplicating that, this patch just doesn't continue for those.

Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?

2020-07-28  Jakub Jelinek  <jakub@redhat.com>

	PR preprocessor/96323
	* lex.c (lex_raw_string): For c == '\n' don't continue after reporting
	an prefix delimiter error.

	* c-c++-common/cpp/pr96323.c: New test.

--- libcpp/lex.c.jj	2020-05-20 10:33:37.000000000 +0200
+++ libcpp/lex.c	2020-07-27 09:12:19.827974203 +0200
@@ -1885,7 +1885,8 @@ lex_raw_string (cpp_reader *pfile, cpp_t
 		 the best failure mode.  */
 	      prefix_len = 0;
 	    }
-	  continue;
+	  if (c != '\n')
+	    continue;
 	}
 
       if (phase != PHASE_NONE)
--- gcc/testsuite/c-c++-common/cpp/pr96323.c.jj	2020-07-27 09:19:42.021557709 +0200
+++ gcc/testsuite/c-c++-common/cpp/pr96323.c	2020-07-27 09:30:37.197052063 +0200
@@ -0,0 +1,8 @@
+/* PR preprocessor/96323 */
+/* { dg-do preprocess } */
+/* { dg-options "-std=gnu99 -Wno-c++-compat -trigraphs" { target c } } */
+/* { dg-options "-std=c++0x" { target c++ } } */
+/* { dg-error "invalid new-line in raw string delimiter" "" { target *-*-* } .+2 } */
+/* { dg-warning "missing terminating . character" "" { target *-*-* } .+2 } */
+const char tu[] = R"a";
+const char tua[] = "(R)a";

	Jakub

Re: [PATCH] libcpp: Fix up raw string literal parsing error-recovery [PR96323]

[Nathan Sidwell]

On 7/28/20 4:26 AM, Jakub Jelinek wrote:
> Hi!
> 
> For (invalid) newline inside of the raw string literal delimiter, doing
> continue means we skip the needed processing of newlines.  Instead of
> duplicating that, this patch just doesn't continue for those.
> 
> Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
>

ok, thanks!  It's hard to think of all the wrong ways code can be written!

nathan

-- 
Nathan Sidwell