[clang] Improve diagnostic for "initializer-string for char array is too long"

[evansmal]

Summary

This patch improves the diagnostic message "initializer-string for char array is too long"
by specifying an expected array length and by indicating that the initializer string implicitly
includes the null terminator.

Fixes #58829

[shafik]

Thank you for this diagnostic fix, please make sure you update clang/test/CXX/dcl.decl/dcl.init/dcl.init.string/p2.cpp.

clang/include/clang/Basic/DiagnosticSemaKinds.td

5968

Note. we are allowed to initialize with a string-literal that is smaller than the array size so maybe something along the lines of array size is %0 but initializer has size %1

[evansmal]

In D141283#4036449, @shafik wrote:

Thank you for this diagnostic fix, please make sure you update clang/test/CXX/dcl.decl/dcl.init/dcl.init.string/p2.cpp.

@shafik I noticed that there are also some tests in: clang/test/SemaCXX/pascal-strings.cpp, clang/test/Sema/array-init.c, and clang/test/Sema/format-strings.c that also check for this diagnostic. I assume these should also be updated?

EDIT: It seems that git-clang-format is failing for the build. Running it re-formats most of clang/lib/Sema/SemaInit.cpp - can I commit these changes here or should I open a seperate PR for that?

[evansmal]

Improve specificity of the diagnostic message.

[tbaeder]

When grepping clang/test/ for "for char array is too long", I get a bunch of other hits for this diagnostic, are the those tests not failing for you (e.g. clang/test/Sema/array-init.c:149)?

[evansmal]

In D141283#4038703, @tbaeder wrote:

When grepping clang/test/ for "for char array is too long", I get a bunch of other hits for this diagnostic, are the those tests not failing for you (e.g. clang/test/Sema/array-init.c:149)?

@tbaeder Yeah, the tests don't fail for me. I can update them, but I'm unsure how to tell if they're working properly.

[shafik]

In D141283#4038703, @tbaeder wrote:

When grepping clang/test/ for "for char array is too long", I get a bunch of other hits for this diagnostic, are the those tests not failing for you (e.g. clang/test/Sema/array-init.c:149)?

The change is specific to C++, so it should not apply to .c file, right?

[gribozavr2]

In D141283#4038703, @tbaeder wrote:

When grepping clang/test/ for "for char array is too long", I get a bunch of other hits for this diagnostic, are the those tests not failing for you (e.g. clang/test/Sema/array-init.c:149)?

"expected-error" matches a substring, that's why they are not failing.

[evansmal]

Thanks all for the review!

[shafik]

LGTM

clang/test/CXX/dcl.decl/dcl.init/dcl.init.string/p2.cpp

2–3

Can you also add:

char test2[1]="";

We expect this to succeed

[evansmal]

Added additional test case

[evansmal]

Hi @shafik! Thanks for the review. I don't have commit access and am unsure what I should do to get this committed. Can you let me know what I need to do next?

[tbaeder]

AFAIK we need a name and an email address so the commit can be attributed properly.

[evansmal]

In D141283#4065117, @tbaeder wrote:

AFAIK we need a name and an email address so the commit can be attributed properly.

The commit can be attributed to Evan Smal <[e-mail]>

[evansmal]

@tbaeder Thank you! :^)

clang/include/clang/Basic/DiagnosticSemaKinds.td

def ext_excess_initializers_for_sizeless_type : ExtWarn<
   "excess elements in initializer for indivisible sizeless type %0">,
   InGroup<ExcessInitializers>;
 def err_excess_initializers_in_char_array_initializer : Error<
   "excess elements in char array initializer">;
 def ext_excess_initializers_in_char_array_initializer : ExtWarn<
   "excess elements in char array initializer">,
   InGroup<ExcessInitializers>;
 def err_initializer_string_for_char_array_too_long : Error<
-  "initializer-string for char array is too long">;
+  "initializer-string for char array is too long, array size is %0 but initializer has size %1 (including the null terminating character)">;
 def ext_initializer_string_for_char_array_too_long : ExtWarn<
   "initializer-string for char array is too long">,
   InGroup<ExcessInitializers>;
 def warn_missing_field_initializers : Warning<
   "missing field %0 initializer">,
   InGroup<MissingFieldInitializers>, DefaultIgnore;
 def warn_braces_around_init : Warning<
   "braces around %select{scalar |}0initializer">,

clang/lib/Sema/SemaInit.cpp

if (StringLiteral *SL = dyn_cast<StringLiteral>(Str->IgnoreParens())) {
       if (SL->isPascal())
         StrLength--;
     }
 
     // [dcl.init.string]p2
     if (StrLength > CAT->getSize().getZExtValue())
       S.Diag(Str->getBeginLoc(),
              diag::err_initializer_string_for_char_array_too_long)
+          << CAT->getSize().getZExtValue() << StrLength
           << Str->getSourceRange();
   } else {
     // C99 6.7.8p14.
     if (StrLength-1 > CAT->getSize().getZExtValue())
       S.Diag(Str->getBeginLoc(),
              diag::ext_initializer_string_for_char_array_too_long)
           << Str->getSourceRange();
   }

clang/test/CXX/dcl.decl/dcl.init/dcl.init.string/p2.cpp

 // RUN: %clang_cc1 -fsyntax-only -verify %s
-char test1[1]="f"; // expected-error {{initializer-string for char array is too long}}
+char test1[1]="f"; // expected-error {{initializer-string for char array is too long, array size is 1 but initializer has size 2 (including the null terminating character)}}
+char test2[1]="";