llvm IR documentation doesn't describe string constants

[llvmbot]

Bugzilla Link	4224
Version	trunk
OS	Windows XP
Blocks	llvm/llvm-bugzilla-archive#10637
Reporter	LLVM Bugzilla Contributor
CC	@jryans,@nlewycky
Fixed by commit(s)	217583

Extended Description

(Pointed out on IRC). The LLVM IR docs don't ever describe the c"..." syntax.

[llvmbot]

There is only example:
@.str = private unnamed_addr constant [13 x i8] c"hello world\0A\00"

Please also specify encoding of characters inside string literal. Is it ASCII or Latin_1? Could one use wide characters ( [6 x i16] c"Привет\00" )? Wide-wide characters [ N x i32]?

[llvmbot]

*** Bug llvm/llvm-bugzilla-archive#12576 has been marked as a duplicate of this bug. ***

[jryans]

The basic character array syntax was added to the IR manual in r217583, but character encoding is not mentioned.

As far as I can tell, character arrays are parsed as an opaque array of i8 until the next quote character. There is no special processing or text decoding of the value.

[JOE1994]

The Array constants subsection of the Complex Constants section of langref describes the c"..." syntax as follows:

As a special case, character array constants may also be represented as a double-quoted string using the c prefix. For example: “c"Hello World\0A\00"”.

The String Constants section of langref has a comment on how the bytes are interpreted:

The interpretation of string constants (e.g. their character encoding) depends on context.

No further action needed to address this ticket.