Is C++ ASCII-aware?

Question

A colleague told me:

C++ is not ASCII-aware.

The source character set of a C++ program is implementation-defined, so to what extent is my colleague incorrect?

@R.MartinhoFernandes: .... dammit. Gonna need more SO accounts. Start a stage show up in this joint. — Lightness Races in Orbit, Apr 02 '13 at 11:48
As a strawman, maybe in the sense that the characters for the digits 0-9 must be in that order, without any gaps? This has nothing much to do with ascii per se, but it would disallow some contrived encodings... — PlasmaHH, Apr 02 '13 at 11:48
@PlasmaHH is there a IANA-registered encoding that would not be allowed by that? — R. Martinho Fernandes, Apr 02 '13 at 11:53

score 6 · Accepted Answer · answered Apr 02 '13 at 11:57

6

The c++ compiler needs to be ASCII aware when it links the numeric 48 value to '0'. So yes, it needs to be ASCII-aware.

But does it always needs to? Imagine you work with EBCDIC ('0' => 240). Then the compiler probably doesn't care about ASCII. Maybe that's what you colleague meant.

C++, generally speaking, does not really care about ASCII. It is an implementation detail.

answered Apr 02 '13 at 11:57

Jean

7,623
6
43
58

3

[This was discussed here](http://stackoverflow.com/questions/15598698/subtracting-0/15598711#15598711) some time ago. [Bo Persson](http://stackoverflow.com/users/597607/bo-persson) said: `On the machine I work on we have '0' => 240, but '9' - '0' is still 9` – Jean Apr 02 '13 at 11:58

score 5 · Answer 2 · answered Apr 02 '13 at 11:45

5

The C++ standard text is "aware" of ASCII in that it makes non-normative mention in a footnote:

[C++11: Footnote 14]: The glyphs for the members of the basic source character set are intended to identify characters from the subset of ISO/IEC 10646 which corresponds to the ASCII character set. However, because the mapping from source file characters to the source character set (described in translation phase 1) is specified as implementation-defined, an implementation is required to document how the basic source characters are represented in source files.

In doing so, it is declaring that the standardised language itself is not ASCII-aware.

"Seems like awareness of ASCII to me!" you might say. Well, no. The mere mention of "ASCII" in the language definition has not made the language ASCII-aware. This is in the same way that you can program a robot to say the words "I am not self-aware" without requiring that the robot become aware of self.

answered Apr 02 '13 at 11:45

Lightness Races in Orbit

378,754
76
643
1,055

inb4 you can program robot to say words. – Bartek Banachewicz Apr 02 '13 at 11:46
can you tell me some examples languages those are not-ascii aware. are low-level assembly? binary ? – Grijesh Chauhan Apr 02 '13 at 11:51
@GrijeshChauhan: C++ is one. – Lightness Races in Orbit Apr 02 '13 at 11:51
1

@LightnessRacesinOrbit ok now I understood. :) after reading again. – Grijesh Chauhan Apr 02 '13 at 11:53
maybe he meant the previous one. Just sayin' – Bartek Banachewicz Apr 02 '13 at 11:55
@MikeSeymour: Interesting! As a superset, is that by definition, or merely by design/through intent? – Lightness Races in Orbit Apr 02 '13 at 11:59
@LightnessRacesinOrbit: By definition I guess - it defines the code points 0-127 to be the same as their ASCII equivalents. (Sorry, I just accidentally deleted my comment; I'll retype it in a second) – Mike Seymour Apr 02 '13 at 12:00
2

However, the definition of a `universal-character-name` is based on the Universal Character Set (ISO/IEC 10646), which is a superset of ASCII. So the language does have some awareness of ASCII. – Mike Seymour Apr 02 '13 at 12:03

Is C++ ASCII-aware?

2 Answers2

Linked