Re: Community Input for the Maintenance and Revision of the Ada Programming Language

[Robert Eachus <rieachus@comcast.net>]

On Friday, August 11, 2017 at 10:48:39 AM UTC-4, jsqu...@gmail.com wrote:

> > On the otherhand Ada.Directories and Ada.Text_IO are woefully outdated. For example, if you have a computer and you speak a language like Japanese, French, German, Spanish, etc. think of how frustrating it is when you simply can't open a file with anything other than ASCII in the path. A program that deals with relative directories might work fine in one directory but get moved and suddenly fail hopelessly. Poor end-user :(
> 
> My mistake String supports more than just ASCII (Duh), but Asian and other esoteric languages obvious not.

There are three decisions about character sets made every time you create Ada source files.  It is nice to have them all the same, but people used to that get their minds wrapped around the axle when it is not true.  The three cases are:

1) The character set used to write the Ada parts of the program.
2) The character sets the executing computer will use when interpreting characters and string values.
3) The character sets used by the programmer to express character and string constants.

In addition, there are the character sets for the keyboard, the display, any printers, etc.  Fortunately most printers today can handle mixed character sets if the compiler knows how to prepare listings for them.  Also a compiler may allow characters in comments which are not allowed elsewhere in the source--but it is nice for the character sets in Strings and comments to match.

Compilers can, and do have switches to set the character set used for the Ada code.  Technically it is not standard compliant to compile in this mode for the international standard, but may be fine for National standards. It is also possible to have switches choose the source and target representation for characters and strings.  Since the name of a file is represented as a String, it is quite possible for a compiler to support ShiftJIS as the character set for String, and thus for file names.  Of course, you need the computer running the program to interpret them the same way...

I hope that Ada is (slowly) moving toward UTF-8 as the default--with the emphasis on slowly.  For example, a standard conversion from UTF-8 to Long_Long_Character and back would be very nice.  (What about Long_Character?  It should be Unicode, and if there is any deviation between UTF-16 and Unicode, stick with Unicode.)

Is it possible for you to find an Ada compiler that will allow programs written in your favorite language and character set?  That is a market issue.  Ada has moved in the direction of supporting multiple character sets a lot better than most standard languages.  But if you don't have the vendor support?  To take an example from above, you should always be able to pass ShiftJIS strings to the file system.*  But to write the programs in Ada, you will need a ShiftJIS capable terminal and printer to stay sane.   I would hope that Unicode and full ISO 10646 is gradually replacing other Japanese encodings.

* Gotchas:  If the second byte of a shift JIS pair is '"', you will need to put an extra " in when writing the string, with the compiler thinking you are using Latin-1 or whatever.  Similarly high page control codes will have to be entered by name.  The best thing to do is write a program that can convert JIS or Shift JIS source to an 8-bit Ada source string.