[m-rev.] for review: define whitespace in the reference manual

[Julien Fischer]

Hi Peter,

On Wed, 14 Jan 2015, Peter Wang wrote:

> On Wed, 14 Jan 2015 11:32:25 +1100 (AEDT), Julien Fischer <jfischer at opturion.com> wrote:
>>
>> @@ -197,6 +198,22 @@ for clauses.
>>   Mercury program source files must be written using the UTF-8 encoding of
>>   the Unicode character set.
>>
>> + at node Whitespace
>> + at section Whitespace
>> +
>> +In Mercury program source files, whitespace is defined to be the following
>> +characters:
>> +
>> + at multitable {Carriage_Return} {CHARACTER_TABULATION} {Unicode_code_point}
>> + at headitem Character       @tab Unicode name              @tab Unicode code point
>> +    @item Space           @tab @sc{space}                @tab U+0020
>> +    @item Horizontal-tab  @tab @sc{character tabulation} @tab U+0009
>> +    @item Line-feed       @tab @sc{line feed}            @tab U+000A
>> +    @item Vertical-tab    @tab @sc{line tabulation}      @tab U+000B
>> +    @item Form-feed       @tab @sc{device control 2}     @tab U+000C
>> +    @item Carriage-return @tab @sc{device control 3}     @tab U+000D
>> + at end multitable
>
> You've mixed up U+000C (FORM FEED) with U+0012 and U+000D (CARRIAGE
> RETURN) with U+0013.

Fixed.

> The code point values are unambiguous so I don't see a need for two
> sets of names anyway.

I will delete the first column and add a new third column that contains
additional information where relevant, for example that the Unicode LINE
TABULATION character is a vertical tab.  (I suspect many people are more
familiar with the latter name.)

Cheers,
Julien.