On 14/11/08 17:50, Charles Campbell wrote:
> [EMAIL PROTECTED] wrote:
>> Here is a sample of what I'm referring to: (need I say to ignore the
>> ruler line?)
>> ----+----1----+----2----+----3----+----4----+----5----+----6----
>> +----7----+----8
>> * This line works fine since it starts in column 7 as does the
>> next
>> *_*_*^**^*_*_*
>>
>> * The 2nd line below, while syntactically correct, is flagged
>> on the screen as being an error.
>> compute A = ((B + C) / D)
>> * ((E * F) + G)
>>
>> I acknowledge that an obvious solution at least for this elementary
>> example would be to move the
>> asterisk from where it is to the end of the previous line but
>> occassionally in real life the line is too
>> long, e.g. due to long field names, forcing the continuation as above.
>>
>> Lynn
>>
>>
>> On Nov 14, 7:01 am, Ben Schmidt<[EMAIL PROTECTED]> wrote:
>>
>>>> In COBOL an asterisk has 2 purposes.
>>>> First if it's in column 7 it signifies that line as being a coment.
>>>> Secondly, if it appears in any colum greater than 11 it is used as a
>>>> mathamatical operator.
>>>>
>>>> The editor correctly flags the lines with the asterisk in column 7 as
>>>> a comment, however, should that character appear anywhere else on the
>>>> line it flags this line as a syntax error which can be fairly
>>>> distracting and annoying.
>>>>
>>> It doesn't always mark it as an error; I can get it to appear normally,
>>> e.g. if I type 3 * 4 with a large indent. But I don't know COBOL, so I
>>> can't tell if what it's doing is right or wrong. Perhaps you could post
>>> some example code where the highlighting is wrong and explain why, and
>>> then we can help with the syntax file? The person listed as the
>>> maintainer of the syntax file may also be worth contacting.
>>>
> I don't really know cobol anymore (I had cobol years ago, and I'm glad
> not to need to mess with it).
> I note that a "-" in column#7 fixes the problem; in other words,
> syntax/cobol.vim expects you to use cobol's continuation character.
>
> Regards,
> Chip Campbell
COBOL's continuation character should only be used (in the several COBOL
standards which I used professionally starting in 1969), if you break a
line in the middle of a word (Vim would say "in the middle of a
keyword"), or in the middle of a non-numeric literal (which is the COBOL
name for a string); in the latter case the double-quotes must not be
closed in column 72 but they must be reopened in colum 12. If you break
the line between words in a sentence, you should not use a continuation
line.
Now a simple way to check whether the * is correctly highlighted: * as
an operator (in columns 12 to 72 of a sentence in the PROCEDURE
DIVISION) is a word (which must be separated by spaces before and after
it except after an opening paren, or before a closing paren, comma,
semicolon or period) and it is equivalent to the slightly more verbose
TIMES or, even longer, MULTIPLIED[ BY]. If an "operator" asterisk is
shown in Error highlight, what happens if you replace just the asterisk
_character_ by the word "TIMES"? (Remember, word-separating spaces or
punctuation must already exist around the asterisk.)
Using the example above, and remembering that I worked on computers
which didn't know about lowercase,
COMPUTE A = ((B + C) / D)
* ((E * F) + G).
is a perfectly valid COBOL sentence provided that it appears in the
PROCEDURE DIVISION, both its lines start no earlier than column 12 (and
end no later than column 72 in fixed-form COBOL), and that all the data
items it uses are defined in the DATA DIVISION. No column-7 dash is
required, but there must be a period followed by a space somewhere (no
later than the end of the paragraph, which is just before the next
procedure-name i.e. label) to end the sentence. In slightly more verbose
COBOL, the above sentence is equivalent to
COMPUTE A = ((B PLUS C) DIVIDED BY D)
MULTIPLIED BY ((E TIMES F) PLUS G.
or even, in a somewhat older version of the standard where the COMPUTE
verb didn't exist, and with two more WORKING-STORAGE SECTION data items
of appropriate PICTURE:
ADD B C GIVING TEMP-1.
DIVIDE D INTO TEMP-1.
MULTIPLY E F GIVING TEMP-2.
ADD G TO TEMP-2.
MULTIPLY TEMP-1 TEMP-2 GIVING A.
Best regards,
Tony.
--
"A fractal is by definition a set for which the Hausdorff Besicovitch
dimension strictly exceeds the topological dimension."
-- Mandelbrot, "The Fractal Geometry of Nature"
--~--~---------~--~----~------------~-------~--~----~
You received this message from the "vim_use" maillist.
For more information, visit http://www.vim.org/maillist.php
-~----------~----~----~----~------~----~------~--~---
