The code to remove commas was included mainly out of habit.
Populating selection lists and combo boxes requires this awareness,
as commas are used to separate list items.
I haven't always used the Export statement when storing to a text
file. Maybe I should from now on, or delimit with some odd character.
Regarding the "for each table1 of table2" question, I think my
preference goes back to that "feeling of control" I mentioned;
whether to select your index or let it be selected for you. I've
heard users say, "Can you speed this up?", too many times in the past.
And what about having to take the "long way around" just because an
index doesn't exist to match your needs. For example, a "company-
partnum-shipdate" index in shipdtl would be helpful.
Sounds impossible, but I remember hearing from someone at Vantage
support that we would be able to add our own to Vantage at some
point, without the dreaded "incorrect CRC value" error.
Lonnie
Populating selection lists and combo boxes requires this awareness,
as commas are used to separate list items.
I haven't always used the Export statement when storing to a text
file. Maybe I should from now on, or delimit with some odd character.
Regarding the "for each table1 of table2" question, I think my
preference goes back to that "feeling of control" I mentioned;
whether to select your index or let it be selected for you. I've
heard users say, "Can you speed this up?", too many times in the past.
And what about having to take the "long way around" just because an
index doesn't exist to match your needs. For example, a "company-
partnum-shipdate" index in shipdtl would be helpful.
Sounds impossible, but I remember hearing from someone at Vantage
support that we would be able to add our own to Vantage at some
point, without the dreaded "incorrect CRC value" error.
Lonnie
--- In vantage@y..., Thad Jacobs <tjacobs@k...> wrote:
> Lonnie, I was also under the impression for a while that CSV
wouldn't store
> our multi-line, comma and quote containing memo fields until I
tried it one
> day, and to my surprise, it worked, and I got to wondering why?
>
> I came to the conclusion that modern CSV import code, once it sees
a new
> field beginning with a " looks for a ", to denote the end of a
field. All
> (or shall I say most) characters until the end of the field are
just plain
> written to the file. Quotes within strings are generally stored in
CSV as
> "" so say you have a description that says:
>
> Thaddeus Jacobs, 6'1", Brown hair, brown eyes
>
> When its exported it will be converted to
>
> ...,"Thaddeus Jacobs, 6'1"", brn hair, brn eyes",....
>
> /*Test Hypothesis...*/
> def var teststring as char initial "Thaddeus Jacobs, 6'1"", brn
hair, brn
> eyes".
>
> message teststring.
>
> output to c:\vntgwrk\csvtest binary no-convert.
> export delimiter "," teststring.
>
> output close.
>
> input from c:\vntgwrk\csvtest.
>
> import delimiter "," teststring.
> input from terminal.
> message teststring.
> /*end test*/
>
> You can actually export a multiline memo field in CSV format as
well. You
> just have to make sure that all quotes within strings are converted
to
> double quotes (which the progress EXPORT statement already does).
>
> Long live CSV!!!
>
> ~Thaddeus
yahoo.com/info/terms/