As a normal practice I wouldn't be posting example code onto the forum unless someone was needing help about a specific problem. We did have a question come in from a customer this week asking for a new conversion code be added to UniData. This brought up a couple of interesting subjects. The customer was not aware that you could write your own code following a template to create your own conversion codes. It took me some time to find where this was buried in the manuals I can understand why they would not have come across it.
So just in case others were not aware you could do this I thought I'd share with you how you do this.
SUBROUTINE SUBNAME( return_val, STATUS, input_val, TYPE )
***************************************************************
*
* Brief functional description
*
* (c) Copyright notice
**************************************************************
*
* Detailed description
*
* Usage, such as return_val = OCONV(input_val,"USUBNAME")
*
**************************************************************
* processing code area:
* Arguments:
* return_val = final data to return.
* STATUS = 0 for success, other for failure.
* input_val= data to be converted or processed.
* TYPE: 0 for ICONV, 1 for OCONV.
*
RETURN
END
The customer was asking for some conversion codes to be added around UTL time stamps. So I'm attaching my example code I provided to the customer to demonstrate some ideas around around UTL conversions and would welcome any comments around UTL and in particular handling time zones.
So my example does the following it takes the input of UTL string in for format 2019-06-06T08:58:42.1878669+09:00 or 2019-06-06T08:58:42+09:00 and it will return on ICONV the internal date and internal time of the UTC time. So ICONV(" 2019-06-06T08:58:42.1878669+09:00", "UUTCL") returns 19912 86322.1878669
where 19912 = 17/07/2022 and 86322.1878669 = 23:58:42
In terms of the OCONV it takes 19912 86322.1878669 and then converts that into UTL time stamp of the timezone that the server is running in, so on my Windows Laptop (being UK based) OCONV("19912 86322.1878669","UUTCL") returns 2022-07-07T23:58:42.1878669+00:00 (Dublin, Edinburgh, Lisbon, London) if I use the same example on one of our Denver based unix servers it returns 2022-07-07T23:58:42.1878669-06:00 (MDT Mountain Daylight Time)
So in terms of the attachments with post the user conversion program is called UTCL, there is a testing program called TIME.TEST and there is a look up file called TIMEZONES, the TIMEZONES and D_TIMEZONES files are the Windows / Linux Versions and the TIMEZONES.UX and D_TIMEZONES.UK files are for AIX / HPUX and Solaris.
I hope they provide an example to those of you who didn't know you could do this and also I'd be interested to hear your thoughts about how you want to determine a timezone programmatically.
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
Page 1 / 1
As a normal practice I wouldn't be posting example code onto the forum unless someone was needing help about a specific problem. We did have a question come in from a customer this week asking for a new conversion code be added to UniData. This brought up a couple of interesting subjects. The customer was not aware that you could write your own code following a template to create your own conversion codes. It took me some time to find where this was buried in the manuals I can understand why they would not have come across it.
So just in case others were not aware you could do this I thought I'd share with you how you do this.
SUBROUTINE SUBNAME( return_val, STATUS, input_val, TYPE )
***************************************************************
*
* Brief functional description
*
* (c) Copyright notice
**************************************************************
*
* Detailed description
*
* Usage, such as return_val = OCONV(input_val,"USUBNAME")
*
**************************************************************
* processing code area:
* Arguments:
* return_val = final data to return.
* STATUS = 0 for success, other for failure.
* input_val= data to be converted or processed.
* TYPE: 0 for ICONV, 1 for OCONV.
*
RETURN
END
The customer was asking for some conversion codes to be added around UTL time stamps. So I'm attaching my example code I provided to the customer to demonstrate some ideas around around UTL conversions and would welcome any comments around UTL and in particular handling time zones.
So my example does the following it takes the input of UTL string in for format 2019-06-06T08:58:42.1878669+09:00 or 2019-06-06T08:58:42+09:00 and it will return on ICONV the internal date and internal time of the UTC time. So ICONV(" 2019-06-06T08:58:42.1878669+09:00", "UUTCL") returns 19912 86322.1878669
where 19912 = 17/07/2022 and 86322.1878669 = 23:58:42
In terms of the OCONV it takes 19912 86322.1878669 and then converts that into UTL time stamp of the timezone that the server is running in, so on my Windows Laptop (being UK based) OCONV("19912 86322.1878669","UUTCL") returns 2022-07-07T23:58:42.1878669+00:00 (Dublin, Edinburgh, Lisbon, London) if I use the same example on one of our Denver based unix servers it returns 2022-07-07T23:58:42.1878669-06:00 (MDT Mountain Daylight Time)
So in terms of the attachments with post the user conversion program is called UTCL, there is a testing program called TIME.TEST and there is a look up file called TIMEZONES, the TIMEZONES and D_TIMEZONES files are the Windows / Linux Versions and the TIMEZONES.UX and D_TIMEZONES.UK files are for AIX / HPUX and Solaris.
I hope they provide an example to those of you who didn't know you could do this and also I'd be interested to hear your thoughts about how you want to determine a timezone programmatically.
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
So just in case others were not aware you could do this I thought I'd share with you how you do this.
SUBROUTINE SUBNAME( return_val, STATUS, input_val, TYPE )
***************************************************************
*
* Brief functional description
*
* (c) Copyright notice
**************************************************************
*
* Detailed description
*
* Usage, such as return_val = OCONV(input_val,"USUBNAME")
*
**************************************************************
* processing code area:
* Arguments:
* return_val = final data to return.
* STATUS = 0 for success, other for failure.
* input_val= data to be converted or processed.
* TYPE: 0 for ICONV, 1 for OCONV.
*
RETURN
END
The customer was asking for some conversion codes to be added around UTL time stamps. So I'm attaching my example code I provided to the customer to demonstrate some ideas around around UTL conversions and would welcome any comments around UTL and in particular handling time zones.
So my example does the following it takes the input of UTL string in for format 2019-06-06T08:58:42.1878669+09:00 or 2019-06-06T08:58:42+09:00 and it will return on ICONV the internal date and internal time of the UTC time. So ICONV(" 2019-06-06T08:58:42.1878669+09:00", "UUTCL") returns 19912 86322.1878669
where 19912 = 17/07/2022 and 86322.1878669 = 23:58:42
In terms of the OCONV it takes 19912 86322.1878669 and then converts that into UTL time stamp of the timezone that the server is running in, so on my Windows Laptop (being UK based) OCONV("19912 86322.1878669","UUTCL") returns 2022-07-07T23:58:42.1878669+00:00 (Dublin, Edinburgh, Lisbon, London) if I use the same example on one of our Denver based unix servers it returns 2022-07-07T23:58:42.1878669-06:00 (MDT Mountain Daylight Time)
So in terms of the attachments with post the user conversion program is called UTCL, there is a testing program called TIME.TEST and there is a look up file called TIMEZONES, the TIMEZONES and D_TIMEZONES files are the Windows / Linux Versions and the TIMEZONES.UX and D_TIMEZONES.UK files are for AIX / HPUX and Solaris.
I hope they provide an example to those of you who didn't know you could do this and also I'd be interested to hear your thoughts about how you want to determine a timezone programmatically.
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
Q1. Why NOT post code?
Q2. Is not almost everything a specific problem?
Please clarify.
------------------------------
Bill Brutzman
IT Manager
Hk Metalcraft Manufacturing Corporation
Lodi NJ US
------------------------------
Jonathan:
Q1. Why NOT post code?
Q2. Is not almost everything a specific problem?
Please clarify.
------------------------------
Bill Brutzman
IT Manager
Hk Metalcraft Manufacturing Corporation
Lodi NJ US
------------------------------
Q1. Why NOT post code?
Q2. Is not almost everything a specific problem?
Please clarify.
------------------------------
Bill Brutzman
IT Manager
Hk Metalcraft Manufacturing Corporation
Lodi NJ US
------------------------------
Thanks
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
Bill ... How would you programmaticly get the system to find out what post code it is in. Finding the systems time zone can be found out via a Unix or Windows command... or you could even enquire on the system or users time zone environment variable. I am not aware of any easy way to get the post code that a system or user is in and then how would you tie translate that to the timezone the post code is in.
Thanks
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
Thanks
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
Thanks for responding but _ my question Q1 is much more general.
I interpreted your message as something like _ please do not reveal any (UniBasic) source code on this forum. Perhaps the intent of your comment is related to intellectual property (legal) considerations. Please clarify.
--Bill
------------------------------
Bill Brutzman
IT Manager
Hk Metalcraft Manufacturing Corporation
Lodi NJ US
------------------------------
Jonathan:
Thanks for responding but _ my question Q1 is much more general.
I interpreted your message as something like _ please do not reveal any (UniBasic) source code on this forum. Perhaps the intent of your comment is related to intellectual property (legal) considerations. Please clarify.
--Bill
------------------------------
Bill Brutzman
IT Manager
Hk Metalcraft Manufacturing Corporation
Lodi NJ US
------------------------------
Thanks for responding but _ my question Q1 is much more general.
I interpreted your message as something like _ please do not reveal any (UniBasic) source code on this forum. Perhaps the intent of your comment is related to intellectual property (legal) considerations. Please clarify.
--Bill
------------------------------
Bill Brutzman
IT Manager
Hk Metalcraft Manufacturing Corporation
Lodi NJ US
------------------------------
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
As a normal practice I wouldn't be posting example code onto the forum unless someone was needing help about a specific problem. We did have a question come in from a customer this week asking for a new conversion code be added to UniData. This brought up a couple of interesting subjects. The customer was not aware that you could write your own code following a template to create your own conversion codes. It took me some time to find where this was buried in the manuals I can understand why they would not have come across it.
So just in case others were not aware you could do this I thought I'd share with you how you do this.
SUBROUTINE SUBNAME( return_val, STATUS, input_val, TYPE )
***************************************************************
*
* Brief functional description
*
* (c) Copyright notice
**************************************************************
*
* Detailed description
*
* Usage, such as return_val = OCONV(input_val,"USUBNAME")
*
**************************************************************
* processing code area:
* Arguments:
* return_val = final data to return.
* STATUS = 0 for success, other for failure.
* input_val= data to be converted or processed.
* TYPE: 0 for ICONV, 1 for OCONV.
*
RETURN
END
The customer was asking for some conversion codes to be added around UTL time stamps. So I'm attaching my example code I provided to the customer to demonstrate some ideas around around UTL conversions and would welcome any comments around UTL and in particular handling time zones.
So my example does the following it takes the input of UTL string in for format 2019-06-06T08:58:42.1878669+09:00 or 2019-06-06T08:58:42+09:00 and it will return on ICONV the internal date and internal time of the UTC time. So ICONV(" 2019-06-06T08:58:42.1878669+09:00", "UUTCL") returns 19912 86322.1878669
where 19912 = 17/07/2022 and 86322.1878669 = 23:58:42
In terms of the OCONV it takes 19912 86322.1878669 and then converts that into UTL time stamp of the timezone that the server is running in, so on my Windows Laptop (being UK based) OCONV("19912 86322.1878669","UUTCL") returns 2022-07-07T23:58:42.1878669+00:00 (Dublin, Edinburgh, Lisbon, London) if I use the same example on one of our Denver based unix servers it returns 2022-07-07T23:58:42.1878669-06:00 (MDT Mountain Daylight Time)
So in terms of the attachments with post the user conversion program is called UTCL, there is a testing program called TIME.TEST and there is a look up file called TIMEZONES, the TIMEZONES and D_TIMEZONES files are the Windows / Linux Versions and the TIMEZONES.UX and D_TIMEZONES.UK files are for AIX / HPUX and Solaris.
I hope they provide an example to those of you who didn't know you could do this and also I'd be interested to hear your thoughts about how you want to determine a timezone programmatically.
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
So just in case others were not aware you could do this I thought I'd share with you how you do this.
SUBROUTINE SUBNAME( return_val, STATUS, input_val, TYPE )
***************************************************************
*
* Brief functional description
*
* (c) Copyright notice
**************************************************************
*
* Detailed description
*
* Usage, such as return_val = OCONV(input_val,"USUBNAME")
*
**************************************************************
* processing code area:
* Arguments:
* return_val = final data to return.
* STATUS = 0 for success, other for failure.
* input_val= data to be converted or processed.
* TYPE: 0 for ICONV, 1 for OCONV.
*
RETURN
END
The customer was asking for some conversion codes to be added around UTL time stamps. So I'm attaching my example code I provided to the customer to demonstrate some ideas around around UTL conversions and would welcome any comments around UTL and in particular handling time zones.
So my example does the following it takes the input of UTL string in for format 2019-06-06T08:58:42.1878669+09:00 or 2019-06-06T08:58:42+09:00 and it will return on ICONV the internal date and internal time of the UTC time. So ICONV(" 2019-06-06T08:58:42.1878669+09:00", "UUTCL") returns 19912 86322.1878669
where 19912 = 17/07/2022 and 86322.1878669 = 23:58:42
In terms of the OCONV it takes 19912 86322.1878669 and then converts that into UTL time stamp of the timezone that the server is running in, so on my Windows Laptop (being UK based) OCONV("19912 86322.1878669","UUTCL") returns 2022-07-07T23:58:42.1878669+00:00 (Dublin, Edinburgh, Lisbon, London) if I use the same example on one of our Denver based unix servers it returns 2022-07-07T23:58:42.1878669-06:00 (MDT Mountain Daylight Time)
So in terms of the attachments with post the user conversion program is called UTCL, there is a testing program called TIME.TEST and there is a look up file called TIMEZONES, the TIMEZONES and D_TIMEZONES files are the Windows / Linux Versions and the TIMEZONES.UX and D_TIMEZONES.UK files are for AIX / HPUX and Solaris.
I hope they provide an example to those of you who didn't know you could do this and also I'd be interested to hear your thoughts about how you want to determine a timezone programmatically.
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
What is the best way to pass parameters into a user-defined conversion via the conversion attribute <3> of a dictionary? (<7> or <8> is A- or S-items).
As an example, consider the "MD" conversion: MD0, MD20P.
Or consider all the various options for a "D" conversion: D [n] [*m] [ s] [fmt [[f1, f2, f3, f4, f5]]] [E] [L]
I've wanted to do do things like that with my own user conversion subroutines.
------------------------------
Chuck Stevenson
DBA / SW Developer
Pomeroy
US
------------------------------
Jonathon,
What is the best way to pass parameters into a user-defined conversion via the conversion attribute <3> of a dictionary? (<7> or <8> is A- or S-items).
As an example, consider the "MD" conversion: MD0, MD20P.
Or consider all the various options for a "D" conversion: D [n] [*m] [ s] [fmt [[f1, f2, f3, f4, f5]]] [E] [L]
I've wanted to do do things like that with my own user conversion subroutines.
------------------------------
Chuck Stevenson
DBA / SW Developer
Pomeroy
US
------------------------------
What is the best way to pass parameters into a user-defined conversion via the conversion attribute <3> of a dictionary? (<7> or <8> is A- or S-items).
As an example, consider the "MD" conversion: MD0, MD20P.
Or consider all the various options for a "D" conversion: D [n] [*m] [ s] [fmt [[f1, f2, f3, f4, f5]]] [E] [L]
I've wanted to do do things like that with my own user conversion subroutines.
------------------------------
Chuck Stevenson
DBA / SW Developer
Pomeroy
US
------------------------------
The example I highlighted was for UniData. UniData does not have the Pick Style dictionaries where 7 and 8 are the conversion and correlative. How to do to write one of the programs and use it is covered in the 'User Exits' part of the 'Using UniData' manual. The first part of any conversion code so M,D etc all tell UniData / UniVerse how to process the conversion.
With a conversion code of Uuser_exit The user_exit just specifies which catalogued program to use it does not have any other options as was never designed to do so. (You may know this , others might not) Originally the user_exit code was put in place to provide Pick style calls such as U50BB. If you sort SYS_BP in UniData you'll see 30 or so programs at the top which are user_exit replacements for old Pick style ones. Given that the user_exit template is pre-defined as well, I do not see an easy way to specify anything other than the user_exit in the conversion code field.
All that being said I could probably think of a few ways of doing what you're asking but in terms of the adding stuff to the user exit code itself to be read by the function program then there is no way to do this.
Thanks,
Jonathan
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
Hi Chuck,
The example I highlighted was for UniData. UniData does not have the Pick Style dictionaries where 7 and 8 are the conversion and correlative. How to do to write one of the programs and use it is covered in the 'User Exits' part of the 'Using UniData' manual. The first part of any conversion code so M,D etc all tell UniData / UniVerse how to process the conversion.
With a conversion code of Uuser_exit The user_exit just specifies which catalogued program to use it does not have any other options as was never designed to do so. (You may know this , others might not) Originally the user_exit code was put in place to provide Pick style calls such as U50BB. If you sort SYS_BP in UniData you'll see 30 or so programs at the top which are user_exit replacements for old Pick style ones. Given that the user_exit template is pre-defined as well, I do not see an easy way to specify anything other than the user_exit in the conversion code field.
All that being said I could probably think of a few ways of doing what you're asking but in terms of the adding stuff to the user exit code itself to be read by the function program then there is no way to do this.
Thanks,
Jonathan
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
The example I highlighted was for UniData. UniData does not have the Pick Style dictionaries where 7 and 8 are the conversion and correlative. How to do to write one of the programs and use it is covered in the 'User Exits' part of the 'Using UniData' manual. The first part of any conversion code so M,D etc all tell UniData / UniVerse how to process the conversion.
With a conversion code of Uuser_exit The user_exit just specifies which catalogued program to use it does not have any other options as was never designed to do so. (You may know this , others might not) Originally the user_exit code was put in place to provide Pick style calls such as U50BB. If you sort SYS_BP in UniData you'll see 30 or so programs at the top which are user_exit replacements for old Pick style ones. Given that the user_exit template is pre-defined as well, I do not see an easy way to specify anything other than the user_exit in the conversion code field.
All that being said I could probably think of a few ways of doing what you're asking but in terms of the adding stuff to the user exit code itself to be read by the function program then there is no way to do this.
Thanks,
Jonathan
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
Yes, that's what I thought. I've always wanted it but didn't think there was a way. Hope springs eternal.
One way you could do it: Allow a syntax like U[userexit]:[options]
[options] would probably be as cryptic as with the out-of-the-box conversion codes.
The existence of :[options] would assume another input argument argument to be passed to the userexit subroutine, or let the 2nd argument, which is currently the status outarg, do double duty as options inarg.
SUBROUTINE Unnnn (answer, status, data, type)
becomes:
SUBROUTINE Unnnn (answer, status, data, type, options)
or
SUBROUTINE Unnnn (answer, optionsIN.statusOUT, data, type)
For UV users listening in, this is best described in appendix A of the "Pick Users Guide" pdf, part of the complete UV Documentation Suite.
------------------------------
Chuck Stevenson
DBA / SW Developer
Pomeroy
US
------------------------------
Jonathon,
Yes, that's what I thought. I've always wanted it but didn't think there was a way. Hope springs eternal.
One way you could do it: Allow a syntax like U[userexit]:[options]
[options] would probably be as cryptic as with the out-of-the-box conversion codes.
The existence of :[options] would assume another input argument argument to be passed to the userexit subroutine, or let the 2nd argument, which is currently the status outarg, do double duty as options inarg.
SUBROUTINE Unnnn (answer, status, data, type)
becomes:
SUBROUTINE Unnnn (answer, status, data, type, options)
or
SUBROUTINE Unnnn (answer, optionsIN.statusOUT, data, type)
For UV users listening in, this is best described in appendix A of the "Pick Users Guide" pdf, part of the complete UV Documentation Suite.
------------------------------
Chuck Stevenson
DBA / SW Developer
Pomeroy
US
------------------------------
Yes, that's what I thought. I've always wanted it but didn't think there was a way. Hope springs eternal.
One way you could do it: Allow a syntax like U[userexit]:[options]
[options] would probably be as cryptic as with the out-of-the-box conversion codes.
The existence of :[options] would assume another input argument argument to be passed to the userexit subroutine, or let the 2nd argument, which is currently the status outarg, do double duty as options inarg.
SUBROUTINE Unnnn (answer, status, data, type)
becomes:
SUBROUTINE Unnnn (answer, status, data, type, options)
or
SUBROUTINE Unnnn (answer, optionsIN.statusOUT, data, type)
For UV users listening in, this is best described in appendix A of the "Pick Users Guide" pdf, part of the complete UV Documentation Suite.
------------------------------
Chuck Stevenson
DBA / SW Developer
Pomeroy
US
------------------------------
As two trains of thought:
1) You can always pass a dynamic array as an argument - e.g. both the data and extended conversion options in separate elements. A Multivalue data set in field1 and conversion options in field 2 for example.
2) CALLC is also an option for anything particularly awkward or performance sensitive.
I have used CALLC many times in the past for tasks such as:
- Time-critical byte-level processing of data blocks.
- TCP communications (before this was made available in BASIC)
- File-transfer (MV-to-Other MV - long before FTP and TCP were available) using peer-to-peer. MV to IBM mainframe ('Pick Angel' of memory)
- Driving auto-diallers, modems, interactive voice synthesisers and a cash-register interface (amongst others).
Regards
JJ
------------------------------
John Jenkins
Thame, Oxfordshire
------------------------------
Chuck,
As two trains of thought:
1) You can always pass a dynamic array as an argument - e.g. both the data and extended conversion options in separate elements. A Multivalue data set in field1 and conversion options in field 2 for example.
2) CALLC is also an option for anything particularly awkward or performance sensitive.
I have used CALLC many times in the past for tasks such as:
Regards
JJ
------------------------------
John Jenkins
Thame, Oxfordshire
------------------------------
As two trains of thought:
1) You can always pass a dynamic array as an argument - e.g. both the data and extended conversion options in separate elements. A Multivalue data set in field1 and conversion options in field 2 for example.
2) CALLC is also an option for anything particularly awkward or performance sensitive.
I have used CALLC many times in the past for tasks such as:
- Time-critical byte-level processing of data blocks.
- TCP communications (before this was made available in BASIC)
- File-transfer (MV-to-Other MV - long before FTP and TCP were available) using peer-to-peer. MV to IBM mainframe ('Pick Angel' of memory)
- Driving auto-diallers, modems, interactive voice synthesisers and a cash-register interface (amongst others).
Regards
JJ
------------------------------
John Jenkins
Thame, Oxfordshire
------------------------------
I appreciate your thoughtful reply. Some further thoughts or explanation of the problem. 1st, I'm on UV, not UD. I realize the thread started with a UD discussion, but if I recall UD correctly, what I'm calling a shortcoming is essentially the same.
From your CALLC comments, it sounds like you're addressing the general case, which includes old Pick-style user exits in Pick-style Procs.
My interest is user-defined conversion codes in dictionaries.
"You can always pass a dynamic array as an argument - e.g. both the data and extended conversion options in separate elements. A Multivalue data set in field1 and conversion options in field 2 for example."
For ICONV/OCONV user exits, the syntax is restricted to:
SUBROUTINE Unnnn (answer, status, data, type)
Retrieve populates & reads the arguments when it calls the subroutine.
In a dictionary, one could monkey with what's passed in the 3rd argument, to include options not just data. But system delimiters must be avoided. Something like:
01: I
02: CATS( RAW.DATA, REUSE( ":[options]" )
03:U[user exit]
but at that point it's easier, clearer to just use a garden variety I-descriptor subroutine. Which is what we do today.
------------------------------
Chuck Stevenson
DBA / SW Developer
Pomeroy
US
------------------------------
Bill ... How would you programmaticly get the system to find out what post code it is in. Finding the systems time zone can be found out via a Unix or Windows command... or you could even enquire on the system or users time zone environment variable. I am not aware of any easy way to get the post code that a system or user is in and then how would you tie translate that to the timezone the post code is in.
Thanks
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
Thanks
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
------------------------------
Stacy Mecklenburg
Senior Developer
Software Solutions Unlimited
San Anselmo CA US
------------------------------
Chuck,
As two trains of thought:
1) You can always pass a dynamic array as an argument - e.g. both the data and extended conversion options in separate elements. A Multivalue data set in field1 and conversion options in field 2 for example.
2) CALLC is also an option for anything particularly awkward or performance sensitive.
I have used CALLC many times in the past for tasks such as:
Regards
JJ
------------------------------
John Jenkins
Thame, Oxfordshire
------------------------------
As two trains of thought:
1) You can always pass a dynamic array as an argument - e.g. both the data and extended conversion options in separate elements. A Multivalue data set in field1 and conversion options in field 2 for example.
2) CALLC is also an option for anything particularly awkward or performance sensitive.
I have used CALLC many times in the past for tasks such as:
- Time-critical byte-level processing of data blocks.
- TCP communications (before this was made available in BASIC)
- File-transfer (MV-to-Other MV - long before FTP and TCP were available) using peer-to-peer. MV to IBM mainframe ('Pick Angel' of memory)
- Driving auto-diallers, modems, interactive voice synthesisers and a cash-register interface (amongst others).
Regards
JJ
------------------------------
John Jenkins
Thame, Oxfordshire
------------------------------
"With a conversion code of Uuser_exit The user_exit just specifies which catalogued program to use it does not have any other options as was never designed to do so. "
UniVerse does have user-defined conversion options. For example, you can specify the conversion as "U1234ABCD" and UniVerse will call $1234 and pass in "U1234ABCD".
For example:
function u1234(branch, value, type)
params = branch[6, len(branch)]
etc.
We use dozens of user-defined conversions and most use such parameters.
Henry
------------------------------
Henry Unger
President
Hitech Systems Inc
Encino CA US
------------------------------
Jonathan - I believe "Bill" was asking about your statement..."As a normal practice I wouldn't be posting example code onto the forum unless" - why? As a developer's forum, this is where I go to bounce ideas/code with other developers, we share code snippets and the like. Yes, we have to be mindful of relevant IP and NDAs, but a snippet of code is not a violation, especially in this case of an OCONV/ICONV hack. I, as well, am curious, "as a normal practice", why not post code? A lot of colleagues on this site, post, we analyze, discuss, bounce ideas off each other, etc. This is more or less our sandbox to play discuss/develop ideas/concepts and the like.
------------------------------
Stacy Mecklenburg
Senior Developer
Software Solutions Unlimited
San Anselmo CA US
------------------------------
------------------------------
Stacy Mecklenburg
Senior Developer
Software Solutions Unlimited
San Anselmo CA US
------------------------------
What i should have said was i wouldn't normally post code on the forum about something unless someone has directly asked a question about it ... however a customer contacted support with a question so i thought I'd share with this everyone as others may not know as well
Thanks
------------------------------
Jonathan Smith
UniData ATS
Rocket Support
------------------------------
Sign up
Already have an account? Login
Welcome to the Rocket Forum!
Please log in or register:
Employee Login | Registration Member Login | RegistrationEnter your E-mail address. We'll send you an e-mail with instructions to reset your password.
77 4th Avenue
Waltham, MA 02451 USA
