ibm-netrexx

Please help with Java problem in NetRexx

Classic

List

Threaded

7 messages Options

Jeff Hennick-3

Please help with Java problem in NetRexx

I am currently working to bring CMS Pipelines Base64 stages 64Encode and 64Decode into NetRexx Pipelines. I am using the java.util.Base64 class to do the work. Encode works great, but decode gives gibberish, and different gibberish each time it is run.

A note on names. CMS calls the stages 64Encode and 64Decode. NetRexx objects to class names beginning with with a digit, so I am using Encode64 and Decode64, and adding compiler based aliases of the CMS names.

Here is what works for Encode

options nostrictcase nostrictargs nostrictsignal
--package org.netrexx.njpipes.stages
import org.netrexx.njpipes.pipes.
import java.util.Base64
class encode64 extends stage
method run()
enc = Base64.getMIMEEncoder()
loop label Main forever
    record_in = Rexx peekto()
    ba = record_in.toByteArray()
    opr = Rexx enc.encodeToString(ba)
    output(opr)
    readto()
catch stageError
    rc = rc()
end Main
Exit(rc*(rc<>12))

Here is what does not work for Decode

options nostrictcase nostrictargs nostrictsignal
--package org.netrexx.njpipes.stages
import org.netrexx.njpipes.pipes.
import java.util.Base64
class decode64 extends stage
method run()
dec = Base64.getMIMEDecoder()
loop label Main forever
    record_in = peekto().toString()
    ob = dec.decode(record_in)
    opr = Rexx ob
    output(opr)
    readto()
catch stageError
    rc = rc()
end Main
Exit(rc*(rc<>12))

Here is the trace, and a second run's result, too

PS C:\Users\Jeff\documents\pipe tests> java ct
    --- decode64.nrx [Thread-7,njPipes]
 33 *=* method run()
 35 *=*   dec = Base64.getMIMEDecoder()
    >v> dec "java.util.Base64$Decoder@11c1b90"
 37 *=*   loop label Main forever
 39 *=*     record_in = peekto().toString()
    >v> record_in "SGVsbG8sIEJhc2U2NCE="
 41 *=*     ob = dec.decode(record_in)
    >>> "SGVsbG8sIEJhc2U2NCE="
    >v> ob "[B@18bd538"
 43 *=*     opr = Rexx ob
    >v> opr "[B@18bd538"
 45 *=*     output(opr)
    >>> "[B@18bd538"
 49 *=*   catch stageError
 50 *=*     rc = rc()
    >v> rc "12"
 51 *=*   end Main
 53 *=*   Exit(rc*(rc<>12))
    >>> "0"
# 2 64decode  **FAIL**  at rec 1 col 1. Less:
Actual:
[B@18bd538
Expected:
Hello, Base64!

# 2 64decode  **FAIL**  at rec 1 col 1. Less:
Actual:
[B@57e382
Expected:
Hello, Base64!

Notice that the two Actuals are different, even though run one right after the other.

I have tried various combinations of casting things to and from Rexx, string, byte[] with no help. I keep getting the [hidden email]-type results. (The Encode took some such shenanigans to coax it to work.)

Any help pointers you can give would be gratefully received. Thanks!

Jeff Hennick

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

Jason Martin

Re: Please help with Java problem in NetRexx

Does this help?

In Base64 encoding, the length of output-encoded String must be a multiple of three. If not, the output will be padded with additional pad characters (=).

Upon decoding, these extra padding characters will be discarded. To dig deeper into padding in Base64, check out this detailed answer on Stack Overflow.

From:

https://www.baeldung.com/java-base64-encode-and-decode

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

rvjansen

Re: Please help with Java problem in NetRexx

In reply to this post by Jeff Hennick-3

Hi Jeff,

These are references - like memory pointers. They need some dereferencing. Let me try something.

René.

On 15 Oct 2020, at 21:37, Jeff Hennick <[hidden email]> wrote:

I am currently working to bring CMS Pipelines Base64 stages 64Encode and 64Decode into NetRexx Pipelines. I am using the java.util.Base64 class to do the work. Encode works great, but decode gives gibberish, and different gibberish each time it is run.

A note on names. CMS calls the stages 64Encode and 64Decode. NetRexx objects to class names beginning with with a digit, so I am using Encode64 and Decode64, and adding compiler based aliases of the CMS names.

Here is what works for Encode

options nostrictcase nostrictargs nostrictsignal
--package org.netrexx.njpipes.stages
import org.netrexx.njpipes.pipes.
import java.util.Base64
class encode64 extends stage
method run()
enc = Base64.getMIMEEncoder()
loop label Main forever
    record_in = Rexx peekto()
    ba = record_in.toByteArray()
    opr = Rexx enc.encodeToString(ba)
    output(opr)
    readto()
catch stageError
    rc = rc()
end Main
Exit(rc*(rc<>12))

Here is what does not work for Decode

options nostrictcase nostrictargs nostrictsignal
--package org.netrexx.njpipes.stages
import org.netrexx.njpipes.pipes.
import java.util.Base64
class decode64 extends stage
method run()
dec = Base64.getMIMEDecoder()
loop label Main forever
    record_in = peekto().toString()
    ob = dec.decode(record_in)
    opr = Rexx ob
    output(opr)
    readto()
catch stageError
    rc = rc()
end Main
Exit(rc*(rc<>12))

Here is the trace, and a second run's result, too

PS C:\Users\Jeff\documents\pipe tests> java ct --- decode64.nrx [Thread-7,njPipes] 33 *=* method run() 35 *=* dec = Base64.getMIMEDecoder() >v> dec "java.util.Base64$Decoder@11c1b90" 37 *=* loop label Main forever 39 *=* record_in = peekto().toString() >v> record_in "SGVsbG8sIEJhc2U2NCE=" 41 *=* ob = dec.decode(record_in) >>> "SGVsbG8sIEJhc2U2NCE=" >v> ob "[B@18bd538" 43 *=* opr = Rexx ob >v> opr "[B@18bd538" 45 *=* output(opr) >>> "[B@18bd538" 49 *=* catch stageError 50 *=* rc = rc() >v> rc "12" 51 *=* end Main 53 *=* Exit(rc*(rc<>12)) >>> "0" # 2 64decode **FAIL** at rec 1 col 1. Less: Actual: [B@18bd538 Expected: Hello, Base64!

# 2 64decode **FAIL** at rec 1 col 1. Less: Actual: [B@57e382 Expected: Hello, Base64!

Notice that the two Actuals are different, even though run one right after the other.

I have tried various combinations of casting things to and from Rexx, string, byte[] with no help. I keep getting the [hidden email]-type results. (The Encode took some such shenanigans to coax it to work.)

Any help pointers you can give would be gratefully received. Thanks!

Jeff Hennick

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

Jeff Hennick-3

Re: Please help with Java problem in NetRexx

In reply to this post by Jason Martin

Thank you, Jason.

Using the MIME-type encoding, specifically in the Java (and implicitly in CMS), handles all that. (There can be additional bit level padding with 0s to get to a full byte.) And it is in the _working_ Encoding that all this is important.

The input string in the test for Decoding, line 39, is "SGVsbG8sIEJhc2U2NCE=", this is the result, with the padding =, from the Encoding of "Hello, Base64!".

For Encoding tests, I also use these short ones: @!*, @!, and @; with the results being QCEq, QCE=, and QA==. (I got all these test phrases from some web page, and the results of encoding agree with theirs.)

I am still left hanging with the incorrect and varying decoding results. I have tried running the short examples too, but still get that enigmatic [B@57e382 - type result from them, too.

On 10/15/2020 3:54 PM, Jason Martin wrote:

Does this help?

In Base64 encoding, the length of output-encoded String must be a multiple of three. If not, the output will be padded with additional pad characters (=).

Upon decoding, these extra padding characters will be discarded. To dig deeper into padding in Base64, check out this detailed answer on Stack Overflow.

From:

https://www.baeldung.com/java-base64-encode-and-decode
_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

rvjansen

Re: Please help with Java problem in NetRexx

Try:

options nostrictcase nostrictargs nostrictsignal

package org.netrexx.njpipes.stages

import org.netrexx.njpipes.pipes.

import java.util.Base64

class decode64 extends stage

method run()

opr=''

dec = Base64.getMIMEDecoder()

loop label Main forever

record_in = peekto().toString()

ob = dec.decode(record_in)

loop i=0 to ob.length-1

opr = opr||char(ob[i])

end

output(opr)

readto()

catch stageError

rc = rc()

end Main

Exit(rc*(rc<>12))

Can be done better without the concatenation, but you get the picture.

best regards,

René.

On 15 Oct 2020, at 22:41, Jeff Hennick <[hidden email]> wrote:
Thank you, Jason.
Using the MIME-type encoding, specifically in the Java (and implicitly in CMS), handles all that. (There can be additional bit level padding with 0s to get to a full byte.) And it is in the _working_ Encoding that all this is important.
The input string in the test for Decoding, line 39, is "SGVsbG8sIEJhc2U2NCE=", this is the result, with the padding =, from the Encoding of "Hello, Base64!".
For Encoding tests, I also use these short ones: @!*, @!, and @; with the results being QCEq, QCE=, and QA==. (I got all these test phrases from some web page, and the results of encoding agree with theirs.)
I am still left hanging with the incorrect and varying decoding results. I have tried running the short examples too, but still get that enigmatic [B@57e382 - type result from them, too.
On 10/15/2020 3:54 PM, Jason Martin wrote:
Does this help?

In Base64 encoding, the length of output-encoded String must be a multiple of three. If not, the output will be padded with additional pad characters (=).
Upon decoding, these extra padding characters will be discarded. To dig deeper into padding in Base64, check out this detailed answer on Stack Overflow.

From:
https://www.baeldung.com/java-base64-encode-and-decode
_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

rvjansen

Re: Please help with Java problem in NetRexx

In reply to this post by rvjansen

Even easier:

options nostrictcase nostrictargs nostrictsignal

package org.netrexx.njpipes.stages

import org.netrexx.njpipes.pipes.

import java.util.Base64

class decode64 extends stage

method run()

dec = Base64.getMIMEDecoder()

loop label Main forever

record_in = peekto().toString()

ob = dec.decode(record_in)

s = String(ob)

output(Rexx s)

readto()

catch stageError

rc = rc()

end Main

Exit(rc*(rc<>12))

String can take a byte[], while Rexx cannot. Might be able to add that in 3.11 (will skip 3.10 for confusion with 3.1)

René.

On 15 Oct 2020, at 22:38, [hidden email] wrote:

Hi Jeff,

These are references - like memory pointers. They need some dereferencing. Let me try something.

René.

On 15 Oct 2020, at 21:37, Jeff Hennick <[hidden email]> wrote:

I am currently working to bring CMS Pipelines Base64 stages 64Encode and 64Decode into NetRexx Pipelines. I am using the java.util.Base64 class to do the work. Encode works great, but decode gives gibberish, and different gibberish each time it is run.
A note on names. CMS calls the stages 64Encode and 64Decode. NetRexx objects to class names beginning with with a digit, so I am using Encode64 and Decode64, and adding compiler based aliases of the CMS names.
Here is what works for Encode
options nostrictcase nostrictargs nostrictsignal
--package org.netrexx.njpipes.stages
import org.netrexx.njpipes.pipes.
import java.util.Base64
class encode64 extends stage
method run()
enc = Base64.getMIMEEncoder()
loop label Main forever
    record_in = Rexx peekto()
    ba = record_in.toByteArray()
    opr = Rexx enc.encodeToString(ba)
    output(opr)
    readto()
catch stageError
    rc = rc()
end Main
Exit(rc*(rc<>12))
Here is what does not work for Decode
options nostrictcase nostrictargs nostrictsignal
--package org.netrexx.njpipes.stages
import org.netrexx.njpipes.pipes.
import java.util.Base64
class decode64 extends stage
method run()
dec = Base64.getMIMEDecoder()
loop label Main forever
    record_in = peekto().toString()
    ob = dec.decode(record_in)
    opr = Rexx ob
    output(opr)
    readto()
catch stageError
    rc = rc()
end Main
Exit(rc*(rc<>12))
Here is the trace, and a second run's result, too
PS C:\Users\Jeff\documents\pipe tests> java ct --- decode64.nrx [Thread-7,njPipes] 33 *=* method run() 35 *=* dec = Base64.getMIMEDecoder() >v> dec "java.util.Base64$Decoder@11c1b90" 37 *=* loop label Main forever 39 *=* record_in = peekto().toString() >v> record_in "SGVsbG8sIEJhc2U2NCE=" 41 *=* ob = dec.decode(record_in) >>> "SGVsbG8sIEJhc2U2NCE=" >v> ob "[B@18bd538" 43 *=* opr = Rexx ob >v> opr "[B@18bd538" 45 *=* output(opr) >>> "[B@18bd538" 49 *=* catch stageError 50 *=* rc = rc() >v> rc "12" 51 *=* end Main 53 *=* Exit(rc*(rc<>12)) >>> "0" # 2 64decode **FAIL** at rec 1 col 1. Less: Actual: [B@18bd538 Expected: Hello, Base64!

# 2 64decode **FAIL** at rec 1 col 1. Less: Actual: [B@57e382 Expected: Hello, Base64!

Notice that the two Actuals are different, even though run one right after the other.
I have tried various combinations of casting things to and from Rexx, string, byte[] with no help. I keep getting the [hidden email]-type results. (The Encode took some such shenanigans to coax it to work.)
Any help pointers you can give would be gratefully received. Thanks!
Jeff Hennick

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

Jeff Hennick-3

Re: Please help with Java problem in NetRexx

René,

Thank you. Of course that works just great.

In reading the Java documentation, I was concentrating on the class of the parameter (either byte[] or String), and glossing over the class (byte[]) that is returned.

As for the 3.11, I'll vote to have anything "string like" be transparently handled as Rexx in NetRexx.

As for 64Encode/64Decode, I'm going to try to add options to cover the other two flavors of Java Base64, beyond MIME -- which will be the default, before submitting them.

Jeff

On 10/15/2020 5:33 PM, René Jansen wrote:

Even easier: options nostrictcase nostrictargs nostrictsignal package...                                                                                                                                                                                     Even easier:

options nostrictcase nostrictargs nostrictsignal

package org.netrexx.njpipes.stages

import org.netrexx.njpipes.pipes.

import java.util.Base64

class decode64 extends stage

method run()

dec = Base64.getMIMEDecoder()

loop label Main forever

record_in = peekto().toString()

ob = dec.decode(record_in)

s = String(ob)

output(Rexx s)

readto()

catch stageError

rc = rc()

end Main

Exit(rc*(rc<>12))

String can take a byte[], while Rexx cannot. Might be able to add that in 3.11 (will skip 3.10 for confusion with 3.1)

René.

On 15 Oct 2020, at 22:38, [hidden email] wrote:

Hi Jeff,

These are references - like memory pointers. They need some dereferencing. Let me try something.

René.

On 15 Oct 2020, at 21:37, Jeff Hennick <[hidden email]> wrote:

I am currently working to bring CMS Pipelines Base64 stages 64Encode and 64Decode into NetRexx Pipelines. I am using the java.util.Base64 class to do the work. Encode works great, but decode gives gibberish, and different gibberish each time it is run.

A note on names. CMS calls the stages 64Encode and 64Decode. NetRexx objects to class names beginning with with a digit, so I am using Encode64 and Decode64, and adding compiler based aliases of the CMS names.

Here is what works for Encode

options nostrictcase nostrictargs nostrictsignal
--package org.netrexx.njpipes.stages
import org.netrexx.njpipes.pipes.
import java.util.Base64
class encode64 extends stage
method run()
enc = Base64.getMIMEEncoder()
loop label Main forever
    record_in = Rexx peekto()
    ba = record_in.toByteArray()
    opr = Rexx enc.encodeToString(ba)
    output(opr)
    readto()
catch stageError
    rc = rc()
end Main
Exit(rc*(rc<>12))

Here is what does not work for Decode

options nostrictcase nostrictargs nostrictsignal
--package org.netrexx.njpipes.stages
import org.netrexx.njpipes.pipes.
import java.util.Base64
class decode64 extends stage
method run()
dec = Base64.getMIMEDecoder()
loop label Main forever
    record_in = peekto().toString()
    ob = dec.decode(record_in)
    opr = Rexx ob
    output(opr)
    readto()
catch stageError
    rc = rc()
end Main
Exit(rc*(rc<>12))

Here is the trace, and a second run's result, too

PS C:\Users\Jeff\documents\pipe tests> java ct --- decode64.nrx [Thread-7,njPipes] 33 *=* method run() 35 *=* dec = Base64.getMIMEDecoder() >v> dec "java.util.Base64$Decoder@11c1b90" 37 *=* loop label Main forever 39 *=* record_in = peekto().toString() >v> record_in "SGVsbG8sIEJhc2U2NCE=" 41 *=* ob = dec.decode(record_in) >>> "SGVsbG8sIEJhc2U2NCE=" >v> ob "[B@18bd538" 43 *=* opr = Rexx ob >v> opr "[B@18bd538" 45 *=* output(opr) >>> "[B@18bd538" 49 *=* catch stageError 50 *=* rc = rc() >v> rc "12" 51 *=* end Main 53 *=* Exit(rc*(rc<>12)) >>> "0" # 2 64decode **FAIL** at rec 1 col 1. Less: Actual: [B@18bd538 Expected: Hello, Base64!

# 2 64decode **FAIL** at rec 1 col 1. Less: Actual: [B@57e382 Expected: Hello, Base64!

Notice that the two Actuals are different, even though run one right after the other.

I have tried various combinations of casting things to and from Rexx, string, byte[] with no help. I keep getting the [hidden email]-type results. (The Encode took some such shenanigans to coax it to work.)

Any help pointers you can give would be gratefully received. Thanks!

Jeff Hennick

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/
_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/

_______________________________________________
Ibm-netrexx mailing list
[hidden email]
Online Archive : http://ibm-netrexx.215625.n3.nabble.com/