$ type z.com
$ write sys$output "''p1'"
$ @z a
A $ @z "a"
a

This is not really surprising.

But are there any way to preserve pn case without
double-quoting it?

Like this hypothetical syntax:

$ set proc/parse=(extended,pnkeepcase)

Arne

On 28/12/2022 16:45, Arne Vajhøj wrote:
> $ type z.com
> $ write sys$output "''p1'"
> $ @z a
> A
> $ @z "a"
> a
>
> This is not really surprising.
>
> But are there any way to preserve pn case without
> double-quoting it?
>
> Like this hypothetical syntax:
>
> $ set proc/parse=(extended,pnkeepcase)
>
> Arne

There is of course

$ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED

Does this not work?

--
Chris

On 12/28/2022 1:29 PM, Chris Townley wrote:
> On 28/12/2022 16:45, Arne Vajhøj wrote:
>> $ type z.com
>> $ write sys$output "''p1'"
>> $ @z a
>> A
>> $ @z "a"
>> a
>>
>> This is not really surprising.
>>
>> But are there any way to preserve pn case without
>> double-quoting it?
>>
>> Like this hypothetical syntax:
>>
>> $ set proc/parse=(extended,pnkeepcase)
>
> There is of course
>
> $ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
>
> Does this not work?

I did not know that. But no.

$ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
%DCL-W-IVKEYW, unrecognized keyword - check validity and spelling
\CASE_LOOKUP\
$ SET PROCESS /CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
$ @z a
%DCL-E-OPENIN, error opening DISK2:[ARNE.gpcp]z.COM; as input
-RMS-E-FNF, file not found
$ @z.com a
A $ @z.com "a"
a

Arne

On 28/12/2022 18:38, Arne Vajhøj wrote:
> On 12/28/2022 1:29 PM, Chris Townley wrote:
>> On 28/12/2022 16:45, Arne Vajhøj wrote:
>>> $ type z.com
>>> $ write sys$output "''p1'"
>>> $ @z a
>>> A
>>> $ @z "a"
>>> a
>>>
>>> This is not really surprising.
>>>
>>> But are there any way to preserve pn case without
>>> double-quoting it?
>>>
>>> Like this hypothetical syntax:
>>>
>>> $ set proc/parse=(extended,pnkeepcase)
>>
>> There is of course
>>
>> $ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
>>
>> Does this not work?
>
> I did not know that. But no.
>
> $ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
> %DCL-W-IVKEYW, unrecognized keyword - check validity and spelling
>  \CASE_LOOKUP\
> $ SET PROCESS /CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
> $ @z a
> %DCL-E-OPENIN, error opening DISK2:[ARNE.gpcp]z.COM; as input
> -RMS-E-FNF, file not found
> $ @z.com a
> A
> $ @z.com "a"
> a
>
> Arne
>

Interesting. What version of VMS are you on?

My example was straight from the VSI DCL dictionary. I was vaguely aware
of something like that.

Presumably the command affects the current process - do you need to have
it in z.com? Although you shouldn't as you are double quoting the symbol
substitution.

--
Chris

On 12/28/22 12:38 PM, Arne Vajhøj wrote:
> On 12/28/2022 1:29 PM, Chris Townley wrote:
>> On 28/12/2022 16:45, Arne Vajhøj wrote:
>>> $ type z.com
>>> $ write sys$output "''p1'"
>>> $ @z a
>>> A
>>> $ @z "a"
>>> a
>>>
>>> This is not really surprising.
>>>
>>> But are there any way to preserve pn case without
>>> double-quoting it?
>>>
>>> Like this hypothetical syntax:
>>>
>>> $ set proc/parse=(extended,pnkeepcase)
>>
>> There is of course
>>
>> $ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
>>
>> Does this not work?
>
> I did not know that. But no.
>
> $ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
> %DCL-W-IVKEYW, unrecognized keyword - check validity and spelling
>  \CASE_LOOKUP\
> $ SET PROCESS /CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
> $ @z a
> %DCL-E-OPENIN, error opening DISK2:[ARNE.gpcp]z.COM; as input
> -RMS-E-FNF, file not found
> $ @z.com a
> A
> $ @z.com "a"
> a

Case sensitivity doesn't really have anything to do with case
preservation. I don't know of any way to do what Arne wants here except
to avoid DCL. Write a C program (or use something written in C, such as
Perl or Python, instead of DCL) and set extended parse in the process
and DECC$ARGV_PARSE_STYLE in the environment before running the C program.

On 12/28/2022 1:45 PM, Chris Townley wrote:
> On 28/12/2022 18:38, Arne Vajhøj wrote:
>> On 12/28/2022 1:29 PM, Chris Townley wrote:
>>> On 28/12/2022 16:45, Arne Vajhøj wrote:
>>>> $ type z.com
>>>> $ write sys$output "''p1'"
>>>> $ @z a
>>>> A
>>>> $ @z "a"
>>>> a
>>>>
>>>> This is not really surprising.
>>>>
>>>> But are there any way to preserve pn case without
>>>> double-quoting it?
>>>>
>>>> Like this hypothetical syntax:
>>>>
>>>> $ set proc/parse=(extended,pnkeepcase)
>>>
>>> There is of course
>>>
>>> $ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
>>>
>>> Does this not work?
>>
>> I did not know that. But no.
>>
>> $ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
>> %DCL-W-IVKEYW, unrecognized keyword - check validity and spelling
>>   \CASE_LOOKUP\
>> $ SET PROCESS /CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
>> $ @z a
>> %DCL-E-OPENIN, error opening DISK2:[ARNE.gpcp]z.COM; as input
>> -RMS-E-FNF, file not found
>> $ @z.com a
>> A
>> $ @z.com "a"
>> a
>
> Interesting. What version of VMS are you on?

$ write sys$output f$getsyi("version")
V8.4-2L2

> My example was straight from the VSI DCL dictionary. I was vaguely aware
> of something like that.
>
> Presumably the command affects the current process - do you need to have
> it in z.com? Although you shouldn't as you are double quoting the symbol
> substitution.

I am pretty sure that the problem has already happened when the first
line in z.com is executed.

Arne

On 12/28/2022 3:07 PM, Craig A. Berry wrote:
>>> On 28/12/2022 16:45, Arne Vajhøj wrote:
>>>> $ type z.com
>>>> $ write sys$output "''p1'"
>>>> $ @z a
>>>> A
>>>> $ @z "a"
>>>> a
>>>>
>>>> This is not really surprising.
>>>>
>>>> But are there any way to preserve pn case without
>>>> double-quoting it?
>>>>
>>>> Like this hypothetical syntax:
>>>>
>>>> $ set proc/parse=(extended,pnkeepcase)

> Case sensitivity doesn't really have anything to do with case
> preservation.  I don't know of any way to do what Arne wants here except
> to avoid DCL.  Write a C program (or use something written in C, such as
> Perl or Python, instead of DCL) and set extended parse in the process
> and DECC$ARGV_PARSE_STYLE in the environment before running the C program.

:-(

But thanks.

Arne

On 28/12/2022 20:35, Arne Vajhøj wrote:
> On 12/28/2022 1:45 PM, Chris Townley wrote:
>> On 28/12/2022 18:38, Arne Vajhøj wrote:
>>> On 12/28/2022 1:29 PM, Chris Townley wrote:
>>>> On 28/12/2022 16:45, Arne Vajhøj wrote:
>>>>> $ type z.com
>>>>> $ write sys$output "''p1'"
>>>>> $ @z a
>>>>> A
>>>>> $ @z "a"
>>>>> a
>>>>>
>>>>> This is not really surprising.
>>>>>
>>>>> But are there any way to preserve pn case without
>>>>> double-quoting it?
>>>>>
>>>>> Like this hypothetical syntax:
>>>>>
>>>>> $ set proc/parse=(extended,pnkeepcase)
>>>>
>>>> There is of course
>>>>
>>>> $ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
>>>>
>>>> Does this not work?
>>>
>>> I did not know that. But no.
>>>
>>> $ SET PROCESS /CASE=CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
>>> %DCL-W-IVKEYW, unrecognized keyword - check validity and spelling
>>>   \CASE_LOOKUP\
>>> $ SET PROCESS /CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
>>> $ @z a
>>> %DCL-E-OPENIN, error opening DISK2:[ARNE.gpcp]z.COM; as input
>>> -RMS-E-FNF, file not found
>>> $ @z.com a
>>> A
>>> $ @z.com "a"
>>> a
>>
>> Interesting. What version of VMS are you on?
>
> $ write sys$output f$getsyi("version")
> V8.4-2L2
>
>> My example was straight from the VSI DCL dictionary. I was vaguely
>> aware of something like that.
>>
>> Presumably the command affects the current process - do you need to
>> have it in z.com? Although you shouldn't as you are double quoting the
>> symbol substitution.
>
> I am pretty sure that the problem has already happened when the first
> line in z.com is executed.
>
> Arne
>

Interesting

$ SET PROCESS /CASE_LOOKUP=SENSITIVE /PARSE_STYLE=EXTENDED
$ aa = "a"
$ show sym aa
AA = "a"

$ @z.com 'aa'
A $ @z.com "''aa'"
a

Not sure what else to suggest...

--
Chris

On 12/28/2022 3:39 PM, Arne Vajhøj wrote:
> On 12/28/2022 3:07 PM, Craig A. Berry wrote:
>>>> On 28/12/2022 16:45, Arne Vajhøj wrote:
>>>>> $ type z.com
>>>>> $ write sys$output "''p1'"
>>>>> $ @z a
>>>>> A
>>>>> $ @z "a"
>>>>> a
>>>>>
>>>>> This is not really surprising.
>>>>>
>>>>> But are there any way to preserve pn case without
>>>>> double-quoting it?
>>>>>
>>>>> Like this hypothetical syntax:
>>>>>
>>>>> $ set proc/parse=(extended,pnkeepcase)
>
>> Case sensitivity doesn't really have anything to do with case
>> preservation.  I don't know of any way to do what Arne wants here except
>> to avoid DCL.  Write a C program (or use something written in C, such as
>> Perl or Python, instead of DCL) and set extended parse in the process
>> and DECC$ARGV_PARSE_STYLE in the environment before running the C
>> program.
>
> :-(
>
> But thanks.

I will use the double quotes.

But just for fun:

$ type zz.com
$ write sys$output "''p1' ''p2' ''p3'"
$ run realp
$ write sys$output "''p1' ''p2' ''p3'"
$ @zz aaaa BBBB CcCc
AAAA BBBB CCCC
aaaa BBBB CcCc

But there is a gazillion things that can go wrong with that
realp.

Code below.

Arne

.title realp
$SSDEF
$DSCDEF
.macro if_end_command,lbl
cmpl r3,r4
bgeq lbl
.endm
.macro if_space,lbl
cmpb (r3),#32
beql lbl
.endm
.psect $LOCAL quad,pic,con,lcl,noshr,noexe,wrt
cmdlen: .blkw 1
cmdbuf: .blkb 256
pnam: .ascid "P1"
pval: .blkw 1
.byte DSC$K_DTYPE_T
.byte DSC$K_CLASS_S
.address pvala
pvala: .blkb 256
.psect $CODE quad,pic,con,lcl,shr,exe,nowrt
.entry realp,^m<r2,r3,r4,r5>
; grab copy of command line
movl #CTL$AG_CLIDATA,r0
movl W^PPD$L_PRC(r0),r0
movl W^PRC_L_RECALLPTR(r0),r0 ; r0 = next command
subl2 #2,r0
cvtwl (r0),r1 ; r1 = length current command
subl2 r1,r0 ; r0 = current command
movw r1,cmdlen ; copy of length of command line
movc3 r1,(r0),cmdbuf ; copy of command line
movab cmdbuf,r3 ; r3 = current character on
command line
cvtwl cmdlen,r1
addl3 r1,r3,r4 ; r4 = end of command line
; skip command itself
100$: if_end_command done
if_space start
incl r3 ; move to next character
brb 100$
; process parameters
start: incl r3 ; r3 = start first argument
movl #1,r2 ; r2 = n = 1
200$: if_end_command done
movw #0,pval ; reset parameter data
movab pvala,r5 ; r5 = where to copy parameter data
300$: if_end_command 500$
if_space 400$
incw pval ; add to pval/pvala
movb (r3),(r5)
incl r3
incl r5
brb 300$
400$: incl r3 ; skip space between parameters
500$: pushl #0
pushab pval
pushab pnam
calls #3,G^LIB$SET_SYMBOL
incb pnam+9 ; Pn -> Pn+1
incl r2 ; n = n + 1
cmpl r2,#8 ; while n <= 8
bleq 200$
done: movl #SS$_NORMAL,r0
ret
.end realp

In article <toisqe$p0r$1@gioia.aioe.org>, =?UTF-8?Q?Arne_Vajh=c3=b8j?= <arne@vajhoej.dk> writes:
>On 12/28/2022 3:39 PM, Arne Vajhøj wrote:
>> On 12/28/2022 3:07 PM, Craig A. Berry wrote:
>>>>> On 28/12/2022 16:45, Arne Vajhøj wrote:
>>>>>> $ type z.com
>>>>>> $ write sys$output "''p1'"
>>>>>> $ @z a
>>>>>> A
>>>>>> $ @z "a"
>>>>>> a
>>>>>>
>>>>>> This is not really surprising.
>>>>>>
>>>>>> But are there any way to preserve pn case without
>>>>>> double-quoting it?
>>>>>>
>>>>>> Like this hypothetical syntax:
>>>>>>
>>>>>> $ set proc/parse=(extended,pnkeepcase)
>>
>>> Case sensitivity doesn't really have anything to do with case
>>> preservation.  I don't know of any way to do what Arne wants here except
>>> to avoid DCL.  Write a C program (or use something written in C, such as
>>> Perl or Python, instead of DCL) and set extended parse in the process
>>> and DECC$ARGV_PARSE_STYLE in the environment before running the C
>>> program.
>>
>> :-(
>>
>> But thanks.
>
>I will use the double quotes.
>
>But just for fun:
>
>$ type zz.com
>$ write sys$output "''p1' ''p2' ''p3'"
>$ run realp
>$ write sys$output "''p1' ''p2' ''p3'"
>$ @zz aaaa BBBB CcCc
>AAAA BBBB CCCC
>aaaa BBBB CcCc
>
>But there is a gazillion things that can go wrong with that
>realp.
>
>Code below.
>
>Arne
>
>
>
> .title realp
> $SSDEF
> $DSCDEF
> .macro if_end_command,lbl
> cmpl r3,r4
> bgeq lbl
> .endm
> .macro if_space,lbl
> cmpb (r3),#32
> beql lbl
> .endm
> .psect $LOCAL quad,pic,con,lcl,noshr,noexe,wrt
>cmdlen: .blkw 1
>cmdbuf: .blkb 256
>pnam: .ascid "P1"
>pval: .blkw 1
> .byte DSC$K_DTYPE_T
> .byte DSC$K_CLASS_S
> .address pvala
>pvala: .blkb 256
> .psect $CODE quad,pic,con,lcl,shr,exe,nowrt
> .entry realp,^m<r2,r3,r4,r5>
>; grab copy of command line
> movl #CTL$AG_CLIDATA,r0
> movl W^PPD$L_PRC(r0),r0
> movl W^PRC_L_RECALLPTR(r0),r0 ; r0 = next command
> subl2 #2,r0
> cvtwl (r0),r1 ; r1 = length current command
> subl2 r1,r0 ; r0 = current command
> movw r1,cmdlen ; copy of length of command line
> movc3 r1,(r0),cmdbuf ; copy of command line
> movab cmdbuf,r3 ; r3 = current character on
>command line
> cvtwl cmdlen,r1
> addl3 r1,r3,r4 ; r4 = end of command line
>; skip command itself
>100$: if_end_command done
> if_space start
> incl r3 ; move to next character
> brb 100$
>; process parameters
>start: incl r3 ; r3 = start first argument
> movl #1,r2 ; r2 = n = 1
>200$: if_end_command done
> movw #0,pval ; reset parameter data
> movab pvala,r5 ; r5 = where to copy parameter data
>300$: if_end_command 500$
> if_space 400$
> incw pval ; add to pval/pvala
> movb (r3),(r5)
> incl r3
> incl r5
> brb 300$
>400$: incl r3 ; skip space between parameters
>500$: pushl #0
> pushab pval
> pushab pnam
> calls #3,G^LIB$SET_SYMBOL
> incb pnam+9 ; Pn -> Pn+1
> incl r2 ; n = n + 1
> cmpl r2,#8 ; while n <= 8
> bleq 200$
>done: movl #SS$_NORMAL,r0
> ret
> .end realp

Good grief! I needed an airsick bag and bottle of Tums after reading that.

--
VAXman- A Bored Certified VMS Kernel Mode Hacker VAXman(at)TMESIS(dot)ORG

I speak to machines with the voice of humanity.

On 12/28/2022 10:49 PM, VAXman-@SendSpamHere.ORG wrote:
> In article <toisqe$p0r$1@gioia.aioe.org>, =?UTF-8?Q?Arne_Vajh=c3=b8j?= <arne@vajhoej.dk> writes:
>> .title realp
>> $SSDEF
>> $DSCDEF
>> .macro if_end_command,lbl
>> cmpl r3,r4
>> bgeq lbl
>> .endm
>> .macro if_space,lbl
>> cmpb (r3),#32
>> beql lbl
>> .endm
>> .psect $LOCAL quad,pic,con,lcl,noshr,noexe,wrt
>> cmdlen: .blkw 1
>> cmdbuf: .blkb 256
>> pnam: .ascid "P1"
>> pval: .blkw 1
>> .byte DSC$K_DTYPE_T
>> .byte DSC$K_CLASS_S
>> .address pvala
>> pvala: .blkb 256
>> .psect $CODE quad,pic,con,lcl,shr,exe,nowrt
>> .entry realp,^m<r2,r3,r4,r5>
>> ; grab copy of command line
>> movl #CTL$AG_CLIDATA,r0
>> movl W^PPD$L_PRC(r0),r0
>> movl W^PRC_L_RECALLPTR(r0),r0 ; r0 = next command
>> subl2 #2,r0
>> cvtwl (r0),r1 ; r1 = length current command
>> subl2 r1,r0 ; r0 = current command
>> movw r1,cmdlen ; copy of length of command line
>> movc3 r1,(r0),cmdbuf ; copy of command line
>> movab cmdbuf,r3 ; r3 = current character on
>> command line
>> cvtwl cmdlen,r1
>> addl3 r1,r3,r4 ; r4 = end of command line
>> ; skip command itself
>> 100$: if_end_command done
>> if_space start
>> incl r3 ; move to next character
>> brb 100$
>> ; process parameters
>> start: incl r3 ; r3 = start first argument
>> movl #1,r2 ; r2 = n = 1
>> 200$: if_end_command done
>> movw #0,pval ; reset parameter data
>> movab pvala,r5 ; r5 = where to copy parameter data
>> 300$: if_end_command 500$
>> if_space 400$
>> incw pval ; add to pval/pvala
>> movb (r3),(r5)
>> incl r3
>> incl r5
>> brb 300$
>> 400$: incl r3 ; skip space between parameters
>> 500$: pushl #0
>> pushab pval
>> pushab pnam
>> calls #3,G^LIB$SET_SYMBOL
>> incb pnam+9 ; Pn -> Pn+1
>> incl r2 ; n = n + 1
>> cmpl r2,#8 ; while n <= 8
>> bleq 200$
>> done: movl #SS$_NORMAL,r0
>> ret
>> .end realp
>
> Good grief! I needed an airsick bag and bottle of Tums after reading that.

That bad? Hmmm.

Well - I have not written much Macro-32 the last 25 years so ...

Arne

On Wednesday, December 28, 2022 at 11:45:45 AM UTC-5, Arne Vajhøj wrote:
> $ type z.com
> $ write sys$output "''p1'"
> $ @z a
> A
> $ @z "a"
> a
>
> This is not really surprising.
>
> But are there any way to preserve pn case without
> double-quoting it?
>
> Like this hypothetical syntax:
>
> $ set proc/parse=(extended,pnkeepcase)
>
> Arne

I don't think it's possible without quotation marks. Upcasing is done in the first phase of command parsing/evaluation/whatever.

See section 12.13 of the v8.4-2L1 Alpha user's manual. It doesn't mention the upcasing, by experiments I did long ago do seem to indicate that the upcasing does indeed occur in phase 1.

See also section 2.3. I could swear there was some example like abc"def"ghi, but maybe you need an older manual to find that.

Found some stuff.
First use SHOW SYMBOL instead of write sys$output "''p1"

$ type z.com
$ write sys$output "''p1'"
$ show symbol p1
$
$ @Z abc"def"ghi
ABCDEFGHI
P1 = "ABC"def"GHI"
$

See Help @ for that example.

On 12/28/22 21:58, Arne Vajhøj wrote:
>
>
> That bad? Hmmm.
>
> Well - I have not written much Macro-32 the last 25 years so ...
>

Gee... My MACRO skills are still pretty sharp.

Oh wait... You meant MACRO-32. Never mind. :-)

bill

On 12/28/2022 9:58 PM, Arne Vajhøj wrote:
> On 12/28/2022 10:49 PM, VAXman-@SendSpamHere.ORG wrote:
>> In article <toisqe$p0r$1@gioia.aioe.org>, =?UTF-8?Q?Arne_Vajh=c3=b8j?=
>> <arne@vajhoej.dk> writes:
[Macro-32 omitted]
>> Good grief!  I needed an airsick bag and bottle of Tums after reading
>> that.
>
> That bad? Hmmm.

You will probably not like the zweiter sehr verbesserte ausgabe either.

.title realp
$SSDEF
$DSCDEF
SPACE=32
PRC_C_CMDBUFSIZ=4097 ; 1025 on VAX
PRC$W_PROCLEVEL=56
.macro if_end_command,lbl
cmpl r3,r4
bgeq lbl
.endm
.macro if_space,lbl
cmpb (r3),#SPACE
beql lbl
.endm
.psect $PDATA quad,pic,con,lcl,shr,noexe,nowrt
nomsg: .ascid "This is not run from first level COM file so no action
taken"
.psect $LOCAL quad,pic,con,lcl,noshr,noexe,wrt
copy1: .blkb PRC_C_CMDBUFSIZ
copy2: .blkb PRC_C_CMDBUFSIZ
pnam: .ascid "P1"
pval: .blkw 1
.byte DSC$K_DTYPE_T
.byte DSC$K_CLASS_S
.address pvala
pvala: .blkb 256
.psect $CODE quad,pic,con,lcl,shr,exe,nowrt
.entry realp,^m<r2,r3,r4,r5,r6>
; phase 1 : grab copy of recall buffer
; out r0 = command length
; out r1 = command address
; temp r2 & r3 for working storage
; temp r6 = pointer recall buffer P1 space
movl #CTL$AG_CLIDATA,r6
movl W^PPD$L_PRC(r6),r6
cmpw W^PRC$W_PROCLEVEL(r6),#1 ; test if run from first level
COM file
bneq nogood
movc3 #PRC_C_CMDBUFSIZ,W^PRC_G_COMMANDS(r6),copy1 ; make two
copies of data to handle circular buffer
movc3 #PRC_C_CMDBUFSIZ,W^PRC_G_COMMANDS(r6),copy2
movab W^PRC_G_COMMANDS(r6),r2 ; find address next command
movl W^PRC_L_RECALLPTR(r6),r3
subl3 r2,r3,r1
movab copy2,r6
addl2 r6,r1
subl2 #2,r1 ; find length current command
cvtwl (r1),r0
subl2 r0,r1 ; find address current command
; phase 2 : parse
; in r0 = command length
; in r1 = command address
; temp r2 = parameter number
; temp r3 = current character in command
; temp r4 = end of command
; temp r5 = current Pn output character
movl r1,r3 ; current character = first
character
addl3 r0,r3,r4 ; find end of command
; skip command itself
100$: if_end_command done
if_space start
incl r3 ; skip if not space
brb 100$
; process parameters
start: incl r3 ; skip space between command
itself and P1
movl #1,r2 ; n = 1
200$: if_end_command done
movw #0,pval ; reset parameter data
movab pvala,r5
300$: if_end_command 500$
if_space 400$
incw pval ; add to parameter data
movb (r3)+,(r5)+
brb 300$
400$: incl r3 ; skip space between Pn and Pn+1
500$: pushl #0
pushab pval ; set Pn symbol
pushab pnam
calls #3,G^LIB$SET_SYMBOL
incb pnam+9 ; update Pn -> Pn+1
incl r2 ; n = n + 1
cmpl r2,#8 ; while n <= 8
bleq 200$
brb done
; message due to to not being run from first level COM file
nogood: pushab nomsg
calls #1,G^LIB$PUT_OUTPUT
; all done
done: movl #SS$_NORMAL,r0
ret
.end realp

Arne

On 31/12/2022 21:03, Arne Vajhøj wrote:
> On 12/28/2022 9:58 PM, Arne Vajhøj wrote:
>> On 12/28/2022 10:49 PM, VAXman-@SendSpamHere.ORG wrote:
>>> In article <toisqe$p0r$1@gioia.aioe.org>,
>>> =?UTF-8?Q?Arne_Vajh=c3=b8j?= <arne@vajhoej.dk> writes:
> [Macro-32 omitted]
>>> Good grief!  I needed an airsick bag and bottle of Tums after reading
>>> that.
>>
>> That bad? Hmmm.
>
> You will probably not like the zweiter sehr verbesserte ausgabe either.
>
>         .title  realp
>         $SSDEF
>         $DSCDEF
> SPACE=32
> PRC_C_CMDBUFSIZ=4097 ; 1025 on VAX
> PRC$W_PROCLEVEL=56
>         .macro  if_end_command,lbl
>         cmpl    r3,r4
>         bgeq    lbl
>         .endm
>         .macro  if_space,lbl
>         cmpb    (r3),#SPACE
>         beql    lbl
>         .endm
>         .psect  $PDATA quad,pic,con,lcl,shr,noexe,nowrt
> nomsg:  .ascid  "This is not run from first level COM file so no action
> taken"
>         .psect  $LOCAL quad,pic,con,lcl,noshr,noexe,wrt
> copy1:  .blkb   PRC_C_CMDBUFSIZ
> copy2:  .blkb   PRC_C_CMDBUFSIZ
> pnam:   .ascid  "P1"
> pval:   .blkw   1
>         .byte   DSC$K_DTYPE_T
>         .byte   DSC$K_CLASS_S
>         .address pvala
> pvala:  .blkb   256
>         .psect  $CODE quad,pic,con,lcl,shr,exe,nowrt
>         .entry  realp,^m<r2,r3,r4,r5,r6>
> ; phase 1 : grab copy of recall buffer
> ;   out  r0 = command length
> ;   out  r1 = command address
> ;   temp r2 & r3 for working storage
> ;   temp r6 = pointer recall buffer P1 space
>         movl    #CTL$AG_CLIDATA,r6
>         movl    W^PPD$L_PRC(r6),r6
>         cmpw    W^PRC$W_PROCLEVEL(r6),#1 ; test if run from first level
> COM file
>         bneq    nogood
>         movc3   #PRC_C_CMDBUFSIZ,W^PRC_G_COMMANDS(r6),copy1 ; make two
> copies of data to handle circular buffer
>         movc3   #PRC_C_CMDBUFSIZ,W^PRC_G_COMMANDS(r6),copy2
>         movab   W^PRC_G_COMMANDS(r6),r2 ; find address next command
>         movl    W^PRC_L_RECALLPTR(r6),r3
>         subl3   r2,r3,r1
>         movab   copy2,r6
>         addl2   r6,r1
>         subl2   #2,r1                   ; find length current command
>         cvtwl   (r1),r0
>         subl2   r0,r1                   ; find address current command
> ; phase 2 : parse
> ;   in   r0 = command length
> ;   in   r1 = command address
> ;   temp r2 = parameter number
> ;   temp r3 = current character in command
> ;   temp r4 = end of command
> ;   temp r5 = current Pn output character
>         movl    r1,r3                   ; current character = first
> character
>         addl3   r0,r3,r4                ; find end of command
> ; skip command itself
> 100$:   if_end_command done
>         if_space start
>         incl    r3                      ; skip if not space
>         brb     100$
> ; process parameters
> start:  incl    r3                      ; skip space between command
> itself and P1
>         movl    #1,r2                   ; n = 1
> 200$:   if_end_command done
>         movw    #0,pval                 ; reset parameter data
>         movab   pvala,r5
> 300$:   if_end_command 500$
>         if_space 400$
>         incw    pval                    ; add to parameter data
>         movb    (r3)+,(r5)+
>         brb     300$
> 400$:   incl    r3                      ; skip space between Pn and Pn+1
> 500$:   pushl   #0
>         pushab  pval                    ; set Pn symbol
>         pushab  pnam
>         calls   #3,G^LIB$SET_SYMBOL
>         incb    pnam+9                  ; update Pn -> Pn+1
>         incl    r2                      ; n = n + 1
>         cmpl    r2,#8                   ; while n <= 8
>         bleq    200$
>         brb     done
> ; message due to to not being run from first level COM file
> nogood: pushab  nomsg
>         calls   #1,G^LIB$PUT_OUTPUT
> ; all done
> done:   movl    #SS$_NORMAL,r0
>         ret
>         .end    realp
>
> Arne
That reminds why I told the development manager, when being interviewed
to join the programming team, that I would not do macro...

I still got the job!

Later when there was a problem with any of the macro, I just rewrote in
a high level language

--
Chris

On 2022-12-31 21:03:24 +0000, Arne Vajhj said:

Didn't review the code, definitely some things I'd do different, but it
seems this code is expecting traditional (#8) and not extended (#16)
parsing.

> .title realp
> $SSDEF
> $DSCDEF
> SPACE=32
> PRC_C_CMDBUFSIZ=4097 ; 1025 on VAX
> PRC$W_PROCLEVEL=56
> .macro if_end_command,lbl
> cmpl r3,r4
> bgeq lbl
> .endm
> .macro if_space,lbl
> cmpb (r3),#SPACE
> beql lbl
> .endm
> .psect $PDATA quad,pic,con,lcl,shr,noexe,nowrt
> nomsg: .ascid "This is not run from first level COM file so no action taken"
> .psect $LOCAL quad,pic,con,lcl,noshr,noexe,wrt
> copy1: .blkb PRC_C_CMDBUFSIZ
> copy2: .blkb PRC_C_CMDBUFSIZ
> pnam: .ascid "P1"
> pval: .blkw 1
> .byte DSC$K_DTYPE_T
> .byte DSC$K_CLASS_S
> .address pvala
> pvala: .blkb 256
> .psect $CODE quad,pic,con,lcl,shr,exe,nowrt
> .entry realp,^m<r2,r3,r4,r5,r6>
> ; phase 1 : grab copy of recall buffer
> ; out r0 = command length
> ; out r1 = command address
> ; temp r2 & r3 for working storage
> ; temp r6 = pointer recall buffer P1 space
> movl #CTL$AG_CLIDATA,r6
> movl W^PPD$L_PRC(r6),r6
> cmpw W^PRC$W_PROCLEVEL(r6),#1 ; test if run from first
> level COM file
> bneq nogood
> movc3 #PRC_C_CMDBUFSIZ,W^PRC_G_COMMANDS(r6),copy1 ; make two
> copies of data to handle circular buffer
> movc3 #PRC_C_CMDBUFSIZ,W^PRC_G_COMMANDS(r6),copy2
> movab W^PRC_G_COMMANDS(r6),r2 ; find address next command
> movl W^PRC_L_RECALLPTR(r6),r3
> subl3 r2,r3,r1
> movab copy2,r6
> addl2 r6,r1
> subl2 #2,r1 ; find length current command
> cvtwl (r1),r0
> subl2 r0,r1 ; find address current command
> ; phase 2 : parse
> ; in r0 = command length
> ; in r1 = command address
> ; temp r2 = parameter number
> ; temp r3 = current character in command
> ; temp r4 = end of command
> ; temp r5 = current Pn output character
> movl r1,r3 ; current character = first character
> addl3 r0,r3,r4 ; find end of command
> ; skip command itself
> 100$: if_end_command done
> if_space start
> incl r3 ; skip if not space
> brb 100$
> ; process parameters
> start: incl r3 ; skip space between command
> itself and P1
> movl #1,r2 ; n = 1
> 200$: if_end_command done
> movw #0,pval ; reset parameter data
> movab pvala,r5
> 300$: if_end_command 500$
> if_space 400$
> incw pval ; add to parameter data
> movb (r3)+,(r5)+
> brb 300$
> 400$: incl r3 ; skip space between Pn and Pn+1
> 500$: pushl #0
> pushab pval ; set Pn symbol
> pushab pnam
> calls #3,G^LIB$SET_SYMBOL
> incb pnam+9 ; update Pn -> Pn+1
> incl r2 ; n = n + 1
> cmpl r2,#8 ; while n <= 8

#16

> bleq 200$
> brb done
> ; message due to to not being run from first level COM file
> nogood: pushab nomsg
> calls #1,G^LIB$PUT_OUTPUT
> ; all done
> done: movl #SS$_NORMAL,r0
> ret
> .end realp
>
> Arne

--
Pure Personal Opinion | HoffmanLabs LLC

(snip)

> But are there any way to preserve pn case without
> double-quoting it?
Can you do it with CLD?

I do remember that TeX uses CLD, and I don't remember quoting command line arguments.

On 12/31/2022 7:07 PM, Stephen Hoffman wrote:
> On 2022-12-31 21:03:24 +0000, Arne Vajhj said:
>> .title realp

> Didn't review the code, definitely some things I'd do different,

Probably also better. But ...

> but it
> seems this code is expecting traditional (#8) and not extended (#16)
> parsing.

>> cmpl r2,#8 ; while n <= 8

> #16

This is interesting. One can get COM files to accept 16 parameters?
How?

Arne

Den 2023-01-01 kl. 16:52, skrev Arne Vajhøj:
> On 12/31/2022 7:07 PM, Stephen Hoffman wrote:
>> On 2022-12-31 21:03:24 +0000, Arne Vajhj said:
>>>          .title  realp
>
>> Didn't review the code, definitely some things I'd do different,
>
> Probably also better. But ...
>
>>                                                                   but it
>> seems this code is expecting traditional (#8) and not extended (#16)
>> parsing.
>
> >>          cmpl    r2,#8                   ; while n <= 8
>
> > #16
>
> This is interesting. One can get COM files to accept 16 parameters?
> How?
>
> Arne
>
>
>

Is it a mixup between number of COM file parameters
and number of sub-directoy levels?

> > This is interesting. One can get COM files to accept 16 parameters?
> > How?

See DCL_CTLFLAGS bit 3

If clear (the default), command procedure supports the default eight optional parameters (that is, (P1,P2,...P8)). If set, command procedure supports up to sixteen optional parameters (that is, (P1,P2,...P16)). This is also applicable when using the CALL command to transfer control to a subroutine.

Reference: https://wiki.vmssoftware.com/DCL_CTLFLAGS

Volker.

On 2023-01-01 15:52:57 +0000, Arne Vajhj said:

> On 12/31/2022 7:07 PM, Stephen Hoffman wrote:
>> On 2022-12-31 21:03:24 +0000, Arne Vajhj said:
>>> .title realp
>
>> Didn't review the code, definitely some things I'd do different,
>
> Probably also better. But ...
>
>> but it seems this code is expecting traditional (#8) and not extended
>> (#16) parsing.
>
> >> cmpl r2,#8 ; while n <= 8
>
> > #16
>
> This is interesting. One can get COM files to accept 16 parameters?
> How?

I'd usually suggest reading the manuals.

Officially: "With OpenVMS Version 8.4, DCL commands and qualifiers have
been enhanced to support upto [sic] 16 parameters as command line input
with command procedures. For more details on DCL enhancements, see the
HPE OpenVMS Version 8.4 New Features and Documentation Overview guide."

Alas, the OpenVMS User's Guide lacks this particular DCL update. "These
symbols, named P1, P2, and so on through P8", " and... P7 = "" P8 = ""
but no P9, and "DCL places parameters passed to command procedures in
the local symbols P1 to P8.", etc. I also see DCL examples with the
INQUIRE command used, too. 🤮

Somebody (and I'm pretty sure I know who it was) did fix the @
documentation in the DCL Dictionary, not that most folks will know or
will remember that @ is documented there.
https://docs.vmssoftware.com/vsi-openvms-dcl-dictionary-a-m/

I'd expect there are some other gaps in the docs for the features added
in the V8 era.

The docs need some help. Some recipes are needed, too. VSI was working
on their version of DEC TIMA/STARS articles (many of which are
effectively recipes), and which will help those with support.

--
Pure Personal Opinion | HoffmanLabs LLC

On 2023-01-01 16:03:05 +0000, Jan-Erik Sderholm said:

> Is it a mixup between number of COM file parameters and number of
> sub-directoy levels?

I've definitely been known to be forgetful, and worse, but I'm pretty
sure DCL supports P1 to P16 as of OpenVMS V8.4 See the DCL dictionary
@ doc, and see my other recent reply here.

Also pretty sure the directory depths on ODS-5 now max out at 255, and
not 8 or 16. See the ODS-5/EFS doc: "There can be up to 255 levels of
directories. The name of each directory can be up to 236 8-bit or 118
16-bit characters long..."

From an earlier discussion of the DCL P1 to P16 parameter support:
https://groups.google.com/g/comp.os.vms/c/KjlzbRje4go/m/ab3NOE7ZDRoJ

Lots of old code around with old limits, and probably some old doc with
old limits, too.

--
Pure Personal Opinion | HoffmanLabs LLC

On 1/1/2023 12:02 PM, Stephen Hoffman wrote:
> On 2023-01-01 15:52:57 +0000, Arne Vajhj said:
>
>> On 12/31/2022 7:07 PM, Stephen Hoffman wrote:
>>> On 2022-12-31 21:03:24 +0000, Arne Vajhj said:
>>>> .title  realp
>>
>>> Didn't review the code, definitely some things I'd do different,
>>
>> Probably also better. But ...
>>
>>> but it seems this code is expecting traditional (#8) and not extended
>>> (#16) parsing.
>>
>>  >>          cmpl    r2,#8                   ; while n <= 8
>>
>>  > #16
>>
>> This is interesting. One can get COM files to accept 16 parameters?
>> How?
>
> I'd usually suggest reading the manuals.

Fair enough.

But sometimes it is not easy to find the right place.

> Officially: "With OpenVMS Version 8.4, DCL commands and qualifiers have
> been enhanced to support upto [sic] 16 parameters as command line input
> with command procedures. For more details on DCL enhancements, see the
> HPE OpenVMS Version 8.4 New Features and Documentation Overview guide."
>
> Alas, the OpenVMS User's Guide lacks this particular DCL update. "These
> symbols, named P1, P2, and so on through P8", "  and... P7 = "" P8 = ""
> but no P9, and "DCL places parameters passed to command procedures in
> the local symbols P1 to P8.", etc.  I also see DCL examples with the
> INQUIRE command used, too. 🤮
>
> Somebody (and I'm pretty sure I know who it was) did fix the @
> documentation in the DCL Dictionary, not that most folks will know or
> will remember that @ is documented there.
> https://docs.vmssoftware.com/vsi-openvms-dcl-dictionary-a-m/

Yes. And it points to DCL_CTLFLAGS bit 3.

And that leads to:

https://wiki.vmssoftware.com/DCL_CTLFLAGS

So:

DCL_CTLFLAGS=8

in modparams.dat and autogen.

Maybe I will try it. Even though I don't really have a need
for 16 parameters.

Arne

On 1/1/2023 12:17 PM, Stephen Hoffman wrote:

> Lots of old code around with old limits, and probably some old doc with old
> limits, too.

CMS, even the version that was updated for ODS-5, still had a hardcoded limit
of 8 depths of directories.

That issue was fixed in some VSI CMS release.

--

--- Rob

On 1/1/2023 11:29 AM, Volker Halle wrote:
>>> This is interesting. One can get COM files to accept 16 parameters?
>>> How?
>
> See DCL_CTLFLAGS bit 3
>
> If clear (the default), command procedure supports the default eight
> optional parameters (that is, (P1,P2,...P8)). If set, command
> procedure supports up to sixteen optional parameters (that is,
> (P1,P2,...P16)). This is also applicable when using the CALL command
> to transfer control to a subroutine. >
> Reference: https://wiki.vmssoftware.com/DCL_CTLFLAGS

So much to learn so little time.

I have a strong feeling that very few has set that bit.

Arne