I'm leaning towards SRST/MVCL.
--
Shmuel (Seymour J.) Metz
http://mason.gmu.edu/~smetz3
עַם יִשְׂרָאֵל חַי
נֵ֣צַח יִשְׂרָאֵ֔ל לֹ֥א יְשַׁקֵּ֖ר
________________________________________
From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> on behalf of
Kirk Wolf <k...@coztoolkit.com>
Sent: Wednesday, October 30, 2024 4:41 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: Bounded string move?
Caution: This email did not originate from George Mason’s mail system. Do not
click links or open attachments unless you recognize the sender and know the
content is safe.
It appears to me that your requirement matches the C function below. (?)
I would probably just use C/C++, but you can see what the IBM C/C++ optimizing
compiler generated. It seems unlikely to me that there is a significantly
better way. You might ask on mvs-asm370.
size_t WordTest::wordCopy(char* out, size_t outLen, char* in, size_t inLen) {
char* pBlank = (char*)memchr(in, ' ', inLen);
size_t wordLen = pBlank
? pBlank - in
: inLen;
if (wordLen > outLen)
wordLen = outLen;
memcpy(out, in, wordLen);
return wordLen;
}
Note: in the pseudo assembly listing, the source doesn't always appear in the
right place in the generated code
OFFSET OBJECT CODE LINE# FILE# P S E U D O A S S E M B L Y L I S
T I N G
000188 | * size_t WordTest::wordCopy(char*
out, size_t outLen, char* in, size_t inLen) {
002B48 @21L0 DS 0D
002B48 00C300C5 =F'12779717'
XPLink entrypoint marker
002B4C 00C500F1 =F'12910833'
002B50 00001398 =F'5016'
002B54 00000088 =F'136'
WordTest::wordCopy(char*,unsigned
int,char*,unsigned int)
002B58 000188 | DS 0D
002B58 906A 4788 000188 | STM r6,r10,1928(r4)
002B5C End of Prolog
002B5C 58A0 484C 000164 | 12 L r10,in(,r4,2124)
002B60 5810 4850 000164 | 12 L r1,inLen(,r4,2128)
002B64 1892 000145 | 12 LR r9,r2
002B66 4100 0040 000164 | 12 LA r0,64
002B6A 182A 000164 | 12 LR r2,r10
002B6C 1851 000164 | 12 LR r5,r1
002B6E 5E50 484C 000164 | 12 AL r5,in(,r4,2124)
002B72 B25E 0052 000164 | 12 SRST r5,r2
002B76 A714 FFFE 000164 | 12 JO *-4
002B7A A744 0004 000164 | 12 JL @21L972
002B7E 4150 0000 000164 | 12 LA r5,0
002B82 000164 | 12 @21L972 DS 0H
000189 | * char* pBlank = (char*)memchr(in, '
', inLen);
000190 | * size_t wordLen = pBlank
002B82 EC58 0005 007E 000190 | CIJE r5,H'0',@21L1258
000191 | *
? pBlank - in
002B88 1F5A 000191 | SLR r5,r10
000192 | *
: inLen;
002B8A 1815 000192 | LR r1,r5
002B8C 000190 | @21L1258 DS 0H
000193 | * if (wordLen > outLen)
002B8C EC13 0004 2077 000193 | CLRJH r1,r3,@21L1259
002B92 1831 000193 | LR r3,r1
002B94 000193 | @21L1259 DS 0H
002B94 1813 000145 | 12 LR r1,r3
002B96 EC38 0016 007E 000145 | 12 CIJE r3,H'0',@21L665
002B9C A71A FFFF 000145 | 12 AHI r1,H'-1'
002BA0 1801 000145 | 12 LR r0,r1
002BA2 8A00 0008 000145 | 12 SRA r0,8
002BA6 A784 000B 000145 | 12 JE @21L974
002BAA 000145 | 12 @21L973 DS 0H
002BAA D2FF 9000 A000 000145 | 12 MVC
(char)(256,r9,0),(char)(r10,0)
002BB0 4190 9100 000145 | 12 LA r9,(char)(,r9,256)
002BB4 41A0 A100 000145 | 12 LA r10,(char)(,r10,256)
002BB8 A706 FFF9 000145 | 12 BRCT r0,@21L973
002BBC 000145 | 12 @21L974 DS 0H
002BBC C610 0000 05FA 000145 | 12 EXRL r1,F'1530'
000194 | * wordLen = outLen;
000195 | * memcpy(out, in, wordLen);
000196 | * return wordLen;
000197 | * }
002BC2 000197 | @21L665 DS 0H
002BC2 989A 4794 000197 | LM r9,r10,1940(r4)
002BC6 47F0 7004 000197 | B 4(,r7)
Kirk Wolf
Dovetailed Technologies
https://coztoolkit.com/
On Wed, Oct 30, 2024, at 1:03 PM, Seymour J Metz wrote:
> No, if the translated character matches it is not stored. In my case I would
> be using a table of table dc 256al1(*-table) with a match character of space
> and explicitly add a space to the end.
>
> --
> Shmuel (Seymour J.) Metz
> http://mason.gmu.edu/~smetz3
> עַם יִשְׂרָאֵל חַי
> נֵ֣צַח יִשְׂרָאֵ֔ל לֹ֥א יְשַׁקֵּ֖ר
>
>
>
> ________________________________________
> From: IBM Mainframe Discussion List <IBM-MAIN@LISTSERV.UA.EDU> on behalf of
> Robin Vowels <0000054c2a8bb600-dmarc-requ...@listserv.ua.edu>
> Sent: Wednesday, October 30, 2024 2:53 AM
> To: IBM-MAIN@LISTSERV.UA.EDU
> Subject: Re: Bounded string move?
>
> Caution: This email did not originate from George Mason’s mail system. Do not
> click links or open attachments unless you recognize the sender and know the
> content is safe.
>
>
> On 2024-10-29 14:48, Seymour J Metz wrote:
> > You don't need to know in advance. The question I asked was about doing
> > a copy of a delimited string (in my use case, blank terminated) with
> > explicit bounds on the source and target sizes. TROO comes closest to
> > doing what I want, but it has that pesky alignment requirement.
>
> Does it move the blank?
>
> ----------------------------------------------------------------------
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
>
>
> ----------------------------------------------------------------------
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
>
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message: INFO IBM-MAIN
