Here is what I think I know about such things, and also may be a clue as to why
I do not understand what is going on between my emulator and ISPF. It's also
horribly mind twisting.
First what I think happens before any fancy encoding or translations.
If I have my emulator set to CP-37 (or CP-1047) and I type a $-sign on my
keyboard then the hex value of the character that the mainframe sees is of
course x'5B'. So if I enter some code that has a $-sign in it then the
compiler sees the $-sign as x'5B' and all is well.
If I change my emulator's code page to Finnish, for example, which is CP-278
and I type a $-sign on my keyboard then the hex value of the character that the
mainframe sees is x'67'. If I use the $-sign in some code then the compiler
sees x'67' and generates a syntax error.
In CP-37 / CP-1047 x'67' shows up as the Swedish Oh (aka A-ring) which looks
like this: Å
An Å _in CP-37_ in a member name or code as x'67' isn't valid. And the same
with CP-278, a $-sign in a member name or code as the same x'67' isn't valid.
It goes the other way, too. If I have my emulator's code page set to Finnish
CP-278 I may look at a list of members in SYS1.MACLIB and this member: IXZÅXPL
I have the "national characters" (@$#) memorized and know that in reality the Å
is x'5B' which means to MVS it is really a $-sign. Conversely, I know if I am
reading some instructions for CP-37 and it says to enter a $-sign that I enter
Å because what it really wants is x'5B' and doesn't care what "I" see.
All the long winded explanation above is how things "normally" behave. The
characters that cause problems are called variant characters because these
characters vary across different code pages. The list of variant characters
are: ! # @ { } \ [ ] | $ ~ ¦
Here is the code page for CP-278 for Finland/Sweden and if you compared this to
all the others you'd see that they all match except for the above characters.
http://www-03.ibm.com/systems/i/software/globalization/pdf/cp00278z.pdf
Dealing with just the national characters is usually ok, but when you get to
code that has braces an brackets and pipe characters it not only gets
confusing, but in many cases the character isn't even on the keyboard.
Now with encoding...
Some products, SAS for example, have a sort of character-WYSIWYG translation.
They know that if you say you are using German or Finnish or Italian to switch
what it expects for a hex value so that it matches what the user sees. So if
in Finnish I need a $-sign, I just type in $ and everything is fine - even
though it is x'67' and not x'5B'.
Again my apologies for being windy, but this is hard to grasp, at least for me.
All this to say that I expect that when my emulator and ISPF are in sync that
with my emulator set to CP-278 Finnish that I could type $ (aka x'67') and it
would treat it as a $-sign even though the hex value is different from CP-37.
This appears not to be the case, either because I'm doing something wrong or I
have understood what should happen incorrectly. So where did I go wrong?
Thanks!
Lindy
-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[EMAIL PROTECTED] On Behalf Of
Hunkeler Peter (KIUK 3)
Sent: 11. heinäkuuta 2008 15:24
To: [email protected]
Subject: Re: How to change the ISPF code page?
>Is it possible that Lindy needs to add translation tables
to ISPF?
I acknowledge my ignorance wiht this ISPF topic. I still
don't have a clue what this translation behind the various
terminal models is intended for. Is it a relict from the
time when real 3270 terminal were standard? Isn't this
obsolete today when everybody is using terminal emulator
software?
Is it correct that this translation is only applied when
displaying data on the terminal but not when writing
data from the terminal? (This is at least what I seem to
see). What does it translate? I tried several combinations
ISPF terminal type vs. emulator-CP but the more I try the
more it confuses me.
I admit that, for once, I'm too lazy to RTFM. So if anyone
is willing to briefly explain a case, I'd appreciate. Otherwise
I'll add it to my "for a lazy rainy sunday" to-do list.
--
Peter Hunkeler
CREDIT SUISSE
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
