On 28-Feb-2014 14:19 -0800, Leif Guldbrand wrote:
Some further information:
Look at: http://think400.dk/adhoc_4.htm#eks0025
As alluded by the "Editor's note" in the document, what is provided
in the presented table of characters is of limited use, possibly being
incorrect for anyone else using a different language environment than
whoever created that table, whenever attempted to be applied generally;
i.e. applied outside of the truly invariant [and mostly also, for the
several almost-invariant] EBCDIC characters. Indeed, the CCSID must be
considered. On a side note, the line starting with the digit x'4' is
seriously incorrect, no matter what CCSID one might presume the table
represents; e.g. the code point x'41', almost universally [as one of the
almost-invariant characters] represents the required-space SP300000
character.
As well, the comments about copy\paste of the resulting character or
otherwise using the characters in messages is potentially problematic,
if not properly accounting for CCSID; i.e. files and messages both have
CCSID support, so pasting characters according to code point
irrespective of the CCSID tagging is prone to errors being manifest, at
least without understanding what is a code point versus a character.
Similar to how presenting the table at that link, without explicitly
naming what Code Page is being represented, was called-out even within
that same text snippet [at least twice] as being problematic.
The char: Capital Letter A Tilde should be x'66', but it depends on
the QCCSID.
Note: The specific hex code point to enter [using the hex-entry] for
a particular character is specific to the code-page\CCSID of the job.
That CCSID may or may not have been resolved from the system value
QCCSID. And whether the expected glyph appears, is dependent on other
factors beyond just the established job CCSID. I suggest that every
user should have their LANGID() and CCSID() set explicitly to reflect
their cultural preference, and in that case, the settings that had been
established for the sysval QCCSID should be moot; as well, if the CHGJOB
had been used set an explicit CCSID value.
My preferred reference doc is the list of code pages shown below;
surely a much more definitive source and representative internationally,
in contrast with the one table in the link given in the quoted message:
<www.ibm.com/software/globalization/cp/cp_cpgid.html>
IBM Software -> Globalization -> Coded character sets and related
resources -> Code page identifiers
_Code page identifiers_
"Code pages by CPGID
..."
Or, indirectly obtained via the CCSID values listed here:
<www.ibm.com/software/globalization/ccsid/ccsid_registered.html>
IBM Software -> Globalization -> Coded character sets and related
resources -> Coded character set identifiers
_Coded character set identifiers_
In either case, further drill down will lead to the specific
documents for a given Code Page, with the file names CP#####.pdf and
CP#####.txt [for an EBCDIC SBCS]; e.g. for the typical USEnglish
language environment on IBM i, the two files:
<
ftp://ftp.software.ibm.com/software/globalization/gcoc/attachments/CP00037.pdf>
<
ftp://ftp.software.ibm.com/software/globalization/gcoc/attachments/CP00037.txt>
As an Amazon Associate we earn from qualifying purchases.