IDS Forum

Re: Informix 12 Unicode

Posted By: Fernando Nunes
Date: Wednesday, 22 October 2014, at 9:33 a.m.

In Response To: Re: Informix 12 Unicode (Fernando Nunes)

Related RFEs are:

Character length semantics instead of byte length semantics:
http://www.ibm.com/developerworks/rfe/execute?use_case=viewRfe&CR_ID=47787

New National Language Support data type with variable size greater then 255
bytes:
http://www.ibm.com/developerworks/rfe/execute?use_case=viewRfe&CR_ID=45681

UTF8 should be handled in background instead of setting SQL_LOGICAL_CHAR
(duplicate of 34405):
http://www.ibm.com/developerworks/rfe/execute?use_case=viewRfe&CR_ID=34406

UTF8 should be handled in background instead of setting SQL_LOGICAL_CHAR
http://www.ibm.com/developerworks/rfe/execute?use_case=viewRfe&CR_ID=34405

This last one is classified as "Under Consideration"
Regards

On Wed, Oct 22, 2014 at 12:50 PM, Fernando Nunes <domusonline@gmail.com>
wrote:

> That is good. But I think the issue the OP is facing (which I think is very
> common), is that it's impossible to know how many characters we're able to
> store in a column, because we don't guarantee that...
> This seems to be a fundamental design structure... Big O for example allows
> you to specify:
>
> CREATE TABLE ... ( col1 VARCHAR2(30 CHAR))
>
> This means you'll always be able to fit 30 characters, no matter how many
> bytes it takes. There is a disclaimer stating that if you do something
> like:
>
> CREATE TABLE... ( col1 VARCHAR2(4000)) -- 4000 is the "hard limit" of
> VARCHAR2.
>
> Then it's not guaranteed that you'll be able to store 4000 characters... it
> will accept up to the maximum bytes (4000) allowed in the datatype).
> I think big O is also able to "extend" CHAR fields... but I'm not really
> sure.
>
> It could be easier to implement this (just for VARCHAR) in Informix....
> Since by design (AFAIK) varchar(N) doesn't impose any physical limit on the
> row.
> My concern is that because our VARCHAR() limit is relatively low (256) that
> this could easily lead to columns exceeding the byte limit....
>
> And this, in turn, leads to the other possible discussion about this: We
> don't raise an error when the user tries to exceed the limit.....
> (INSERTing a VARCHAR(200) in a VARCHAR(100) column). We just truncate the
> data without warning.
> >From my point of view (and I recently had customers complaining about
> that)
> this is an undesirable behavior (not a bug as it works as design).
> There is an RFE for this.... and because we raise an error on ANSI
> databases, I'd expect this change to classify as "very possible", but
> without seeing the code I can only imagine.
>
> So.... in short I would:
> 1- Open an RFE to request the specification of CHAR (as opposed to bytes)
> for variable length fields (VARCHAR/LVARCHAR)
> 2- Vote for the existing RFE to create a configuration option ($ONCONFIG,
> SET ENVIRONMENT....) to force the raise of an error if we insert a
> character string that does not fit the column. That RFE is:
> http://www.ibm.com/developerworks/rfe/execute?use_case=viewRfe&CR_ID=33830
>
> There seems to be a duplicate (not yet classified as such):
>
> http://www.ibm.com/developerworks/rfe/execute?use_case=viewRfe&CR_ID=53804
>
> Hopefully if the first one gets implemented, R&D will consider my
> comments.... I think the original request for a "warning" is not enough. An
> error is necessary in order to be useful for existing applications.
>
> Regarding the first RFE, AFIK it's not inserted yet... but a carefull
> search would be needed. Implementing it for VARCHAR seems much easier than
> for CHAR...
> Regards.
>
> On Wed, Oct 22, 2014 at 11:41 AM, Martin Fuerderer <MARTINFU@de.ibm.com>
> wrote:
>
> > Hi,
> >
> > you may use function ifx_gl_mb_loc_max() in
> > an application to determine, how "wide" a
> > character can be in the current locale.
> >
> > Regards, Martin
> > --
> > Martin Fuerderer
> > IBM Informix Development Munich, Germany
> > Information Management
> >
> > Read about the Informix Warehouse Accelerator:
> > http://tinyurl.com/the-iwa-blog
> >
> > IBM Deutschland Research & Development GmbH
> > Chairman of the Supervisory Board: Martina Koederitz
> > Board of Management: Dirk Wittkopp
> > Corporate Seat: Boeblingen, Germany
> > Reg.-Gericht: Amtsgericht Stuttgart, HRB 243294
> >
> > From: "MATTI JAATINEN" <matti.jaatinen@norelco.fi>
> > To: ids@iiug.org
> > Date: 10/22/2014 10:14
> > Subject: Re: Informix 12 Unicode [34012]
> > Sent by: ids-bounces@iiug.org
> >
> > Hi,
> >
> > http://www.iiug.org/forums/ids/index.cgi/read/34002
> >
> > Char_length return length of the characters in data not the length of
> > column. Example you have
> >
> > column sar1 char(5) and there is row which data is '4ä4'.
> > Char_length(sar1)
> > return 3 pecause there is 3 characters in that row. Length(sar1) return 4
> > pecause there is 4 bytes in that row.
> >
> > The problem is how to determine how many characted can be put into the
> > column
> > sar1. Informix reports column size as byte-size so i appears that there
> is
> >
> > enought space for 5 characters, but when we use myltibyte characters eq.
> > '&#39740;' there can be put only 1-characted into column sar1.
> >
> >
> >
> >
>
> *******************************************************************************
> >
> > Forum Note: Use "Reply" to post a response in the discussion forum.
> >
> >
> >
> >
>
> *******************************************************************************
> > Forum Note: Use "Reply" to post a response in the discussion forum.
> >
> >
>
> --
> Fernando Nunes
> Portugal
>
> http://informix-technology.blogspot.com
> My email works... but I don't check it frequently...
>
> --001a113f7d5e6069310506018ca1
>
>
>
> *******************************************************************************
> Forum Note: Use "Reply" to post a response in the discussion forum.
>
>

--
Fernando Nunes
Portugal

http://informix-technology.blogspot.com
My email works... but I don't check it frequently...

--001a113ee16eaa6d39050602fdf0

Messages In This Thread

• Informix 12 Unicode
  MATTI JAATINEN -- Tuesday, 21 October 2014, at 5:34 a.m.
  • Re: Informix 12 Unicode
    MATTI JAATINEN -- Tuesday, 21 October 2014, at 9:16 a.m.
    • Re: Informix 12 Unicode
      MATTI JAATINEN -- Tuesday, 21 October 2014, at 9:46 a.m.
      • Re: Informix 12 Unicode
        Fernando Nunes -- Tuesday, 21 October 2014, at 11:15 a.m.
        • Re: Informix 12 Unicode
          Fernando Nunes -- Tuesday, 21 October 2014, at 11:58 a.m.
          • Re: Informix 12 Unicode
            MATTI JAATINEN -- Wednesday, 22 October 2014, at 1:50 a.m.
            • Re: Informix 12 Unicode
              Martin Fuerderer -- Wednesday, 22 October 2014, at 3:56 a.m.
              • Re: Informix 12 Unicode
                MATTI JAATINEN -- Wednesday, 22 October 2014, at 4:14 a.m.
                • Re: Informix 12 Unicode
                  Martin Fuerderer -- Wednesday, 22 October 2014, at 6:41 a.m.
                  • Re: Informix 12 Unicode
                    Fernando Nunes -- Wednesday, 22 October 2014, at 7:50 a.m.
                    • Re: Informix 12 Unicode
                      Fernando Nunes -- Wednesday, 22 October 2014, at 9:33 a.m.
                  • Re: Informix 12 Unicode
                    Fernando Nunes -- Wednesday, 22 October 2014, at 9:24 a.m.
                    • Re: Informix 12 Unicode
                      MATTI JAATINEN -- Thursday, 23 October 2014, at 1:11 a.m.
            • Re: Informix 12 Unicode
              Fernando Nunes -- Wednesday, 22 October 2014, at 5:30 a.m.
  • Re: Informix 12 Unicode
    Fernando Nunes -- Tuesday, 21 October 2014, at 9:58 a.m.
    • Re: Informix 12 Unicode
      MATTI JAATINEN -- Tuesday, 21 October 2014, at 10:13 a.m.
      • Re: Informix 12 Unicode
        Fernando Nunes -- Tuesday, 21 October 2014, at 11:07 a.m.

IDS Forum is maintained by Administrator with WebBBS 5.12.