Part III: Xerces, XML4C, C++, V6R1 - IBM AS400

This is a discussion on Part III: Xerces, XML4C, C++, V6R1 - IBM AS400 ; Success, at least for now, but I'm mightily pissed about it all. I went into the samples and just started copying options from the CL. These included OPTIMIZE, INLINE, TERASPACE, STGMDL, CSOPT. I'm doubtful about OPTIMIZE, probably INLINE and TERASPACE. ...

+ Reply to Thread
Results 1 to 4 of 4

Thread: Part III: Xerces, XML4C, C++, V6R1

  1. Part III: Xerces, XML4C, C++, V6R1

    Success, at least for now, but I'm mightily pissed about it all. I
    went into the samples and just started copying options from the CL.
    These included OPTIMIZE, INLINE, TERASPACE, STGMDL, CSOPT. I'm
    doubtful about OPTIMIZE, probably INLINE and TERASPACE. STGMDL -
    don't know that one but it's likely this CSOPT is to blame. What
    ticks me off is that there is no explanation, no documentation as to
    why these options were chosen. That's what has caused me to burn
    plenty of days on this project.

    CRTCPPMOD
    MODULE(CREATE)
    SRCSTMF('/qibm/proddata/xmltoolkit2/xml5_6_0/samples/CreateDOMDocument/
    CreateDOMDocument.cpp')
    TEXT(*SRCMBRTXT)
    OUTPUT(*print)
    INCDIR('/qibm/proddata/xmltoolkit2/xml5_6_0/include')
    OPTIMIZE(40)
    INLINE(*OFF *AUTO 30 2000)
    TERASPACE(*YES)
    STGMDL(*INHERIT)
    CSOPT('-qnamemangling=v5')

    The CRTPGM command is not what was causing my any trouble, turns out.
    I can leave all OPTIONs off and the result is the same.

    CRTPGM
    PGM(CREATE)
    BNDSRVPGM(QXMLLIB/QXML4C560)
    DETAIL(*FULL)
    OPTION(*DUPVAR *DUPPROC *NOWARN)

    So now instead of compiling samples, I'll try some of my own code.
    Cross your fingers.


  2. Re: Part III: Xerces, XML4C, C++, V6R1

    Mr. K.V.B.L. wrote:
    > CSOPT('-qnamemangling=v5')


    This option tells the compiler howC++ external names should be
    "mangled". In C++ data types, what classes or namespaces contain the
    object, and other things are encoded into the external names. The
    compiler supports multiple ways of doing this and in more recent
    compiler releases the default changed to a newer way of doing it to
    match IBMs other compilers. -qnamemangling=v5 tells the compiler to
    use an older method, probably to match what was done in earlier
    releases of the XML product.


  3. Re: Part III: Xerces, XML4C, C++, V6R1

    On Sep 3, 10:52*am, "Mr. K.V.B.L." wrote:
    > ...What
    > ticks me off is that there is no explanation, no documentation as to
    > why these options were chosen. *That's what has caused me to burn
    > plenty of days on this project.


    /QIBM/ProdData/xmltoolkitv4/xml5_6_3/doc/index.html

    The documentation should be installed into an IFS directory named
    something like that on your system. The 5_6_3 part and xmltoolkitv4
    part will vary depending on which releases you have installed. Don't
    know if what you are looking for is in there.

  4. Re: Part III: Xerces, XML4C, C++, V6R1

    On Sep 4, 7:46*am, WDS wrote:
    > On Sep 3, 10:52*am, "Mr. K.V.B.L." wrote:
    >
    > > ...What
    > > ticks me off is that there is no explanation, no documentation as to
    > > why these options were chosen. *That's what has caused me to burn
    > > plenty of days on this project.

    >
    > /QIBM/ProdData/xmltoolkitv4/xml5_6_3/doc/index.html
    >
    > The documentation should be installed into an IFS directory named
    > something like that on your system. *The 5_6_3 part and xmltoolkitv4
    > part will vary depending on which releases you have installed. *Don't
    > know if what you are looking for is in there.


    /QIBM/ProdData/xmltoolkitv4/xml5_6_3/doc/contents.html is supposed to
    list all the various installed licensed program options that the
    toolkit contains and what compiler options to use. But on my system
    it is named "/QIBM/ProdData/xmltoolkitv4/xml5_6_3/doc/
    contents.htmlgen" for some reason. Maybe the install didn't work
    quite right.

+ Reply to Thread