Re: Commit: some of Larin's work

From: Dom Lachowicz (domlachowicz@yahoo.com)
Date: Fri Mar 07 2003 - 20:08:44 EST

  • Next message: William Lachance: "Commit: Add Column Import to OpenWriter importer"

    Using this is fine (even encouraged by me):

    static const foo_t foo = somevalue;

    However, AbiWord's coding guidelines forbid the use of
    C++ templates, and even if we do decide to use them,
    it's not happening before 2.0 is out the door. All
    patches to this effect will be sent to /dev/null

    Also, note that these types of macros:

    1) Usually set some member variable to an error state
    (this one does)
    2) Return from the calling method (this one does too)

    Template functions won't suffice here and aren't an
    applicable solution to the problem at hand.

    Dom

    --- Larin Hennessy <larin@science.oregonstate.edu>
    wrote:
    > OK, another C -> C++ change comes to mind. What
    > about
    > changing #define to const and/or inline template
    > functions
    > depending on the situation?
    >
    > For example, the WordPerfect importer/exporter has a
    > group
    > of formatting codes in ie_impexp_WordPerfect.h:
    >
    > #define WP_TOP_SOFT_SPACE 128 // (0x80)
    > ...
    > #define WP_TOP_DISPLAY_NUMBER_REFERENCE_GROUP 0xDA
    >
    > Should these be converted to:
    >
    > const char WP_TOP_SOFT_SPACE = 0x80;
    > ...
    > const char WP_TOP_DISPLAY_NUMBER_REFERENCE_GROUP =
    > 0xDA;
    > ?
    >
    > It seems that if these values represent character
    > values in files, using
    > the hex
    > representation would be more intuitive.
    >
    > I would also like to look at the feasability of
    > changing some of the
    > define statements
    > to templates. For example in ie_imp_DocBook.cpp
    > there is X_CheckDocument:
    >
    > #define X_CheckError(v) do { if
    > (!(v)) \
    >
    > {
    > UT_DEBUGMSG(("DOM: X_CheckError failed: %s\n", #v));
    > \
    >
    >
    > m_error = UT_ERROR; \
    >
    >
    > return; } } while (0)
    >
    > into:
    >
    > UT_Error X_CheckError(char * v)
    > {
    > do
    > {
    > if (!v)
    > {
    > UT_DEBUGMSG("DOM:_CheckError failed:
    > %s\n", v);
    > return UT_ERROR;
    > }
    > } while (0);
    > }
    >
    > Note that the above is just an idea and has not been
    > tested.
    >
    > Any thoughts on wether this would be acceptable?
    >
    > -Larin
    >
    > Dom Lachowicz wrote:
    >
    > >>I think Larin meant ie_exp_WordPerfect.cpp, which
    > >>does have such a list.
    > >>Marc would know more about this (he wrote the
    > code),
    > >>but I believe the
    > >>magic is only a temporary hack until we actually
    > >>figure out what the
    > >>magic means. :-) It should become history..
    > >>eventually.
    > >>
    > >>
    > >
    > >Indeed it does. The (char) should stay for the
    > above
    > >reason, and the fact that 192
    > reinterpret_cast<char>()
    > >is much much worse than 192 (char) from a
    > readability
    > >perspective. reinterpret_cast gains us nothing
    > here.
    > >
    > >On a related note, I just committed a dos2unix
    > version
    > >of exp_WordPerfect.cpp to remove the ^M problem.
    > >
    > >Dom
    > >
    > >__________________________________________________
    > >Do you Yahoo!?
    > >Yahoo! Tax Center - forms, calculators, tips, more
    > >http://taxes.yahoo.com/
    > >
    > >
    >
    >

    __________________________________________________
    Do you Yahoo!?
    Yahoo! Tax Center - forms, calculators, tips, more
    http://taxes.yahoo.com/

    This archive was generated by hypermail 2.1.4 : Fri Mar 07 2003 - 20:15:09 EST