in reply to Re^2: DBD::CSV Update
in thread DBD::CSV Update

Would it be a sensible suggestion to use a byte offset (as in seek) for this option?
That would make it possible to have binary data in that "header".

Replies are listed 'Best First'.
Re^4: DBD::CSV Update
by Tux (Canon) on Jun 01, 2015 at 13:19 UTC

    Sounds like a recipe for failure. How do you think Joe Programmer will check the exact amount of *bytes* to the real data? Think UTF-8 and JPEG.

    As an alternative, one could think of a start-of-data-tag.

    Both however will only make life for DBD::CSV harder, as it has to put the same data back on the same place on updates. I'm still not convinced this would all be worth the trouble.


    Enjoy, Have FUN! H.Merijn
[reply]

      Don't know about Joe, but I'd fire up a hex editor/viewer to look at the position :-)

      The thought leading to my suggestion was something with an SFX-like header, where the byte count would be known in advance. However, that was just a theoretical idea, so …

[reply]

In Section Seekers of Perl Wisdom