KCodecs - Quick Review

[John Layt]

Hi,

I know nothing about text codecs, but I've had a *very* quick look at KCodecs:

* Original code by Lars dated 1999!
* One method marked as deprecated to be removed for KDE4
* "###FIXME KDE4: the name of the encodings should mostly be uppercase"
* Code generated by script generate_string_table.pl located in kdesdk/scripts
* Algorithms marked as copyright by RSA Data Security and others, but no 
mention what the original licence was or real link to original source
* Encoding probers and lookup tables marked as copyright Mozilla 1998, X11 
license
* kentities.c is documented as generated by gperf from either kentities.gperf 
and/or khtmlentities.gperf but neither are in kcodecs, instead they are in 
khtml as is another copy of kentities.c.
* Public API using boolean parms

This suggests it could do with some attention:
* Check still valid to remove deprecated code?
* Check if encoding names should be made uppercase?
* The generate_string_table.pl script should probably be moved into KCodecs, 
unless it has more general use?
* The RSA and other algorithms may need checking for licensing issues, or at 
least improve the license documentation?
* The probers may need to be checked they are still up to date with the 
original Mozilla code and look-up tables?
* kentities.c needs investigation and I suspect moving all the files from 
khtml to kcodecs, with khtml then using kcodecs?  Or at least docs added that 
this is where it comes from and should be kept in sync.

I wonder how much of this functionality is now done in Qt5?  Would it benefit 
from a functional review by someone who knows what they're doing, like Thiago 
or David?

Cheers!

John.