I've run across an issue when dealing with the Onecode barcode in
BWIPP, and was wondering if something could be done about it.  The
issue is that the speed performance of the algorithm seems to be far
less efficient than that of the postnet barcode.

When we generate, say, 100 consecutive barcodes (lets say I have a
mailing run, for example), the time to generate and print the Onecode
barcodes will be at least three to four times the time it takes to
generate postnet.  Multiply this to a thousand or ten thousand
barcodes--it makes a lot of difference in time!

Is there a way the algorithm can be improved to have a better
performance?

Thanks,
Jean-François

I've just run some basic profiling on the encoder and (unsurprisingly)
I find that almost all of the processing time is spent in the two
routines that look like this:

--------8<--------
    % Conversion from codewords to characters
    /tab513 1287 dict def
    /lo 0 def
    /hi 1286 def
    0 1 8191 {
        { % no loop - provides common exit point
            /i exch def
            /onbits 0 def
            0 1 12 {
                i exch 1 exch bitshift and 0 ne {
                    /onbits onbits 1 add def
                } if
            } for
            onbits 5 ne {exit} if
            /j i def
            /rev 0 def
            16 {
                /rev rev 1 bitshift j 1 and or def
                /j j -1 bitshift def
            } repeat
            /rev rev -3 bitshift def
            rev i lt {exit} if
            rev i eq {
                tab513 hi i put
                /hi hi 1 sub def
            } {
                tab513 lo i put
                tab513 lo 1 add rev put
                /lo lo 2 add def
            } ifelse
            exit
        } loop
-------->8--------

Whilst this code could be slightly optimised by converting the dicts
into arrays and through some cunning bitwise logic, all the algorithm
really does is to generates static lookup tables which are not worth
the huge processing expense. I have simply replaced the algorithms
with hardcoded tables of constants in the new release.

Please confirm that the latest release (2008-07-10) works for you.

Warm regards,

Tez

Yes, the latest release of BWIPP has a huge increase in performance
when dealing with the OneCode barcode!

Thanks!
Jean-François

On 10 Jul, 16:15, "Terry Burton" <t...@terryburton.co.uk> wrote: