Danek Duvall wrote:
Brock Pytlik wrote:
Webrev:
http://cr.opensolaris.org/~bpytlik/ips-14307-v1/
As far as I can tell, this looks okay. I don't see how the enumerate()
calls in transform_main_dict_line() are useful (previously, but now no
longer used?), and I feel like there has to be some way to collapse the
four for loops into something smaller, but given that one of them requires
the quote()/unquote() call, and they're readable enough as it is, it's
not a huge deal.
Yeah, I'm not sure how at the moment. I suppose a stack of list
comprehensions is at least possible, but I'm not sure that'd be any more
understandable.
To test this, I've diffed the files this code writes on both sparc
and x86 and verified that the indexes. produced are identical.
What's the test dataset?
When you asked, it was my desktop and a sparc machine. Now, I've also
indexed the entire dev repo and compared it. The dictionary produced is
identical. The time taken has dropped from 1 hr 40 mins, to 1hr 22 mins.
Brock
Thanks,
Danek
_______________________________________________
pkg-discuss mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/pkg-discuss
