mykael
07-Jun-2007, 01:47 AM
I've got a version of BC that identifies itself as 2.4 (build 240). Is that the latest one available?
I'm getting an alignment problem using the standard rules, where it won't match two lines that are skewed by about 800 lines. This is using a custom ruleset, skew offset is set to about 4000 lines and there are alignment rules that match the lines, with a weight of about 20,000. In one of the files is a line classified as 'unimportant' (actually a blank line) and that seems to be what's triggering it to get its comparison wrong.
I'm seeing:
ddddddddddddddd dddddddddddddd
uuuuuuuuuuuuuuu uuuuuuuuuuuuuu
aaaaaaaaaaaaaaa aaaaaaaaaaaaaa
bbbbbbbbbbbbbbb //////////////
bbbbbbbbbbbbbbb //////////////
bbbbbbbbbbbbbbb //////////////
/////////////// uuuuuuuuuuuuuu
/////////////// dddddddddddddd
/////////////// cccccccccccccc
/////////////// cccccccccccccc
/////////////// cccccccccccccc
uuuuuuuuuuuuuuu uuuuuuuuuuuuuu
ddddddddddddddd dddddddddddddd
ccccccccccccccc //////////////
ccccccccccccccc //////////////
ccccccccccccccc //////////////
uuuuuuuuuuuuuuu //////////////
ddddddddddddddd //////////////
eeeeeeeeeeeeeee eeeeeeeeeeeeee
eeeeeeeeeeeeeee eeeeeeeeeeeeee
..it does get back into sync a few thousand lines later.
Using the advanced comparison does get the result I expect (aligning properly on the ddddddd lines), but it takes a long time before there's any screen output and I'm dealing with big (20 MB+) files (program trace outputs).
I'm getting an alignment problem using the standard rules, where it won't match two lines that are skewed by about 800 lines. This is using a custom ruleset, skew offset is set to about 4000 lines and there are alignment rules that match the lines, with a weight of about 20,000. In one of the files is a line classified as 'unimportant' (actually a blank line) and that seems to be what's triggering it to get its comparison wrong.
I'm seeing:
ddddddddddddddd dddddddddddddd
uuuuuuuuuuuuuuu uuuuuuuuuuuuuu
aaaaaaaaaaaaaaa aaaaaaaaaaaaaa
bbbbbbbbbbbbbbb //////////////
bbbbbbbbbbbbbbb //////////////
bbbbbbbbbbbbbbb //////////////
/////////////// uuuuuuuuuuuuuu
/////////////// dddddddddddddd
/////////////// cccccccccccccc
/////////////// cccccccccccccc
/////////////// cccccccccccccc
uuuuuuuuuuuuuuu uuuuuuuuuuuuuu
ddddddddddddddd dddddddddddddd
ccccccccccccccc //////////////
ccccccccccccccc //////////////
ccccccccccccccc //////////////
uuuuuuuuuuuuuuu //////////////
ddddddddddddddd //////////////
eeeeeeeeeeeeeee eeeeeeeeeeeeee
eeeeeeeeeeeeeee eeeeeeeeeeeeee
..it does get back into sync a few thousand lines later.
Using the advanced comparison does get the result I expect (aligning properly on the ddddddd lines), but it takes a long time before there's any screen output and I'm dealing with big (20 MB+) files (program trace outputs).