I'm running Mdk 7.0 and I'm loading some text files into mysql.
1) I cleaned up the files a whole bunch (quick perl scrpt.)

2) Their's a little "dirty" data left, but MySQL jsut gives me a warning
count, not the rows in question, so I'm extracting the loaded tables
back and doing a compare on the two files (should be equal, no?)

I compared one smaller file (420 000 rows x 4 columns) OK, but diff
seems to choke on memory on anything bigger.  The files are similiar,
with no more than 1 or 2 chars different in a couple of rows.

Is there any difference tool where I can say "DO not look more than 30
lines ahead or back to match up lines".  I would think this would keep
memory to almost nothing (60 lines + some overhead).

I tried -d and -H with diff, no luck.  I have 64M + about 100M swap
space.  My only other choice is to
a) Try and add a special swap-space file on a partition
b) Grep and split the files into small chunks
c) Bring the files to work and use diff on AIX with 2 Gb of Ram.
d) Get the real answer from this list.

Please help me out on d) !!!

Eric Aksomitis
http://www.dlcwest.com/~jed/re_answer.shtml

Keep in touch with http://mandrakeforum.com: 
Subscribe the "[EMAIL PROTECTED]" mailing list.

Reply via email to