Hi all:
After giving up trying to reach the developers on GitHub, I git cloned
the xsv build and created a new module "lengths.rs" by hacking the
fixlengths.rs module and carefully following conventions in the files
critical to enabling it as a command option.
That said, the build successfully compiled and I have a new xsv option
"lengths" which lists the record # and field length for each record in
the csv file. By piping this command:
% xsv lengths my_file.csv | sort | uniq -c
One can quickly get an idea and count of the actual field lengths in the
csv file. If some records omit or add an extraneous field, then
rerunning the xsv lengths and using grep -v {good length} will show
which records need fixing.
The command is fast, a 1.19 gig file with 2,100,878 records with 79
fields took 21 secs to run.
- Randall
