daff fails to report differences with duplicate column names

[annatindall]

Hi @paulfitz ,

I've been trying to use daff 1.4.2 on the command line with csvs that have duplicate column names, and am seeing some strange behaviour, I think similar to those reported by bluunk in https://github.com/paulfitz/daff/issues/61.

Example:

a.csv:

One,Two,Two A,B,B

b.csv:

One,Two,Two A,C,C

daff --fail-if-diff a.csv b.csv exits with code 0

If I change the headers to "One,Two,Three", daff works as expected

If I change A's data row to "A,B,C", daff outputs

! ,(A)
@@,One

but it does exit with code 1

What's the current state around how daff handles duplicate column names? Thanks!

[oslo254804746]

the same as me

root@320a3160c844:/code/bin# cat test/a.csv 
level1,newEnCol,fs,fs,123,123,123,123
sfs,1,,,,,,
sfs,2,,fs,,,,
sfs,3,,,,,,
root@320a3160c844:/code/bin# cat test/b.csv 
level1,newEnCol,fs,fs,123,123,123,额外列
sfs,1,,,,,,
sfs,2,,fs,,,,
sfs,3,,,,,,

just rename the end 123 col to 额外列, diff output is (I build daff cpp binary)

root@320a3160c844:/code/bin# ./Coopy diff test/a.csv test/b.csv 
@@,level1,newEnCol,fs
root@320a3160c844:/code/bin# ./Coopy diff test/a.csv test/b.csv  --index
@:@,,A:A,B:B,D:D
,@@,level1,newEnCol,fs
root@320a3160c844:/code/bin#