Add allowDuplicateHeaders option to configuration and update tests

[tnrich]

The automatic duplicate header parsing introduced in a recent version, while helpful for new development, has broken existing legacy code in our codebase.

This PR adds a flag allowDuplicateHeaders which can be set to true to revert to the earlier functionality (as seen in v5.2.0 - that's the version we were on)

Let me know if this PR needs anything else and thanks for the consideration! Love papaparse :)

[pokoli]

A list of renamedHeaders is returned as part of the result, so you can read the values and keep the previous behaviour.

You just need to adapt your legacy code to read this new data.

There is nothing to do from the PapaParse side.

[tnrich]

@pokoli that isn't entirely true.. the other benefit to my legacy code from this option would be to prevent transformHeader from being run multiple times for the same header, see - #1029

Unfortunately the code added to rename the headers causes the tranformHeader function to be called a 2nd time which breaks code in our codebase that is depending on that function being run only once per header (which I don't think was an unreasonable assumption and allowed us to do custom logic in our CSV parser)

Why not add this simple option to optionally undo a BREAKING change that was pushed in a MINOR version? This doesn't seem like an unreasonable request and I believe would give an immediate fix to a whole host of tickets related to the changes made after v5.3.2 -

#1054 #1054 #998 #1006

I'm not suggesting that those shouldn't be fixed but it isn't a good practice to push a breaking change in a minor version without any recourse for end users.

[tnrich]

Hope the above comment doesn't come off as harsh. Appreciate your stewardship of this important repo @pokoli! Overall I agree that the header de-dupe is a step in the right direction, unfortunately a lot of legacy code will be broken by the two changes (automatic header dedupe and additional calling of transformHeader).

[Targzp]

I am disappointed that the author did not respond to your follow-up questions. I also encountered the same problem.

[danieljbrooks-GitHub]

And I as well