LazyTree construction errors : UndefVarError: TLeafS / UndefVarError: TBasket not defined

[Momox357]

Hello, first of all thanks for this amazing package and the fast reaction times (I've started looking into it on monday and experienced the compling error as my first obstacle)!

I've managed to read in the sample file in the quick start sections and compare that data / a generated histogram with one generated from ROOT / with the TBrowser.

Now I'm trying to do that with .root files from ATLAS OpenData (since they have the notebooks for comparison), but for the files I've tried it reads the files but doesn't display it in a "tree-like fashion" though I'm not sure if that limited / just not done for files too large for example.
My main issue is that I'm getting an error when trying to contruct a LazyTree.

For my first case (as far as I understand it "grabs" the whole tree, while ftree=LazyTree(f, "mini", ["lep_n"]) is the specific branch)

Using UnROOT

f=ROOTFile("http://opendata.atlas.cern/release/samples/MC/mc_105987.WZ.root")
ftree=LazyTree(f, "mini")

I get "UndefVarError: TLeafS not defined" .

For two others (https://atlas-opendata.web.cern.ch/atlas-opendata/samples/2020/GamGam/Data/data_A.GamGam.root and https://atlas-opendata.web.cern.ch/atlas-opendata/samples/2020/1largeRjet1lep/MC/mc_361106.Zee.1largeRjet1lep.root )
I get "UndefVarError: TBasket not defined" in that case.

I've found this issue while looking through the github.
That said I'm not sure if it's a potential bug, a file issue or an user issue (since I'm new to Root and Julia :) ).

[Moelf]

thanks for the report, I think it's the same problem with:

• ERROR: TBasket not defined #14
  basically we need to just open/skip a TBasket when we see one

[Moelf]

can you try the PR, that fixes your first issue.

[Moelf]

https://opendata.cern.ch/record/12341

this is an opendata in a widely used format (nanoAOD), we have extensive tests with nanoAOD format so everything should just work

[Momox357]

can you try the PR, that fixes your first issue.

Yes it works, thanks! That was really fast.