So, the downloader is failing, and without the downloader working it's hard to test that flat_tree is good!
The part of the downloader that is failing presently seems unrelated to flat_tree. I'm stepping up in the callstack. Because by happenstance I made the length_id_paris a generator expression rather than a list, it's actually executing code farther down the function when the variable is used. It's evaluating lazily. When I get far enough up, I can evaluate entryct. (Pdb) p entryct 16503103374363018573310452198762115323676072158390965418376578654689494076191 It looks like there is an error parsing the bundle, that is not being caught. It's possible the bundle format changed! Or it could just be corrupt data.
