"upload-process --exclude" flags don't seem to work when uploading reports

[mckern]

While trying to exclude paths from search when running upload-process, I found that excluded paths were still being searched (and producing false-positive uploads of un-processable data). I tried passing them as relative and absolute paths before rolling up my sleeves and splashing a hundred logger.debug() calls through the code to trace it out. What I found follows:

When select_file_finder() is called, it gets a list of Path objects. But search_files() compares this as a list of strings; if d in folders_to_ignore is never true, so excluded paths still end up in the search tree when scanning for reports. I don't really know Python as much as I know how to be clumsy with Python but search_files() seems to expect a List(str). While I'm unfamiliar with Typing it looks like there's no implicit casting for a List(Path) to a List(str)? Here's a very, very trivial example I worked up based on the code from folder_searcher.py:

from pathlib import Path

paths = [Path("example1"), Path("example2"), Path("example3")]
strings = ["example1", "example2", "example3"]

matches = set(d for d in paths if d in strings)
print(f"paths that match: {matches}")

$ python3 example.py
paths that match: set()

Compared to:

from pathlib import Path

paths = [Path("example1"), Path("example2"), Path("example3")]
strings = ["example1", "example2", "example3"]

matches = set(d for d in paths if str(d) in strings)
print(f"paths that match: {matches}")

$ python3 example.py
paths that match: {PosixPath('example1'), PosixPath('example3'), PosixPath('example2')}

---

I'm not sure it's the cleanest way, but everything "just works" if I cast everything passed as folders_to_ignore to strings during FileFinder initialization but this feels kind of dodgy/hamfisted to me (again, not actually good at Python):

def select_file_finder(
    root_folder_to_search,
    folders_to_ignore,
    explicitly_listed_files,
    disable_search,
    report_type="coverage",
):
    # cast everything passed in folders_to_ignore as a string
    folders_to_ignore = [str(dir) for dir in folders_to_ignore]
    return FileFinder(
        root_folder_to_search,
        folders_to_ignore,
        explicitly_listed_files,
        disable_search,
        report_type,
    )