multiprocessing code looks to be broken on windows

[jbellis]

phasher = PHash()
encodings = phasher.encode_images(image_dir)
duplicates = phasher.find_duplicates(image_dir)

->

2023-05-03 08:55:03,150: INFO Start: Calculating hashes...
Traceback (most recent call last):
  File "<string>", line 1, in <module>
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\spawn.py", line 116, in spawn_main
    exitcode = _main(fd, parent_sentinel)
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\spawn.py", line 125, in _main
    prepare(preparation_data)
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\spawn.py", line 236, in prepare
    _fixup_main_from_path(data['init_main_from_path'])
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\spawn.py", line 287, in _fixup_main_from_path
    main_content = runpy.run_path(main_path,
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\runpy.py", line 289, in run_path
    return _run_module_code(code, init_globals, run_name,
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\runpy.py", line 96, in _run_module_code
    _run_code(code, mod_globals, init_globals,
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "E:\Projects\pydedup\pydedup.py", line 52, in <module>
    encodings = phasher.encode_images(image_dir)
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\site-packages\imagededup\methods\hashing.py", line 161, in encode_images
    hashes = parallelise(function=self.encode_image, data=files, verbose=self.verbose, num_workers=num_enc_workers)
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\site-packages\imagededup\utils\general_utils.py", line 65, in parallelise
    pool = Pool(processes=num_workers)
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\context.py", line 119, in Pool
    return Pool(processes, initializer, initargs, maxtasksperchild,
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\pool.py", line 215, in __init__
    self._repopulate_pool()
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\pool.py", line 306, in _repopulate_pool
    return self._repopulate_pool_static(self._ctx, self.Process,
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\pool.py", line 329, in _repopulate_pool_static
    w.start()
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\process.py", line 121, in start
    self._popen = self._Popen(self)
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\context.py", line 336, in _Popen
    return Popen(process_obj)
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\popen_spawn_win32.py", line 45, in __init__
    prep_data = spawn.get_preparation_data(process_obj._name)
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\spawn.py", line 154, in get_preparation_data
    _check_not_importing_main()
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\multiprocessing\spawn.py", line 134, in _check_not_importing_main
    raise RuntimeError('''
RuntimeError:
        An attempt has been made to start a new process before the
        current process has finished its bootstrapping phase.

        This probably means that you are not using fork to start your
        child processes and you have forgotten to use the proper idiom
        in the main module:

            if __name__ == '__main__':
                freeze_support()
                ...

        The "freeze_support()" line can be omitted if the program
        is not going to be frozen to produce an executable.

and then

Traceback (most recent call last):
  File "E:\Projects\pydedup\pydedup.py", line 52, in <module>
    duplicates = phasher.find_duplicates(image_dir)
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\site-packages\imagededup\methods\hashing.py", line 303, in find_duplicates
    result = self._find_duplicates_dir(
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\site-packages\imagededup\methods\hashing.py", line 364, in _find_duplicates_dir
    results = self._find_duplicates_dict(
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\site-packages\imagededup\methods\hashing.py", line 230, in _find_duplicates_dict
    result_set = HashEval(
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\site-packages\imagededup\handlers\search\retrieval.py", line 82, in __init__
    self._fetch_nearest_neighbors_bktree()
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\site-packages\imagededup\handlers\search\retrieval.py", line 153, in _fetch_nearest_neighbors_bktree
    built_tree = BKTree(self.test, self.distance_invoker)  # construct bktree
  File "C:\Users\jbell\mambaforge\envs\pydedup\lib\site-packages\imagededup\handlers\search\bktree.py", line 38, in __init__
    self.ROOT = self.all_keys[0]
IndexError: list index out of range

[jbellis]

apparently this is a "feature" of the underlying multiprocessing library, changing my code to

if __name__ == '__main__':
  ...
  phasher = PHash()
  encodings = phasher.encode_images(image_dir)
  duplicates = phasher.find_duplicates(image_dir)

fixes it. might want to drop a warning in the docs tho.

[tanujjain]

Thanks for the report, will make a note in the documentation.

[christopherquenneville]

@jbellis, I'm sorry, could you help my smooth brain understand what I have to modify from the workflow to fix this?

from imagededup.methods import PHash
phasher = PHash()

encodings = phasher.encode_images(image_dir='path/to/image/directory')

duplicates = phasher.find_duplicates(encoding_map=encodings)

from imagededup.utils import plot_duplicates
plot_duplicates(image_dir='path/to/image/directory',
                duplicate_map=duplicates,
                filename='ukbench00120.jpg')

Tried to integrate what you wrote but I don't understand where to put it and what to replace

if __name__ == '__main__':
  ...
  phasher = PHash()
  encodings = phasher.encode_images(image_dir)
  duplicates = phasher.find_duplicates(image_dir)

[jbellis]

Just put everything (or everything except the imports, either way) in an "if name == 'main'" block. https://stackoverflow.com/questions/419163/what-does-if-name-main-do

this prevents multiprocessing from entering an infinite loop when it imports the script in the subprocesses: https://stackoverflow.com/questions/20222534/python-multiprocessing-on-windows-if-name-main

[tanujjain]

Updated documentation in #238