Function key_split

dask/utils.py:1945–1999

>>> key_split('x') 'x' >>> key_split('x-1') 'x' >>> key_split('x-1-2-3') 'x' >>> key_split(('x-2', 1)) 'x' >>> key_split("('x-2', 1)") 'x' >>> key_split("('x', 1)") 'x' >>> key_split('hello-world-1') 'hello-world' >>> key_split(b'hello-wor

(s)

@functools.lru_cache(100000)
def key_split(s):
    """
    >>> key_split('x')
    'x'
    >>> key_split('x-1')
    'x'
    >>> key_split('x-1-2-3')
    'x'
    >>> key_split(('x-2', 1))
    'x'
    >>> key_split("('x-2', 1)")
    'x'
    >>> key_split("('x', 1)")
    'x'
    >>> key_split('hello-world-1')
    'hello-world'
    >>> key_split(b'hello-world-1')
    'hello-world'
    >>> key_split('ae05086432ca935f6eba409a8ecd4896')
    'data'
    >>> key_split('<module.submodule.myclass object at 0xdaf372')
    'myclass'
    >>> key_split(None)
    'Other'
    >>> key_split('x-abcdefab')  # ignores hex
    'x'
    >>> key_split('_(x)')  # strips unpleasant characters
    'x'
    """
    # If we convert the key, recurse to utilize LRU cache better
    if type(s) is bytes:
        return key_split(s.decode())
    if type(s) is tuple:
        return key_split(s[0])
    try:
        words = s.split("-")
        if not words[0][0].isalpha():
            result = words[0].split(",")[0].strip("_'()\"")
        else:
            result = words[0]
        for word in words[1:]:
            if word.isalpha() and not (
                len(word) == 8 and hex_pattern.match(word) is not None
            ):
                result += f"-{word}"
            else:
                break
        if len(result) == 32 and re.match(r"[a-f0-9]{32}", result):
            return "data"
        else:
            if result[0] == "<":
                result = result.strip("<>").split()[0].split(".")[-1]
            return sys.intern(result)
    except Exception:
        return "Other"


def stringify(obj, exclusive: Iterable | None = None):