sgi.insert_session crashes if first name has a space

[zoldello]

Describe the bug
if experimenter has a space in hs/her first name in the NWB file, like "Kyu Hyun" (this is Kyu's real first name), spyglass crashes in sgi.insert_session. This is related is #809 and may be a band-aid for it.

To Reproduce

1. create a nwb file with an experimenter named "Kyu Hyun Lee"
2. sgi.insert_session crashes and leaving the data insertion in an complete state

Expected behavior
sgi.insert_session to not crash and populate relevant tables

Potential solution
I am trying to meet a deadline and no one else is identifying this issue. So, I will polish up the potential fix, write unit test and make a PR later in the week. A potential untested solution is common_lab.py -

def decompose_name(full_name: str) -> tuple:
    """Return the full, first and last name parts of a full name.

    Names capitalized. Decomposes either comma- or space-separated.

    Parameters
    ----------
    full_name: str
        The full name to decompose.

    Returns
    -------
    tuple
        A tuple of the full, first, last name parts.
        
        
    Raises
    ------
    KeyError
        When full_name is in an unsupported format
    
    """
    
    # Notes - 
    #(1) spyglass supports having space in a first name, but not the last name.
    #(2) regex was considered but abandoned as it is overkill
    
    full_name_trimmed = None if not full_name else full_name.strip()  
    
    if not full_name_trimmed or len(full_name_trimmed) == 0:
        raise ValueError(f"Name is none/null or empty, Skipping {full_name}")
    elif full_name_trimmed.count(", ") > 1 or full_name_trimmed.count(",") > 1:
        raise ValueError(
            f"Names has multiple commas. Skipping {full_name}"
        )
    elif full_name_trimmed.count(", ") == 0 and full_name_trimmed.count(' ') == 0:
        raise ValueError(
            f"One-syllable name not supported. Skipping {full_name}"
        )
    elif ", " in full_name_trimmed:
        last, first = full_name_trimmed.title().rsplit(", ", 1)
    else:
        first, last = full_name_trimmed.title().rsplit(" ", 1)

    full = f"{first} {last}"

    return full, first, last

[edeno]

Also related to #304

[CBroz1]

I have some suggestions for the draft above to reduce the number of checks/raises

def decompose_name(full_name: str) -> tuple:
    """Return the full, first and last name parts of a full name.

    Names capitalized. Decomposes either comma- or space-separated.
    If name includes spaces, must use exactly one comma.

    Parameters
    ----------
    full_name: str
        The full name to decompose.

    Returns
    -------
    tuple
        A tuple of the full, first, last name parts.

    Raises
    ------
    ValueError
        When full_name is in an unsupported format
    """

    # default to empty string if None so that .count() doesn't fail
    full_trimmed = "" if not full_name else full_name.strip()

    # decide case early to avoid multiple checks
    # catches empty string, multiple commas, multispace w/o comma
    delim = (
        ", "
        if full_trimmed.count(", ") == 1
        else " "
        if full_trimmed.count(" ") == 1
        else None
    )

    # rsplit -> split bc exactly one enforced above
    # .title removed to avoid enforcing unwanted capitalization
    #     on names like "McDonald" -> "Mcdonald"
    # running split w/None delim still works
    parts = full_trimmed.split(delim)

    if delim is None or len(parts) != 2:  # catch unsupported format
        raise ValueError(
            f"Name has unsupported format for {full_name}. \n\t"
            + "Must use exactly one comma+space (i.e., ', ') or space.\n\t"
            + "And provide exactly two name parts.\n\t"
            + "For example, 'First Last' or 'Last1 Last2, First1 First2' "
            + "if name(s) includes spaces."
        )

    first, last = parts if delim == " " else parts[::-1]
    full = f"{first} {last}"

    return full, first, last