ENH: Improve table to dict parser to accept custom XPath for table cell.

[hhslepicka]

Allow users to specify a custom XPath for the table cell elements for situations in which th is mixed with td on rows.

Here is a real use case from the Nasdaq Screener table:

screener_table = self.find_element(selector="nasdaq-screener__table", by=By.CLASS_NAME)
# Data from the table using a custom cell selector
data = table_to_dict(
    table=screener_table,
    has_header=True,
    cell_xpath="//*[contains(@class,'nasdaq-screener__cell')]"
)