ALCourt Objects#

class ALCourt(Court)

Object representing a court in Massachusetts. TODO: it could be interesting to store a jurisdiction on a court. But this is non-trivial. Should it be geo boundaries? A list of cities? A list of counties? Instead, we use a function on the CourtList object that filters courts by address and can use any of those three features of the court to do the filtering.


| short_label() -> str

Returns a string that represents a nice, disambiguated label for the court. This may not match the court's name. If the name omits city, we append city name to the court name. This is good for a drop-down selection list.


| short_label_and_address() -> str

Returns a markdown formatted string with the name and address of the court. More concise version without description; suitable for a responsive case.


| short_description() -> str

Returns a Markdown formatted string that includes the disambiguated name and the description of the court, for inclusion in the results page with radio buttons.


| from_row(df_row, ensure_lat_long=True) -> None

Loads data from a single Pandas Dataframe into a court object. Note: It will try to convert column names that don't make valid attributes. Best practice is to use good attribute names (no spaces) that don't interfere with existing attributes or methods of DAObject

ALCourtLoader Objects#

class ALCourtLoader(DAObject)

Object to hold some methods surrounding loading/filtering courts.

Built around Pandas dataframe.


| filter_courts(court_types: list, column='department') -> list

Return a subset of courts, only the name column and index.

If you do not want the list to be filtered, set court_types to None (or falsy value)


| as_court(intrinsicName, index, ensure_lat_long=True)

Return the court at the specified index as an ALCourt object