We've talked about switching to only allowing lookup of IUCR code with ILCS bits. If we were to make this change, it would put the responsibility on the client to do their own parsing of ILCS reference strings, parsing which we could at least try to help with.

The simplest thing would be to provide a regex with which to parse ILCS reference strings as a constant, like ILCS_FORMAT. Another possibility would be to expose a method which is capable of doing some parsing for the client.

Problems arise in that there might be multiple formats in which ILCS data could appear, which suggests a more involved approach as opposed to providing a single regex.

In an effort to handle #3, I believe a bug has been introduced. Adding subsection arguments into the lookup_by_ilcs method signature (which @ghing did initially) has caused a problem, in that it appears to have broken the functionality of looking up an IUCR code given an ILCS reference, for statutes which are specified past the section level.

For now I am ignoring this, but an example would be the following ILCS statute in the crosswalk document:

720-550/4(a)

I found the following section in the Cook County Jail data, but not in our IUCR crosswalk:

625-5/4-103.2

This section exists, and is visible in the online ILCS reference, if you scroll down sufficiently. But it is not in our IUCR crosswalk document. Instead, only the following section exists:

625-5/4-103

Now, we could add this example to our dataset. Maybe we should see if there are other missing sections from the IUCR crosswalk we have, based on the two datasets we have available, and try and add them all?

Hannah Cushman at datamade pointed out that the state released new IUCR codes. We should update this library to reflect the new codes.