Lib.rs

Background of the Free Law Reporters Database

A long, long time ago near a courthouse not too far away, people started keeping books of every important opinion that was ever written. These books became known as reporters and were generally created by librarian-types of yore such as Mr. William Cranch and Alex Dallas.

These people were busy for the next few centuries and created thousands of these books, culminating in what we know today as West's reporters or as regional reporters like the "Dakota Reports" or the thoroughly-named, "Synopses of the Decisions of the Supreme Court of Texas Arising from Restraints by Conscript and Other Military Authorities (Robards)."

In this repository we've taken a look at all these reporters and tried to sort out what we know about them and convert that to data. This data is available as a JSON file, as Python variables, and can be browsed in an unofficial CSV (it's usually out of date).

Naturally, converting several centuries' history into clean data results in a mess, but we've done our best and this mess is in use in a number of projects as listed below. As of version 2.0, this data contains information about 733 reporters, including 1,466 name variations, and 830 editions.

We hope you'll find this useful to your endeavors and that you'll share your work with the community if you improve or use this work.

Data Sourcing

This project has been enhanced several times with data from several sources:

1. The original data came from parsing the citation fields for millions of cases in CourtListener.
2. A second huge push came from parsing metadata obtained from two major legal publishers, and by parsing the citation fields of Havard's Case.law database.
3. An audit was performed and additional fields were added by using regular expressions to find number-word-number strings in the entire Harvard Case.law database. The results of this were sorted by frequency, with the top omissions fixed.

Along the way, small and subdry improvements have been made as gaps were identified and fixed.

The result is that this database should thus be very complete when it comes to reporter abbreviations and variations. It has the data from CourtListener, two major legal publishers, and Harvard's Case.law. Hundreds of hours have gone into this database to make it complete.

CSV

You can make a CSV of this data by running:

make_csv.py

We keep a copy of this CSV in this repository (reporters.csv), but it is not kept up to date. It should, however, provide a good idea of what's here.

Known Implementations

1. This work was originally deployed in the CourtListener citation finder beginning in about 2012. It has been used literally millions of times to identify citations between cases.
2. An extension for Firefox known as the Free Law Ferret uses this code to find citations in your browser as you read things -- all over the Web.
3. A Node module called Walverine uses an iteration of this code to find citations using the V8 JavaScript engine.

Additional usages can be found via Github.

Some Notes on the Data

Some things to bear in mind as you are examining the Free Law Reporters Database:

1. Each Reporter key maps to a list of reporters that that key can represent. In some cases (especially in early reporters), the key is ambiguous, referring to more than one possible reporter.
2. Formats follow the Blue Book standard, with variations listed for local rules and other ways lawyers abbreviate it over the years or accidentally.
3. The variations key consists of data from local rules, found through organic usage in our corpus and from the Cardiff Index to Legal Abbreviations. We have used a dict for these values due to the fact that there can be variations for each series.
4. mlz_jurisdiction corresponds to the work that is being done for Multi-Lingual Zotero. This field is maintained by Frank Bennett and may sometimes be missing values.
5. Some reporters have href or notes fields to provide a link to the best available reference (often Wikipedia) or to provide notes about the reporter itself.
6. Regarding dates of the editions, there are a few things to know. In reporters with multiple series, if multiple volumes have the same dates, this indicates that the point where one series ends and the other begins is unknown. If an edition has 1750 as its start date, this indicates that the actual start date is unknown. Likewise, if an edition has null as its end date, that indicates the actual end date is either unknown, or it's known that the series has not completed. These areas need research before we can release version 1.1 of this database. Finally, dates are inclusive, so the first and last opinions in a reporter series have the same dates as the database.

A complete data point has fields like so:

"$citation": [
    {
        "cite_type": "state|federal|neutral|specialty|SpecialtyWest|specialty_lexis|state_regional|ScotusEarly",
        "editions": {
            "$citation": {
                "end": null,
                "regexes": [],
                "start": "1750-01-01T00:00:00"
            },
            "$citation 2d": {
                "end": null,
                "regexes": [],
                "start": "1750-01-01T00:00:00"
            }
        },
        "examples": [],
        "mlz_jurisdiction": [],
        "name": "",
        "variations": {},
        "notes": "",
        "href": "",
        "publisher": ""
    }
],

Installation (Python)

You can install the Free Law Reporters Database with a few simple commands:

pip install reporters-db

Once installed you can use it in your code with something like:

from reporters_db import REPORTERS

You can see all of the variables that can be imported by looking in __init__.py. Other variables currently include: STATE_ABBREVIATIONS, CASE_NAME_ABBREVIATIONS, SPECIAL_FORMATS, VARIATIONS_ONLY, and EDITIONS. These latter two are convenience variables that you can use to get different views of the REPORTERS data.

Of course, if you're not using Python, the data is in the json format, so you should be able to import it using your language of choice.

Tests

We have a few tests that make sure things haven't completely broken. They are automatically run by Travis CI each time a push is completed and should be run by developers as well before pushing. They can be run with:

python tests.py

It's pretty simple, right?

License

This repository is available under the permissive BSD license, making it easy and safe to incorporate in your own libraries.

Pull and feature requests welcome. Online editing in Github is possible (and easy!)