bryanlammon / supra Goto Github PK

If the same pincite is used in subsequent footnotes, the second one will use a pincite instead of just Id.

Input

Text.^[[@source1] at 1.]

Text.
Text.
Text.^[*See, e.g.*, [@source1] at 1; [@source2] at 2; [@source3] at 3.]

Output

Text.1

Text. Text. Text.2

n.1: Source1, 1 U.S. 1, 1 (2001).

n.2: See, e.g., id. at 1; Source2, 2 U.S. 2, 2 (2002); Source3, 3 U.S. 3 (2003).

Note 2 should start with "See, e.g., id.;".

Sources in parentheticals do not appear to be processed. For example:

[@AlliedAirFreight1968] at 444 (citing [@UnitedStatesWallace1949] at 794).

Renders only the first source, not the second:

Allied Air Freight, Inc. v. Pan Am. World Airways, Inc., 393 F.2d 441, 444 (2d Cir. 1968) (citing [@UnitedStatesWallace1949] at 794).

If (1) a citation is rendered as an Id. without a pincite and (2) the citation ends with a period, and extra period is rendered.

E.g.:

• Input: *Compare* signal.^[Compare [@PlaintiffDefendant2000] at 1002, with [@PlaintiffDefendant2000] at 1002.]
• Expected: *Compare* signal.^[Compare id. at 1002, with id.]
• Result: *Compare* signal.^[Compare id. at 1002, with id..]

There's an error message for when a footnote does not end with a closing bracket. But the same error is thrown thrown if the footnote ends with a citation that lacks any ending punctuation. The error message should be expanded to address both situations.

It looks like BetterBibTex's CSL JSON output is storing dates as either strings or integers. The deserializing code expects integers. So the deserializing code needs to account for both types.

The Makefile example in the readme has an output in markdown format (output.md). It should be a .docx file.

*Id.*s are using pincites even if the previous citation cited to the same page.

If an Id. has the same pincite as the prior cite, it should not render the pincite.

If the input uses a cross-reference that does not exist, the program crashes with an unhelpful error message.

Weight-of-authority parentheticals (e.g., "per curiam," "en banc") are not rendered.

Pincites that include a comma (e.g., at 550, 551 n.12) are not supported.

This was done so that the lexer would capture citations that end with a comma. The thought was that these were more common than pincites with a comma.

But can the lexer capture both?