Update citation model to handle volumes as strings

[quevon24]

While running columbia merger we found citations in which the volume also contains letters so the current citation model cannot handle that volume number since the field is defined as SmallIntegerField, throws a ValueError exception.

For example:
1985-1 Trade Cases P 66,472
88-1 U.S. Tax Cas. (CCH) P9373

We need to analyze if just changing the type of field is feasible and how it would affect the system, or what change is necessary to store the appointments in which the volume is a string.

This issue was seen before in #1162

[quevon24]

I counted the citations from the columbia matches and we only have five citations where we have a hyphen:

• 1985-1 Trade Cas. (CCH) 66,472
• 1986-2 Trade Cas. (CCH) 67,324
• 1995-2 Trade Cas. (CCH) 71,147
• 1986-1 Trade Cas. (CCH) 66,944
• 1985-2 Trade Cas. (CCH) 66,765

The number of citations of this type in the Columbia dataset is insignificant.

Now I need to count the ones in the lexis and westlaw dataset to see if we have a greater number of citations of this type.