Issue 242 - Fix court detection when parenthetical includes date

CHANGES.md

@@ -21,6 +21,7 @@ Fixes:
   citations #237
 - Update `maybe_balance_style_tags` to account for party names and intro words
   inside the style tag #231
+- Fixes court detection with court/date parenthetical contains the month and day, e.g., `(C.D. Cal. Feb. 9, 2015)` #242
 
 ## Current
 

eyecite/regexes.py

@@ -219,32 +219,43 @@ def short_cite_re(regex):
 # Post full citation regex:
 # Capture metadata after a full cite. For example given the citation "1 U.S. 1"
 # with the following text:
-#   1 U.S. 1, 4-5, 2 S. Ct. 2, 6-7 (4th Cir. 2012) (overruling foo)
+#   1 U.S. 1, 4-5, 2 S. Ct. 2, 6-7 (4th Cir. Jan. 1, 2012) (overruling foo)
 # we want to capture:
 #   pin_cite = 4-5
 #   extra = 2 S. Ct. 2, 6-7
 #   court = 4th Cir.
+#   month = Jan.
+#   day = 1
 #   year = 2012
 #   parenthetical = overruling foo
 POST_FULL_CITATION_REGEX = rf"""
-    (?:  # handle a full cite with a valid year paren:
-        # content before year paren:
+    (?: # handle a full cite with a valid court+date paren:
+        # content before court+date paren:
         (?:
             # pin cite with comma and extra:
             {PIN_CITE_REGEX}?
             ,?\ ?
             (?P<extra>[^(;]*)
         )
-        # content within year paren:
-        [\(\[](?:
-            # court and year:
-            (?P<court>[^)]+)\ {YEAR_REGEX}|
-            # just year:
-            {YEAR_REGEX}
-        )[\)\]]
+        # content within court+date paren:
+        [\(\[] # opening paren or bracket
+        (?:
+            (?:
+                (?P<court>.*?) # treat anything before date as court
+                (?= # lookahead to stop when we see a month or year
+                    \s+{MONTH_REGEX} |
+                    \s+{YEAR_REGEX}
+                )
+            )?
+            \ ?
+            (?P<month>{MONTH_REGEX})?\ ?   # optional month
+            (?P<day>\d{{1,2}})?\,?\ ?      # optional day and comma
+            (?P<year>{YEAR_REGEX})         # year is required
+        )
+        [\)\]] # closing paren or bracket
         # optional parenthetical comment:
         {PARENTHETICAL_REGEX}
-    |  # handle a pin cite with no valid year paren:
+    |  # handle a pin cite with no valid court+date paren:
         {PIN_CITE_REGEX}
     )
 """

tests/test_CourtsTest.py

@@ -32,6 +32,10 @@ def test_parenthetical_court_parser(self):
             "State Kennedy, 666 P.2d 1316, 1326 (Or. 1983)": "or",
             "State v. Michael J., 875 A.2d 510, 534-35 (Conn. 2005)": "conn",
             "Commonwealth v. Muniz, 164 A.3d 1189 (Pa. 2017)": "pa",
+            "Commonwealth v. Shaffer, 209 A.3d 957, 969 (Pa. Jan. 1, 2019)": "pa",
+            "Sixty-Eight Liquors, Inc. v. Colvin, 118 S.W.3d 171 (Ky. Aug. 2, 2003)": "ky",
+            "Wisniewski v. Johns-Manville Corp., 812 F.2d 81, 83 (3rd Cir. June 30, 1987).": "ca3",
+            "Wallace v. Cellco P'ship, No. CV 14-8052-DSF (AS), 2015 WL 13908106, at *7 (C.D. Cal. Feb. 9, 2015)": "cacd",
         }
         for key in samples.keys():
             eyecite_result = get_citations(key)

tests/test_FindTest.py

@@ -151,6 +151,12 @@ def test_find_citations(self):
                             metadata={'plaintiff': 'Commonwealth',
                                       'defendant': 'Muniz',
                                       'court': 'pa'})]),
+            # Test with month/day in court parenthetical
+            ('Commonwealth v. Muniz, 164 A.3d 1189 (Pa. Feb. 9, 2017)',
+             [case_citation(page='1189', reporter='A.3d', volume='164', year=2017,
+                            metadata={'plaintiff': 'Commonwealth',
+                                      'defendant': 'Muniz',
+                                      'court': 'pa'})]),
             # Parallel cite with parenthetical
             ('bob lissner v. test 1 U.S. 12, 347-348, 1 S. Ct. 2, 358 (4th Cir. 1982) (overruling foo)',
              [case_citation(page='12', year=1982,