Attempt str conversion when matching in column from regex, capture group optional

annotator/utils.py

@@ -348,9 +348,10 @@ def substituteColumnValues(referenceList, mod):
     return referenceList
 
 
-def colFromRegex(referenceList, regex):
+def colFromRegex(referenceList, regex, silent=False):
     """ Return a list created by mapping a regular expression to another list.
-    The regular expression must contain at least one capture group.
+    If a capture group is included, contents matching the capture group will
+    be used to derive a new value. Otherwise the entire expression is used.
 
     Parameters
     ----------
@@ -364,12 +365,19 @@ def colFromRegex(referenceList, regex):
     The list resulting from mapping `regex` to `referenceList`.
     """
     p = re.compile(regex)
-    if not p.groups:
-        raise RuntimeError("`regex` must have at least one capture group.")
     newCol = []
+    if not all([isinstance(i, str) for i in referenceList]):
+        print("Warning: Attempting to convert values to str.")
+        print("May cause unexpected results.")
     for s in referenceList:
-        m = p.search(s) if isinstance(s, str) else None
-        if not m and isinstance(s, str):
-            print("{} does not match regex.".format(s))
-        newCol.append(m.group(1)) if m else newCol.append(None)
+        m = p.search(str(s))
+        if not m:
+            if not silent:
+                print("{} does not match regex.".format(str(s)))
+            newCol.append(None)
+        else:
+            if p.groups:
+                newCol.append(m.group(1))
+            else:
+                newCol.append(m.group())
     return newCol

tests/unit/test_utils.py

@@ -218,3 +218,12 @@ def test_colFromRegex(self, values):
         # gives a list of the first vowel of each word
         result = annotator.utils.colFromRegex(values, r"([aeiou])")
         assert result == ['u', 'u', 'e', 'e', 'o']
+
+    def test_colFromRegex_non_str(self):
+        values = [[1,2,3], {4,5,6}]
+        result = annotator.utils.colFromRegex(values, r"(\d)")
+        assert result == ["1", "4"]
+
+    def test_colFromRegex_no_capture_group(self, values):
+        result = annotator.utils.colFromRegex(values, r"\we\w")
+        assert result == [None, None, 'red', 'ree', None]