bugfixs; closes #16

CHANGELOG.md

@@ -1,5 +1,11 @@
 # Changes
 
+## [Unreleased]
+
+- Fixed bug whereby charstate labels with illegal tokens were incorrectly read.
+- Fixed bug whereby the last charstate label was never detected as duplicate.
+
+
 ## [v1.9.1] - 2023-11-25
 
 - Fixed support for writing comments.

src/commonnexus/blocks/characters.py

@@ -549,10 +549,15 @@ def __init__(self, tokens, nexus=None):
                 if isinstance(w, Token) and w.text == '/':
                     in_states = True
                     continue
+                if name:
+                    raise ValueError(
+                        'Illegal token in charstatelabel: "{}{}"'.format(name, w))
                 name = w
             except StopIteration:
                 break
         if num:
+            if name and name in names:
+                duplicate_charlabel(name, 'CHARSTATELABELS', nexus)
             self.characters.append(types.SimpleNamespace(number=num, name=name, states=states))
         elif comma:  # There was a comma, but no new label.
             warnings.warn('Trailing comma in CHARSTATELABELS command')

src/commonnexus/tokenizer.py

@@ -11,7 +11,7 @@
 
     The following punctuation marks have special properties: [ ] do not break a
     word; + and - are allowed as state symbols, but none of the rest are allowed; - is
-    considered punctuation except were it is the minus sign in a negative number.
+    considered punctuation except where it is the minus sign in a negative number.
 """
 import enum
 import itertools

tests/test_blocks_characters.py

@@ -267,3 +267,56 @@ def test_Data_with_mixed_charlabels(nexus):
     nex = nexus(DATA="DIMENSIONS NCHAR=2; CHARSTATELABELS 1 x, 2 ; MATRIX t1 1 1;")
     m = nex.characters.get_matrix()
     assert '2' in m['t1'], 'unspecified character label'
+
+
+def test_illegal_charstatelabel():
+    nex = Nexus("""\
+#NEXUS
+
+BEGIN DATA;
+    DIMENSIONS NTAX=3 NCHAR=3;
+    FORMAT DATATYPE=STANDARD MISSING=? GAP=- SYMBOLS="01";
+    CHARSTATELABELS
+        1 hand_1,
+        2 burn(tr.)_3,
+        3 claw(nail)_3
+        ;
+MATRIX
+A   001
+B   000
+C   000
+;
+END;""")
+    with pytest.raises(ValueError) as e:
+        _ = nex.DATA.CHARSTATELABELS
+    assert 'burn(' in str(e)
+
+    with pytest.raises(ValueError):
+        nex.validate()
+
+
+def test_duplicate_charstatelabels():
+    nex = Nexus("""\
+#NEXUS
+
+BEGIN DATA;
+    DIMENSIONS NTAX=3 NCHAR=3;
+    FORMAT DATATYPE=STANDARD MISSING=? GAP=- SYMBOLS="01";
+    CHARSTATELABELS
+        1 hand_1,
+        2 _3,
+        3 _3
+        ;
+MATRIX
+A   001
+B   000
+C   000
+;
+END;""")
+    with warnings.catch_warnings(record=True) as w:
+        _ = nex.DATA.CHARSTATELABELS
+        assert len(w) == 1, 'Expected 1 warning, got %r' % w
+
+    nex.cfg.strict = True
+    with pytest.raises(ValueError):
+        _ = nex.DATA.CHARSTATELABELS