Fix bad start or end symbol error (#167)

* Fix ValueError error message * Better test to for ValueError message * Update the changelog * Raw format string to appease flake8
allenai · Nov 16, 2020 · 236034f · 236034f
1 parent 8a305a0
commit 236034f
Show file tree

Hide file tree

Showing 3 changed files with 3 additions and 2 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -15,6 +15,7 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Fixed
 
 - Fixed `pair-classification-esim` pretrained model.
+- Fixed `ValueError` error message in `Seq2SeqDatasetReader`.
 
 ### Added
 

diff --git a/allennlp_models/generation/dataset_readers/seq2seq.py b/allennlp_models/generation/dataset_readers/seq2seq.py
@@ -97,7 +97,7 @@ def __init__(
                 )
             except ValueError:
                 raise ValueError(
-                    f"Bad start or end symbol ({'start_symbol', 'end_symbol'}) "
+                    f"Bad start or end symbol ('{start_symbol}', '{end_symbol}') "
                     f"for tokenizer {self._source_tokenizer}"
                 )
 

diff --git a/tests/generation/dataset_readers/seq2seq_test.py b/tests/generation/dataset_readers/seq2seq_test.py
@@ -218,5 +218,5 @@ def test_correct_quote_handling(self, line):
             ]
 
     def test_bad_start_or_end_symbol(self):
-        with pytest.raises(ValueError, match="Bad start or end symbol"):
+        with pytest.raises(ValueError, match=r"Bad start or end symbol \('BAD SYMBOL"):
             Seq2SeqDatasetReader(start_symbol="BAD SYMBOL")