Better str representation (#27)

* Prevent very long outputs when transcription evaluators are printed

* Change str to repr

python/stringalign/evaluation.py

@@ -196,9 +196,11 @@ def from_strings(
             tokenizer=tokenizer,
         )
 
-    def __str__(self) -> str:
+    def __repr__(self) -> str:
         return f"LineError('{self.reference}', '{self.predicted}', metadata={self.metadata})"
 
+    __str__ = __repr__
+
 
 @dataclass(frozen=True, slots=True)
 class TranscriptionEvaluator:
@@ -278,3 +280,11 @@ def from_strings(
             predictions=predictions,
             line_errors=line_errors,
         )
+
+    def __len__(self) -> int:
+        return len(self.line_errors)
+
+    def __repr__(self) -> str:
+        return f"TranscriptionEvaluator(len={len(self)})"
+
+    __str__ = __repr__