Use new property in parallel_text_corpus

Author: Enkidu93

machine/corpora/n_parallel_text_corpus.py

@@ -79,7 +79,7 @@ class NParallelTextCorpus(NParallelTextCorpusBase):
     def __init__(self, corpora: Sequence[TextCorpus], row_ref_comparer: Optional[Callable[[Any, Any], int]] = None):
         self._corpora = corpora
         self._row_ref_comparer = row_ref_comparer if row_ref_comparer is not None else default_row_ref_comparer
-        self._all_rows: Sequence[bool] = [False for _ in range(len(corpora))]
+        self._all_rows: List[bool] = [False for _ in range(len(corpora))]
 
     def is_tokenized(self, i: int) -> bool:
         return self.corpora[i].is_tokenized
@@ -97,7 +97,7 @@ def row_ref_comparer(self) -> Callable[[Any, Any], int]:
         return self._row_ref_comparer
 
     @property
-    def all_rows(self) -> Sequence[bool]:
+    def all_rows(self) -> List[bool]:
         return self._all_rows
 
     def _get_text_ids_from_corpora(self) -> Set[str]:

machine/corpora/standard_parallel_text_corpus.py

@@ -27,7 +27,8 @@ def __init__(
         self._all_source_rows = all_source_rows
         self._all_target_rows = all_target_rows
         self._n_parallel_text_corpus = NParallelTextCorpus([source_corpus, target_corpus])
-        self._n_parallel_text_corpus._all_rows = [self._all_source_rows, self.all_target_rows]
+        self._n_parallel_text_corpus.all_rows[0] = self.all_source_rows
+        self._n_parallel_text_corpus.all_rows[1] = self.all_target_rows
         self._row_ref_comparer = row_ref_comparer or default_row_ref_comparer
 
     @property