Support for UD_OLD_FRENCH added

flair/datasets/__init__.py

@@ -118,6 +118,7 @@
 from .treebanks import UD_NORTH_SAMI
 from .treebanks import UD_MALTESE
 from .treebanks import UD_AFRIKAANS
+from .treebanks import UD_OLD_FRENCH
 from .treebanks import UD_GOTHIC
 
 # Expose all text-text datasets

flair/datasets/treebanks.py

@@ -1226,6 +1226,7 @@ def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True):
 
         super(UD_NORTH_SAMI, self).__init__(data_folder, in_memory=in_memory)
 
+
 class UD_MALTESE(UniversalDependenciesCorpus):
     def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True):
 
@@ -1252,6 +1253,7 @@ def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True):
 
         super(UD_MALTESE, self).__init__(data_folder, in_memory=in_memory)
 
+
 class UD_AFRIKAANS(UniversalDependenciesCorpus):
     def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True):
 
@@ -1277,8 +1279,9 @@ def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True):
         )
 
         super(UD_AFRIKAANS, self).__init__(data_folder, in_memory=in_memory)
+
 
-class UD_GOTHIC(UniversalDependenciesCorpus):
+class UD_GOTHIC(UniversalDependenciesCorpus):  
     def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True):
 
         if type(base_path) == str:
@@ -1303,3 +1306,30 @@ def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True):
         )
 
         super(UD_GOTHIC, self).__init__(data_folder, in_memory=in_memory)
+
+
+class UD_OLD_FRENCH(UniversalDependenciesCorpus):
+    def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True):
+
+        if type(base_path) == str:
+            base_path: Path = Path(base_path)
+
+        # this dataset name
+        dataset_name = self.__class__.__name__.lower()
+
+        # default dataset folder is the cache root
+        if not base_path:
+            base_path = Path(flair.cache_root) / "datasets"
+        data_folder = base_path / dataset_name
+
+        # download data if necessary
+        web_path = "https://github.com/UniversalDependencies/UD_Old_French-SRCMF/tree/master"
+        cached_path(f"{web_path}/fro_srcmf-ud-dev.conllu", Path("datasets") / dataset_name)
+        cached_path(
+            f"{web_path}/fro_srcmf-ud-test.conllu", Path("datasets") / dataset_name
+        )
+        cached_path(
+            f"{web_path}/fro_srcmf-ud-train.conllu", Path("datasets") / dataset_name
+        )
+
+        super(UD_OLD_FRENCH, self).__init__(data_folder, in_memory=in_memory)