Merge branch 'master' into feature/UD_Belarusian

flair/datasets/__init__.py

@@ -136,6 +136,8 @@
 from .treebanks import UD_GOTHIC
 from .treebanks import UD_WOLOF
 from .treebanks import UD_BELARUSIAN
+from .treebanks import UD_OLD_CHURCH_SLAVONIC
+from .treebanks import UD_COPTIC
 
 # Expose all text-text datasets
 from .text_text import ParallelTextCorpus

flair/datasets/treebanks.py

@@ -251,6 +251,32 @@ def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True, s
 
         super(UD_ENGLISH, self).__init__(data_folder, in_memory=in_memory, split_multiwords=split_multiwords)
 
+class UD_OLD_CHURCH_SLAVONIC(UniversalDependenciesCorpus):
+    def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True, split_multiwords: 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://raw.githubusercontent.com/UniversalDependencies/UD_Old_Church_Slavonic-PROIEL/master"
+        cached_path(f"{web_path}/cu_proiel-ud-dev.conllu", Path("datasets") / dataset_name)
+        cached_path(
+            f"{web_path}/cu_proiel-ud-test.conllu", Path("datasets") / dataset_name
+        )
+        cached_path(
+            f"{web_path}/cu_proiel-ud-train.conllu", Path("datasets") / dataset_name
+        )
+
+        super(UD_OLD_CHURCH_SLAVONIC, self).__init__(data_folder, in_memory=in_memory, split_multiwords=split_multiwords)
+
 
 class UD_ESTONIAN(UniversalDependenciesCorpus):
     def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True, split_multiwords: bool = True):
@@ -1485,7 +1511,7 @@ def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True, s
 
 
 class UD_BELARUSIAN(UniversalDependenciesCorpus):
-    def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True, split_multiwords: bool = True):
+      def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True, split_multiwords: bool = True):
 
         if type(base_path) == str:
             base_path: Path = Path(base_path)
@@ -1508,4 +1534,31 @@ def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True, s
             f"{web_path}/be_hse-ud-train.conllu", Path("datasets") / dataset_name
         )
 
-        super(UD_BELARUSIAN, self).__init__(data_folder, in_memory=in_memory, split_multiwords=split_multiwords)
+        super(UD_BELARUSIAN, self).__init__(data_folder, in_memory=in_memory, split_multiwords=split_multiwords)
+
+
+class UD_COPTIC(UniversalDependenciesCorpus):
+    def __init__(self, base_path: Union[str, Path] = None, in_memory: bool = True, split_multiwords: 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://raw.githubusercontent.com/UniversalDependencies/UD_Coptic-Scriptorium/master"
+        cached_path(f"{web_path}/cop_scriptorium-ud-dev.conllu", Path("datasets") / dataset_name)
+        cached_path(
+            f"{web_path}/cop_scriptorium-ud-test.conllu", Path("datasets") / dataset_name
+        )
+        cached_path(
+            f"{web_path}/cop_scriptorium-ud-train.conllu", Path("datasets") / dataset_name
+        )
+
+        super(UD_COPTIC, self).__init__(data_folder, in_memory=in_memory, split_multiwords=split_multiwords)