added support for xz and bzip2 compression

frictionless/platform.py

@@ -41,6 +41,12 @@ class Platform:
 
     # Core
 
+    @cached_property
+    def bz2(self):
+        import bz2
+
+        return bz2
+
     @cached_property
     def chardet(self):
         import chardet
@@ -149,6 +155,12 @@ def jsonschema_validators(self):
 
         return jsonschema.validators
 
+    @cached_property
+    def lzma(self):
+        import lzma
+
+        return lzma
+
     @cached_property
     def marko(self):
         import marko

frictionless/settings.py

@@ -13,7 +13,7 @@
 NAME_PATTERN = "^([-a-z0-9._/])+$"
 TYPE_PATTERN = "^([-a-z/])+$"
 PACKAGE_PATH = "datapackage.json"
-COMPRESSION_FORMATS = ["zip", "gz"]
+COMPRESSION_FORMATS = ["zip", "gz", "bz2", "xz"]
 
 # Defaults
 

frictionless/system/loader.py

@@ -231,6 +231,28 @@ def read_byte_stream_decompress(
             byte_stream = platform.gzip.open(byte_stream)  # type: ignore
             return byte_stream
 
+        # bzip2 compression
+        if self.resource.compression == "bz2":
+            # Stats
+            if not self.remote:
+                bytes = True
+                while bytes:
+                    bytes = byte_stream.read1(io.DEFAULT_BUFFER_SIZE)  # type: ignore
+                byte_stream.seek(0)
+            byte_stream = platform.bz2.open(byte_stream)  # type: ignore
+            return byte_stream
+
+        # XZ compression
+        if self.resource.compression == "xz":
+            # Stats
+            if not self.remote:
+                bytes = True
+                while bytes:
+                    bytes = byte_stream.read1(io.DEFAULT_BUFFER_SIZE)  # type: ignore
+                byte_stream.seek(0)
+            byte_stream = platform.lzma.open(byte_stream)  # type: ignore
+            return byte_stream
+
         # Not supported compression
         note = f'compression "{self.resource.compression}" is not supported'
         raise FrictionlessException(errors.CompressionError(note=note))

tests/resources/table/test_compression.py

@@ -63,6 +63,28 @@ def test_resource_compression_local_csv_gz():
         ]
 
 
+def test_resource_compression_local_csv_xz():
+    with TableResource(path="data/table.csv.xz") as resource:
+        assert resource.compression == "xz"
+        assert resource.innerpath is None
+        assert resource.header == ["id", "name"]
+        assert resource.read_rows() == [
+            {"id": 1, "name": "english"},
+            {"id": 2, "name": "中国人"},
+        ]
+
+
+def test_resource_compression_local_csv_bz2():
+    with TableResource(path="data/table.csv.bz2") as resource:
+        assert resource.compression == "bz2"
+        assert resource.innerpath is None
+        assert resource.header == ["id", "name"]
+        assert resource.read_rows() == [
+            {"id": 1, "name": "english"},
+            {"id": 2, "name": "中国人"},
+        ]
+
+
 def test_resource_compression_stream_csv_zip():
     with open("data/table.csv.zip", "rb") as file:
         with TableResource(data=file, format="csv", compression="zip") as resource:
@@ -83,6 +105,26 @@ def test_resource_compression_stream_csv_gz():
             ]
 
 
+def test_resource_compression_stream_csv_xz():
+    with open("data/table.csv.xz", "rb") as file:
+        with TableResource(data=file, format="csv", compression="xz") as resource:
+            assert resource.header == ["id", "name"]
+            assert resource.read_rows() == [
+                {"id": 1, "name": "english"},
+                {"id": 2, "name": "中国人"},
+            ]
+
+
+def test_resource_compression_stream_csv_bz2():
+    with open("data/table.csv.bz2", "rb") as file:
+        with TableResource(data=file, format="csv", compression="bz2") as resource:
+            assert resource.header == ["id", "name"]
+            assert resource.read_rows() == [
+                {"id": 1, "name": "english"},
+                {"id": 2, "name": "中国人"},
+            ]
+
+
 @pytest.mark.vcr
 def test_resource_compression_remote_csv_zip():
     source = "https://raw.githubusercontent.com/frictionlessdata/tabulator-py/master/data/table.csv.zip"