AVRO-2921: Type Fixes for avro.io (apache#1264)

lang/py/avro/datafile.py

@@ -25,13 +25,13 @@
 import io
 import json
 from types import TracebackType
-from typing import BinaryIO, MutableMapping, Optional, Type
+from typing import BinaryIO, MutableMapping, Optional, Type, cast
 
 import avro.codecs
 import avro.errors
 import avro.io
 import avro.schema
-import avro.utils
+from avro.utils import TypedDict, randbytes
 
 VERSION = 1
 MAGIC = bytes(b"Obj" + bytearray([VERSION]))
@@ -60,6 +60,12 @@
 SCHEMA_KEY = "avro.schema"
 
 
+class HeaderType(TypedDict):
+    magic: bytes
+    meta: MutableMapping[str, bytes]
+    sync: bytes
+
+
 class _DataFileMetadata:
     """
     Mixin for meta properties.
@@ -183,7 +189,7 @@ def __init__(
             writer.seek(0, 2)
             self._header_written = True
             return
-        self.sync_marker = avro.utils.randbytes(16)
+        self.sync_marker = randbytes(16)
         self.codec = codec
         self.schema = str(writers_schema)
         self.datum_writer.writers_schema = writers_schema
@@ -358,16 +364,10 @@ def _read_header(self) -> None:
         self.reader.seek(0, 0)
 
         # read header into a dict
-        header = self.datum_reader.read_data(META_SCHEMA, META_SCHEMA, self.raw_decoder)
-
-        # check magic number
+        header = cast(HeaderType, self.datum_reader.read_data(META_SCHEMA, META_SCHEMA, self.raw_decoder))
         if header.get("magic") != MAGIC:
             raise avro.errors.AvroException(f"Not an Avro data file: {header.get('magic')!r} doesn't match {MAGIC!r}.")
-
-        # set metadata
         self._meta = header["meta"]
-
-        # set sync marker
         self.sync_marker = header["sync"]
 
     def _read_block_header(self) -> None:
@@ -393,6 +393,8 @@ def __next__(self) -> object:
                 raise StopIteration
             self._read_block_header()
 
+        if self.datum_decoder is None:
+            raise avro.errors.DataFileException("DataFile is not ready to read because it has no decoder")
         datum = self.datum_reader.read(self.datum_decoder)
         self.block_count -= 1
         return datum

lang/py/avro/errors.py

@@ -56,7 +56,11 @@ def __init__(self, *args):
             expected_schema, datum = args[:2]
         except (IndexError, ValueError):
             return super().__init__(*args)
-        return super().__init__(f"The datum {datum} is not an example of the schema {_safe_pretty(expected_schema)}")
+        return super().__init__(f"The datum {datum} of the type {type(datum)} is not an example of the schema {_safe_pretty(expected_schema)}")
+
+
+class InvalidDefaultException(AvroTypeException):
+    """Raised when a default value isn't a suitable type for the schema."""
 
 
 class AvroOutOfScaleException(AvroTypeException):
@@ -81,6 +85,10 @@ class DataFileException(AvroException):
     """Raised when there's a problem reading or writing file object containers."""
 
 
+class IONotReadyException(AvroException):
+    """Raised when attempting an avro operation on an io object that isn't fully initialized."""
+
+
 class AvroRemoteException(AvroException):
     """Raised when an error message is sent by an Avro requestor or responder."""