Merge pull request #30 from BristolComputing/kreczko-issue-5-p2

Fixes issue #5

.pre-commit-config.yaml

@@ -58,6 +58,7 @@ repos:
         args: []
         additional_dependencies:
           - codetiming
+          - crc32c
           - fasthep_logging
           - pyhdfs
           - pytest

pyproject.toml

@@ -33,6 +33,7 @@ dynamic = ["version"]
 
 dependencies = [
     "codetiming",
+    "crc32c",
     "fasthep-logging >= 0.1.1",
     "typer >= 0.4.0",
     "requests",

src/xrdsum/checksums/__init__.py

@@ -1,10 +1,13 @@
 """xrdsum.checksums package"""
+
 from __future__ import annotations
 
 from ._base import Checksum
+from ._crc32c import CRC32C
 from .adler32 import Adler32
 
 AVAILABLE_CHECKSUM_TYPES = {
     "adler32": Adler32,
+    "crc32c": CRC32C,
 }
-__all__ = ["Adler32", "Checksum", "AVAILABLE_CHECKSUM_TYPES"]
+__all__ = ["AVAILABLE_CHECKSUM_TYPES", "CRC32C", "Adler32", "Checksum"]

src/xrdsum/checksums/_base.py

@@ -1,26 +1,21 @@
 """Definition of the Checksum protocol."""
+
 from __future__ import annotations
 
+from abc import ABC, abstractmethod
 from collections.abc import Iterable
-from typing import Any, Protocol
+from typing import Any
 
 
-class Checksum(Protocol):
+class Checksum(ABC):
     """Base protocol for checksum implementations."""
 
     name: str = "Unknown"
     value: str = "N/A"
     bytes_read: int = 0
     number_of_buffers_read: int = 0
 
-    def int_to_hex(self, value: int) -> str:
-        """Converts integer to hex representation"""
-        raise NotImplementedError()
-
-    def hex_to_int(self, value: str) -> int:
-        """Converts hex representation to integer"""
-        raise NotImplementedError()
-
+    @abstractmethod
     def calculate(self, file_buffer: Iterable[Any]) -> str:
         """Calculates the checksum"""
         raise NotImplementedError()

src/xrdsum/checksums/_crc32c.py

@@ -0,0 +1,37 @@
+from __future__ import annotations
+
+from collections.abc import Iterable
+from typing import Any
+
+import crc32c
+
+from ..logger import APP_LOGGER_NAME, get_logger
+from ._base import Checksum
+
+log = get_logger(APP_LOGGER_NAME)
+
+
+class CRC32C(Checksum):
+    name: str = "crc32c"
+
+    def calculate(self, file_buffer: Iterable[Any]) -> str:
+        value = crc32c.CRC32CHash()
+        bytes_read = 0
+        number_of_buffers_read = 0
+        for buffer in file_buffer:
+            value.update(buffer)
+            bytes_read += len(buffer)
+            number_of_buffers_read += 1
+            log.trace(
+                "%s: %s %s %s",
+                self.name,
+                value.hexdigest(),
+                len(buffer),
+                bytes_read,
+            )
+
+        self.value = value.hexdigest()
+        self.bytes_read = bytes_read
+        self.number_of_buffers_read = number_of_buffers_read
+
+        return self.value