cfinterface 1.10.3__tar.gz → 1.10.4__tar.gz

{cfinterface-1.10.3 → cfinterface-1.10.4}/CHANGELOG.md

@@ -7,6 +7,12 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+## [1.10.4] - 2026-11-14
+
+### Fixed
+
+- `SectionReading.read()` and `BlockReading.read()` now reset their internal data before each call, as `RegisterReading.read()` does since 1.10.0; a Latin-1 file whose first non-ASCII byte comes after the first 8 KiB used to return the sections or blocks read before the `UnicodeDecodeError` twice
+
 ## [1.10.3] - 2026-10-30
 
 ### Fixed

{cfinterface-1.10.3 → cfinterface-1.10.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.5
 Name: cfinterface
-Version: 1.10.3
+Version: 1.10.4
 Summary: Interface for handling custom formatted files
 Project-URL: Documentation, https://rjmalves.github.io/cfinterface/
 Project-URL: Repository, https://github.com/rjmalves/cfinterface/

{cfinterface-1.10.3 → cfinterface-1.10.4}/cfinterface/__init__.py

@@ -6,7 +6,7 @@ cfi is a Python module for handling custom formatted files
 and provide reading, storing and writing utilities.
 """
 
-__version__ = "1.10.3"
+__version__ = "1.10.4"
 
 from . import components  # noqa
 from . import data  # noqa

{cfinterface-1.10.3 → cfinterface-1.10.4}/cfinterface/reading/blockreading.py

@@ -78,6 +78,7 @@ class BlockReading:
         :return: The data from the blocks found in the file
         :rtype: BlockData
         """
+        self.__data = BlockData(DefaultBlock(data=""))
         self.__repository = factory(self.__storage)(
             content, not isfile(content), encoding
         )

{cfinterface-1.10.3 → cfinterface-1.10.4}/cfinterface/reading/sectionreading.py

@@ -75,6 +75,7 @@ class SectionReading:
         :return: The data from the sections found in the file
         :rtype: SectionData
         """
+        self.__data = SectionData(DefaultSection(data=""))
         self.__repository = factory(self.__storage)(
             content, not isfile(content), encoding
         )

cfinterface-1.10.4/tests/files/test_encoding_fallback.py

@@ -0,0 +1,49 @@
+"""The fallback encoding retry must not keep what the failed attempt read."""
+
+from pathlib import Path
+
+import pytest
+
+from cfinterface.files.blockfile import BlockFile
+from cfinterface.files.registerfile import RegisterFile
+from cfinterface.files.sectionfile import SectionFile
+
+NUM_LINES = 100
+
+
+class _Blocks(BlockFile):
+    BLOCKS = []
+    ENCODING = ["utf-8", "latin-1", "ascii"]
+
+
+class _Registers(RegisterFile):
+    REGISTERS = []
+    ENCODING = ["utf-8", "latin-1", "ascii"]
+
+
+class _Sections(SectionFile):
+    SECTIONS = []
+    ENCODING = ["utf-8", "latin-1", "ascii"]
+
+
+def _latin1_file(tmp_path: Path) -> Path:
+    lines = [f"{f'LINE {i}':<500}\n" for i in range(NUM_LINES - 1)]
+    lines.append("1º FCF / 2º SF\n")
+    content = "".join(lines).encode("latin-1")
+    # Past the text decoder's first chunk, so UTF-8 fails mid-file.
+    assert content.index("º".encode("latin-1")) > 8192
+    path = tmp_path / "file.dat"
+    path.write_bytes(content)
+    return path
+
+
+@pytest.mark.parametrize("file_type", [_Blocks, _Registers, _Sections])
+def test_latin1_file_is_read_once(tmp_path: Path, file_type: type) -> None:
+    path = _latin1_file(tmp_path)
+
+    data = file_type.read(str(path)).data
+
+    lines = [str(component.data) for component in data][1:]
+    assert len(lines) == NUM_LINES
+    assert lines[0].startswith("LINE 0")
+    assert lines[-1].startswith("1º FCF")