Release version 1.8.0

This setting was added in flake8 3.8.0 and allows adding entries to the
exclude list without also removing the default entries.

.editorconfig

@@ -8,3 +8,7 @@ insert_final_newline = true
 [*.rst]
 indent_style = space
 indent_size = 4
+
+[*.yml]
+indent_style = space
+indent_size = 2

.github/workflows/ci.yml

@@ -0,0 +1,20 @@
+on: [pull_request, push]
+jobs:
+  test:
+    strategy:
+      matrix:
+        platform: [macos-latest, ubuntu-latest, windows-latest]
+    runs-on: ${{ matrix.platform }}
+    steps:
+      - uses: actions/checkout@v2
+      - uses: actions/setup-python@v1
+        with:
+          python-version: "3.6"
+      - uses: actions/setup-python@v1
+        with:
+          python-version: "3.7"
+      - uses: actions/setup-python@v1
+        with:
+          python-version: "3.8"
+      - run: python -m pip install --upgrade tox
+      - run: tox

.gitignore

@@ -2,7 +2,13 @@
 *.py[co]
 __pycache__/
 
+# tox
+.tox/
+
 # setuptools
 *.egg-info/
 build/
 dist/
+
+# mypy
+.mypy_cache/

MANIFEST.in

@@ -0,0 +1,6 @@
+# Note: See the PyPA documentation for a list of file names that are included/excluded by default:
+# https://packaging.python.org/guides/using-manifest-in/#how-files-are-included-in-an-sdist
+# Please only add entries here for files that are *not* already handled by default.
+
+recursive-include tests *.py
+recursive-include tests/data *.rsrc

README.rst

@@ -56,7 +56,7 @@ Simple example
     >>> rf
     <rsrcfork.ResourceFile at 0x1046e6048, attributes ResourceFileAttrs.0, containing 4 resource types: [b'utxt', b'utf8', b'TEXT', b'drag']>
     >>> rf[b"TEXT"]
-    <rsrcfork.ResourceFile._LazyResourceMap at 0x10470ed30 containing one resource: rsrcfork.Resource(resource_type=b'TEXT', resource_id=256, name=None, attributes=ResourceAttrs.0, data=b'Here is some text')>
+    <rsrcfork.ResourceFile._LazyResourceMap at 0x10470ed30 containing one resource: rsrcfork.Resource(type=b'TEXT', id=256, name=None, attributes=ResourceAttrs.0, data=b'Here is some text')>
 
 Automatic selection of data/resource fork
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
@@ -108,25 +108,88 @@ Writing resource data is not supported at all.
 Further info on resource files
 ------------------------------
 
-Sources of information about the resource fork data format, and the structure of common resource types:
-
-* Inside Macintosh, Volume I, Chapter 5 "The Resource Manager". This book can probably be obtained in physical form somewhere, but the relevant chapter/book is also available in a few places online:
-
-  * `Apple's legacy documentation <https://developer.apple.com/legacy/library/documentation/mac/pdf/MoreMacintoshToolbox.pdf>`_
-  * pagetable.com, a site that happened to have a copy of the book: `info blog post <http://www.pagetable.com/?p=50>`_, `direct download <http://www.weihenstephan.org/~michaste/pagetable/mac/Inside_Macintosh.pdf>`_
-
-* `Wikipedia <https://en.wikipedia.org/wiki/Resource_fork>`_, of course
-* The `Resource Fork <http://fileformats.archiveteam.org/wiki/Resource_Fork>`_ article on "Just Solve the File Format Problem" (despite the title, this is a decent site and not clickbait)
-* The `KSFL <https://github.com/kreativekorp/ksfl>`_ library (and `its wiki <https://github.com/kreativekorp/ksfl/wiki/Macintosh-Resource-File-Format>`_), written in Java, which supports reading and writing resource files
-* Alysis Software Corporation's article on resource compression (found on `the company's website <http://www.alysis.us/arctechnology.htm>`_ and in `MacTech Magazine's online archive <http://preserve.mactech.com/articles/mactech/Vol.09/09.01/ResCompression/index.html>`_) has some information on the structure of certain kinds of compressed resources.
-* Apple's macOS SDK, which is distributed with Xcode. The latest version of Xcode is available for free from the Mac App Store. Current and previous versions can be downloaded from `the Apple Developer download page <https://developer.apple.com/download/more/>`_. Accessing these downloads requires an Apple ID with (at least) a free developer program membership.
-* Apple's MPW (Macintosh Programmer's Workshop) and related developer tools. These were previously available from Apple's FTP server at ftp://ftp.apple.com/, which is no longer functional. Because of this, these downloads are only available on mirror sites, such as http://staticky.com/mirrors/ftp.apple.com/.
-
-If these links are no longer functional, some are archived in the `Internet Archive Wayback Machine <https://archive.org/web/>`_ or `archive.is <http://archive.is/>`_ aka `archive.fo <https://archive.fo/>`_.
+For technical info and documentation about resource files and resources, see the `"resource forks" section of the mac_file_format_docs repo <https://github.com/dgelessus/mac_file_format_docs/blob/master/README.md#resource-forks>`_.
 
 Changelog
 ---------
 
+Version 1.8.0
+^^^^^^^^^^^^^
+
+* Removed the old (non-subcommand-based) CLI syntax.
+* Added filtering support to the ``list`` subcommand.
+* Added a ``resource-info`` subcommand to display technical information about resources (more detailed than what is displayed by ``list`` and ``read``).
+* Added a ``raw-compress-info`` subcommand to display technical header information about standalone compressed resource data.
+* Made the library PEP 561-compliant by adding a py.typed file.
+* Fixed an incorrect ``AssertionError`` when using the ``--no-decompress`` command-line options.
+
+Version 1.7.0
+^^^^^^^^^^^^^
+
+* Added a ``raw-decompress`` subcommand to decompress compressed resource data stored in a standalone file rather than as a resource.
+* Optimized lazy loading of ``Resource`` objects. Previously, resource data would be read from disk whenever a ``Resource`` object was looked up, even if the data itself is never used. Now the resource data is only loaded once the ``data`` (or ``data_raw``) attribute is accessed.
+
+  * The same optimization applies to the ``name`` attribute, although this is unlikely to make a difference in practice.
+  * As a result, it is no longer possible to construct ``Resource`` objects without a resource file. This was previously possible, but had no practical use.
+* Fixed a small error in the ``'dcmp' (0)`` decompression implementation.
+
+Version 1.6.0
+^^^^^^^^^^^^^
+
+* Added a new subcommand-based command-line syntax to the ``rsrcfork`` tool, similar to other CLI tools such as ``git`` or ``diskutil``.
+
+  * This subcommand-based syntax is meant to replace the old CLI options, as the subcommand structure is easier to understand and more extensible in the future.
+  * Currently there are three subcommands: ``list`` to list resources in a file, ``read`` to read/display resource data, and ``read-header`` to read a resource file's header data. These subcommands can be used to perform all operations that were also available with the old CLI syntax.
+  * The old CLI syntax is still supported for now, but it will be removed soon.
+  * The new syntax no longer supports reading CLI arguments from a file (using ``@args_file.txt``), abbreviating long options (e. g. ``--no-d`` instead of ``--no-decompress``), or the short option ``-f`` instead of ``--fork``. If you have a need for any of these features, please open an issue.
+
+Version 1.5.0
+^^^^^^^^^^^^^
+
+* Added stream-based decompression methods to the ``rsrcfork.compress`` module.
+
+  * The internal decompressor implementations have been refactored to use streams.
+  * This allows for incremental decompression of compressed resource data. In practice this has no noticeable effect yet, because the main ``rsrcfork`` API doesn't support incremental reading of resource data.
+
+* Fixed the command line tool always displaying an incorrect error "Cannot specify an explicit fork when reading from stdin" when using ``-`` (stdin) as the input file.
+
+Version 1.4.0
+^^^^^^^^^^^^^
+
+* Added ``length`` and ``length_raw`` attributes to ``Resource``. These attributes are equivalent to the ``len`` of ``data`` and ``data_raw`` respectively, but may be faster to access.
+
+  * Currently, the only optimized case is ``length`` for compressed resources, but more optimizations may be added in the future.
+
+* Added a ``compressed_info`` attribute to ``Resource`` that provides access to the header information of compressed resources.
+* Improved handling of compressed resources when listing resource files with the command line tool.
+
+  * Metadata of compressed resources is now displayed even if no decompressor implementation is available (as long as the compressed data header can be parsed).
+  * Performance has been improved - the data no longer needs to be fully decompressed to get its length, this information is now read from the header.
+  * The ``'dcmp'`` ID used to decompress each resource is displayed.
+
+* Fixed an incorrect ``options.packages`` in ``setup.cfg``, which made the library unusable except when installing from source using ``--editable``.
+* Fixed ``ResourceFile.__enter__`` returning ``None``, which made it impossible to use ``ResourceFile`` properly in a ``with`` statement.
+* Fixed various minor errors reported by type checking with ``mypy``.
+
+Version 1.3.0.post1
+^^^^^^^^^^^^^^^^^^^
+
+* Fixed an incorrect ``options.packages`` in ``setup.cfg``, which made the library unusable except when installing from source using ``--editable``.
+
+Version 1.2.0.post1
+^^^^^^^^^^^^^^^^^^^
+
+* Fixed an incorrect ``options.packages`` in ``setup.cfg``, which made the library unusable except when installing from source using ``--editable``.
+
+Version 1.3.0
+^^^^^^^^^^^^^
+
+* Added a ``--group`` command line option to group resources in list format by type (the default), ID, or with no grouping.
+* Added a ``dump-text`` output format to the command line tool. This format is identical to ``dump``, but instead of a hex dump, it outputs the resource data as text. The data is decoded as MacRoman and classic Mac newlines (``\r``) are translated. This is useful for examining resources that contain mostly plain text.
+* Changed the command line tool to sort resources by type and ID, and added a ``--no-sort`` option to disable sorting and output resources in file order (which was the previous behavior).
+* Renamed the ``rsrcfork.Resource`` attributes ``resource_type`` and ``resource_id`` to ``type`` and ``id``, respectively. The old names have been deprecated and will be removed in the future, but are still supported for now.
+* Changed ``--format=dump`` output to match ``hexdump -C``'s format - spacing has been adjusted, and multiple subsequent identical lines are collapsed into a single ``*``.
+
 Version 1.2.0
 ^^^^^^^^^^^^^
 

pyproject.toml

@@ -0,0 +1,6 @@
+[build-system]
+requires = [
+	"setuptools >= 46.4.0",
+	"wheel >= 0.32.0",
+]
+build-backend = "setuptools.build_meta"

rsrcfork/__init__.py

@@ -1,6 +1,26 @@
 """A pure Python, cross-platform library/tool for reading Macintosh resource data, as stored in resource forks and ``.rsrc`` files."""
 
-__version__ = "1.2.0"
+# To release a new version:
+# * Remove the .dev suffix from the version number in this file.
+# * Update the changelog in the README.rst (rename the "next version" section to the correct version number).
+# * Remove the ``dist`` directory (if it exists) to clean up any old release files.
+# * Run ``python3 setup.py sdist bdist_wheel`` to build the release files.
+# * Run ``python3 -m twine check dist/*`` to check the release files.
+# * Fix any errors reported by the build and/or check steps.
+# * Commit the changes to master.
+# * Tag the release commit with the version number, prefixed with a "v" (e. g. version 1.2.3 is tagged as v1.2.3).
+# * Fast-forward the release branch to the new release commit.
+# * Push the master and release branches.
+# * Upload the release files to PyPI using ``python3 -m twine upload dist/*``.
+# * On the GitHub repo's Releases page, edit the new release tag and add the relevant changelog section from the README.rst. (Note: The README is in reStructuredText format, but GitHub's release notes use Markdown, so it may be necessary to adjust the markup syntax.)
+
+# After releasing:
+# * (optional) Remove the build and dist directories from the previous release as they are no longer needed.
+# * Bump the version number in this file to the next version and add a .dev suffix.
+# * Add a new empty section for the next version to the README.rst changelog.
+# * Commit and push the changes to master.
+
+__version__ = "1.8.0"
 
 __all__ = [
 	"Resource",
@@ -11,8 +31,8 @@ __all__ = [
 	"open",
 ]
 
-from . import api, compress
 from .api import Resource, ResourceAttrs, ResourceFile, ResourceFileAttrs
+from . import compress
 
 # noinspection PyShadowingBuiltins
 open = ResourceFile.open

rsrcfork/api.py

@@ -4,6 +4,7 @@ import enum
 import io
 import os
 import struct
+import types
 import typing
 import warnings
 
@@ -58,9 +59,11 @@ STRUCT_RESOURCE_REFERENCE = struct.Struct(">hHI4x")
 # 1 byte: Length of following resource name.
 STRUCT_RESOURCE_NAME_HEADER = struct.Struct(">B")
 
+
 class InvalidResourceFileError(Exception):
 	pass
 
+
 class ResourceFileAttrs(enum.Flag):
 	"""Resource file attribute flags. The descriptions for these flags are taken from comments on the map*Bit and map* enum constants in <CarbonCore/Resources.h>."""
 	
@@ -81,6 +84,7 @@ class ResourceFileAttrs(enum.Flag):
 	_BIT_1 = 1 << 1
 	_BIT_0 = 1 << 0
 
+
 class ResourceAttrs(enum.Flag):
 	"""Resource attribute flags. The descriptions for these flags are taken from comments on the res*Bit and res* enum constants in <CarbonCore/Resources.h>."""
 	
@@ -93,23 +97,37 @@ class ResourceAttrs(enum.Flag):
 	resChanged = 1 << 1 # "Existing resource changed since last update", "Resource changed?"
 	resCompressed = 1 << 0 # "indicates that the resource data is compressed" (only documented in https://github.com/kreativekorp/ksfl/wiki/Macintosh-Resource-File-Format)
 
+
 class Resource(object):
 	"""A single resource from a resource file."""
 	
-	__slots__ = ("resource_type", "resource_id", "name", "attributes", "data_raw", "_data_decompressed")
+	_resfile: "ResourceFile"
+	type: bytes
+	id: int
+	name_offset: int
+	_name: typing.Optional[bytes]
+	attributes: ResourceAttrs
+	data_raw_offset: int
+	_data_raw: bytes
+	_compressed_info: compress.common.CompressedHeaderInfo
+	_data_decompressed: bytes
 	
-	def __init__(self, resource_type: bytes, resource_id: int, name: typing.Optional[bytes], attributes: ResourceAttrs, data_raw: bytes):
-		"""Create a new resource with the given type code, ID, name, attributes, and data."""
+	def __init__(self, resfile: "ResourceFile", resource_type: bytes, resource_id: int, name_offset: int, attributes: ResourceAttrs, data_raw_offset: int) -> None:
+		"""Create a resource object representing a resource stored in a resource file.
+		
+		External code should not call this constructor manually. Resources should be looked up through a ResourceFile object instead.
+		"""
 		
 		super().__init__()
 		
-		self.resource_type: bytes = resource_type
-		self.resource_id: int = resource_id
-		self.name: typing.Optional[bytes] = name
-		self.attributes: ResourceAttrs = attributes
-		self.data_raw: bytes = data_raw
+		self._resfile = resfile
+		self.type = resource_type
+		self.id = resource_id
+		self.name_offset = name_offset
+		self.attributes = attributes
+		self.data_raw_offset = data_raw_offset
 	
-	def __repr__(self):
+	def __repr__(self) -> str:
 		try:
 			data = self.data
 		except compress.DecompressError:
@@ -119,14 +137,85 @@ class Resource(object):
 			decompress_ok = True
 		
 		if len(data) > 32:
-			data_repr = f"<{len(data)} bytes: {data[:32]}...>"
+			data_repr = f"<{len(data)} bytes: {data[:32]!r}...>"
 		else:
 			data_repr = repr(data)
 		
 		if not decompress_ok:
 			data_repr = f"<decompression failed - compressed data: {data_repr}>"
 		
-		return f"{type(self).__module__}.{type(self).__qualname__}(resource_type={self.resource_type}, resource_id={self.resource_id}, name={self.name}, attributes={self.attributes}, data={data_repr})"
+		return f"<{type(self).__qualname__} type {self.type!r}, id {self.id}, name {self.name!r}, attributes {self.attributes}, data {data_repr}>"
+	
+	@property
+	def resource_type(self) -> bytes:
+		warnings.warn(DeprecationWarning("The resource_type attribute has been deprecated and will be removed in a future version. Please use the type attribute instead."))
+		return self.type
+	
+	@property
+	def resource_id(self) -> int:
+		warnings.warn(DeprecationWarning("The resource_id attribute has been deprecated and will be removed in a future version. Please use the id attribute instead."))
+		return self.id
+	
+	@property
+	def name(self) -> typing.Optional[bytes]:
+		try:
+			return self._name
+		except AttributeError:
+			if self.name_offset == 0xffff:
+				self._name = None
+			else:
+				self._resfile._stream.seek(self._resfile.map_offset + self._resfile.map_name_list_offset + self.name_offset)
+				(name_length,) = self._resfile._stream_unpack(STRUCT_RESOURCE_NAME_HEADER)
+				self._name = self._resfile._read_exact(name_length)
+			
+			return self._name
+	
+	@property
+	def data_raw(self) -> bytes:
+		try:
+			return self._data_raw
+		except AttributeError:
+			self._resfile._stream.seek(self._resfile.data_offset + self.data_raw_offset)
+			(data_raw_length,) = self._resfile._stream_unpack(STRUCT_RESOURCE_DATA_HEADER)
+			self._data_raw = self._resfile._read_exact(data_raw_length)
+			return self._data_raw
+	
+	@property
+	def compressed_info(self) -> typing.Optional[compress.common.CompressedHeaderInfo]:
+		"""The compressed resource header information, or None if this resource is not compressed.
+		
+		Accessing this attribute may raise a DecompressError if the resource data is compressed and the header could not be parsed. To access the unparsed header data, use the data_raw attribute.
+		"""
+		
+		if ResourceAttrs.resCompressed in self.attributes:
+			try:
+				return self._compressed_info
+			except AttributeError:
+				self._compressed_info = compress.common.CompressedHeaderInfo.parse(self.data_raw)
+				return self._compressed_info
+		else:
+			return None
+	
+	@property
+	def length_raw(self) -> int:
+		"""The length of the raw resource data, which may be compressed.
+		
+		Accessing this attribute may be faster than computing len(self.data_raw) manually.
+		"""
+		
+		return len(self.data_raw)
+	
+	@property
+	def length(self) -> int:
+		"""The length of the resource data. If the resource data is compressed, this is the length of the data after decompression.
+		
+		Accessing this attribute may be faster than computing len(self.data) manually.
+		"""
+		
+		if self.compressed_info is not None:
+			return self.compressed_info.decompressed_length
+		else:
+			return self.length_raw
 	
 	@property
 	def data(self) -> bytes:
@@ -135,72 +224,84 @@ class Resource(object):
 		Accessing this attribute may raise a DecompressError if the resource data is compressed and could not be decompressed. To access the compressed resource data, use the data_raw attribute.
 		"""
 		
-		if ResourceAttrs.resCompressed in self.attributes:
+		if self.compressed_info is not None:
 			try:
 				return self._data_decompressed
 			except AttributeError:
-				self._data_decompressed = compress.decompress(self.data_raw)
+				self._data_decompressed = compress.decompress_parsed(self.compressed_info, self.data_raw[self.compressed_info.header_length:])
 				return self._data_decompressed
 		else:
 			return self.data_raw
 
-class ResourceFile(collections.abc.Mapping):
+
+class _LazyResourceMap(typing.Mapping[int, Resource]):
+	"""Internal class: Read-only wrapper for a mapping of resource IDs to resource objects.
+	
+	This class behaves like a normal read-only mapping. The main difference to a plain dict (or similar mapping) is that this mapping has a specialized repr to avoid excessive output when working in the REPL.
+	"""
+	
+	type: bytes
+	_submap: typing.Mapping[int, Resource]
+	
+	def __init__(self, resource_type: bytes, submap: typing.Mapping[int, Resource]) -> None:
+		"""Create a new _LazyResourceMap that wraps the given mapping."""
+		
+		super().__init__()
+		
+		self.type = resource_type
+		self._submap = submap
+	
+	def __len__(self) -> int:
+		"""Get the number of resources with this type code."""
+		
+		return len(self._submap)
+	
+	def __iter__(self) -> typing.Iterator[int]:
+		"""Iterate over the IDs of all resources with this type code."""
+		
+		return iter(self._submap)
+	
+	def __contains__(self, key: object) -> bool:
+		"""Check if a resource with the given ID exists for this type code."""
+		
+		return key in self._submap
+	
+	def __getitem__(self, key: int) -> Resource:
+		"""Get a resource with the given ID for this type code."""
+		
+		return self._submap[key]
+	
+	def __repr__(self) -> str:
+		if len(self) == 1:
+			contents = f"one resource: {next(iter(self.values()))}"
+		else:
+			contents = f"{len(self)} resources with IDs {list(self)}"
+		
+		return f"<Resource map for type {self.type!r}, containing {contents}>"
+
+
+class ResourceFile(typing.Mapping[bytes, typing.Mapping[int, Resource]], typing.ContextManager["ResourceFile"]):
 	"""A resource file reader operating on a byte stream."""
 	
-	# noinspection PyProtectedMember
-	class _LazyResourceMap(collections.abc.Mapping):
-		"""Internal class: Lazy mapping of resource IDs to resource objects, returned when subscripting a ResourceFile."""
-		
-		def __init__(self, resfile: "ResourceFile", restype: bytes):
-			"""Create a new _LazyResourceMap "containing" all resources in resfile that have the type code restype."""
-			
-			super().__init__()
-			
-			self._resfile: "ResourceFile" = resfile
-			self._restype: bytes = restype
-			self._submap: typing.Mapping[int, typing.Tuple[int, ResourceAttrs, int]] = self._resfile._references[self._restype]
-		
-		def __len__(self):
-			"""Get the number of resources with this type code."""
-			
-			return len(self._submap)
-		
-		def __iter__(self):
-			"""Iterate over the IDs of all resources with this type code."""
-			
-			return iter(self._submap)
-		
-		def __contains__(self, key: int):
-			"""Check if a resource with the given ID exists for this type code."""
-			
-			return key in self._submap
-		
-		def __getitem__(self, key: int) -> Resource:
-			"""Get a resource with the given ID for this type code."""
-			
-			name_offset, attributes, data_offset = self._submap[key]
-			
-			if name_offset == 0xffff:
-				name = None
-			else:
-				self._resfile._stream.seek(self._resfile.map_offset + self._resfile.map_name_list_offset + name_offset)
-				(name_length,) = self._resfile._stream_unpack(STRUCT_RESOURCE_NAME_HEADER)
-				name = self._resfile._read_exact(name_length)
-			
-			self._resfile._stream.seek(self._resfile.data_offset + data_offset)
-			(data_length,) = self._resfile._stream_unpack(STRUCT_RESOURCE_DATA_HEADER)
-			data = self._resfile._read_exact(data_length)
-			
-			return Resource(self._restype, key, name, attributes, data)
-		
-		def __repr__(self):
-			if len(self) == 1:
-				return f"<{type(self).__module__}.{type(self).__qualname__} at {id(self):#x} containing one resource: {next(iter(self.values()))}>"
-			else:
-				return f"<{type(self).__module__}.{type(self).__qualname__} at {id(self):#x} containing {len(self)} resources with IDs: {list(self)}>"
+	_close_stream: bool
+	_stream: typing.BinaryIO
+	
+	data_offset: int
+	map_offset: int
+	data_length: int
+	map_length: int
+	header_system_data: bytes
+	header_application_data: bytes
+	
+	map_type_list_offset: int
+	map_name_list_offset: int
+	file_attributes: ResourceFileAttrs
+	
+	_reference_counts: typing.MutableMapping[bytes, int]
+	_references: typing.MutableMapping[bytes, typing.MutableMapping[int, Resource]]
 	
 	@classmethod
-	def open(cls, filename: typing.Union[str, bytes, os.PathLike], *, fork: str="auto", **kwargs) -> "ResourceFile":
+	def open(cls, filename: typing.Union[str, os.PathLike], *, fork: str = "auto", **kwargs: typing.Any) -> "ResourceFile":
 		"""Open the file at the given path as a ResourceFile.
 		
 		The fork parameter controls which fork of the file the resource data will be read from. It accepts the following values:
@@ -259,7 +360,7 @@ class ResourceFile(collections.abc.Mapping):
 		else:
 			raise ValueError(f"Unsupported value for the fork parameter: {fork!r}")
 	
-	def __init__(self, stream: typing.io.BinaryIO, *, close: bool=False):
+	def __init__(self, stream: typing.BinaryIO, *, close: bool = False) -> None:
 		"""Create a ResourceFile wrapping the given byte stream.
 		
 		To read resource file data from a bytes object, wrap it in an io.BytesIO.
@@ -273,8 +374,7 @@ class ResourceFile(collections.abc.Mapping):
 		
 		super().__init__()
 		
-		self._close_stream: bool = close
-		self._stream: typing.io.BinaryIO
+		self._close_stream = close
 		if stream.seekable():
 			self._stream = stream
 		else:
@@ -298,7 +398,7 @@ class ResourceFile(collections.abc.Mapping):
 			raise InvalidResourceFileError(f"Attempted to read {byte_count} bytes of data, but only got {len(data)} bytes")
 		return data
 	
-	def _stream_unpack(self, st: struct.Struct) -> typing.Tuple:
+	def _stream_unpack(self, st: struct.Struct) -> tuple:
 		"""Unpack data from the stream according to the struct st. The number of bytes to read is determined using st.size, so variable-sized structs cannot be used with this method."""
 		
 		try:
@@ -306,17 +406,11 @@ class ResourceFile(collections.abc.Mapping):
 		except struct.error as e:
 			raise InvalidResourceFileError(str(e))
 	
-	def _read_header(self):
+	def _read_header(self) -> None:
 		"""Read the resource file header, starting at the current stream position."""
 		
 		assert self._stream.tell() == 0
 		
-		self.data_offset: int
-		self.map_offset: int
-		self.data_length: int
-		self.map_length: int
-		self.header_system_data: bytes
-		self.header_application_data: bytes
 		(
 			self.data_offset,
 			self.map_offset,
@@ -329,25 +423,23 @@ class ResourceFile(collections.abc.Mapping):
 		if self._stream.tell() != self.data_offset:
 			raise InvalidResourceFileError(f"The data offset ({self.data_offset}) should point exactly to the end of the file header ({self._stream.tell()})")
 	
-	def _read_map_header(self):
+	def _read_map_header(self) -> None:
 		"""Read the map header, starting at the current stream position."""
 		
 		assert self._stream.tell() == self.map_offset
 		
-		self.map_type_list_offset: int
-		self.map_name_list_offset: int
 		(
 			_file_attributes,
 			self.map_type_list_offset,
 			self.map_name_list_offset,
 		) = self._stream_unpack(STRUCT_RESOURCE_MAP_HEADER)
 		
-		self.file_attributes: ResourceFileAttrs = ResourceFileAttrs(_file_attributes)
+		self.file_attributes = ResourceFileAttrs(_file_attributes)
 	
-	def _read_all_resource_types(self):
+	def _read_all_resource_types(self) -> None:
 		"""Read all resource types, starting at the current stream position."""
 		
-		self._reference_counts: typing.MutableMapping[bytes, int] = collections.OrderedDict()
+		self._reference_counts = collections.OrderedDict()
 		
 		(type_list_length_m1,) = self._stream_unpack(STRUCT_RESOURCE_TYPE_LIST_HEADER)
 		type_list_length = (type_list_length_m1 + 1) % 0x10000
@@ -361,13 +453,13 @@ class ResourceFile(collections.abc.Mapping):
 			count = (count_m1 + 1) % 0x10000
 			self._reference_counts[resource_type] = count
 	
-	def _read_all_references(self):
+	def _read_all_references(self) -> None:
 		"""Read all resource references, starting at the current stream position."""
 		
-		self._references: typing.MutableMapping[bytes, typing.MutableMapping[int, typing.Tuple[int, ResourceAttrs, int]]] = collections.OrderedDict()
+		self._references = collections.OrderedDict()
 		
 		for resource_type, count in self._reference_counts.items():
-			resmap: typing.MutableMapping[int, typing.Tuple[int, ResourceAttrs, int]] = collections.OrderedDict()
+			resmap: typing.MutableMapping[int, Resource] = collections.OrderedDict()
 			self._references[resource_type] = resmap
 			for _ in range(count):
 				(
@@ -379,9 +471,9 @@ class ResourceFile(collections.abc.Mapping):
 				attributes = attributes_and_data_offset >> 24
 				data_offset = attributes_and_data_offset & ((1 << 24) - 1)
 				
-				resmap[resource_id] = (name_offset, ResourceAttrs(attributes), data_offset)
+				resmap[resource_id] = Resource(self, resource_type, resource_id, name_offset, ResourceAttrs(attributes), data_offset)
 	
-	def close(self):
+	def close(self) -> None:
 		"""Close this ResourceFile.
 		
 		If close=True was passed when this ResourceFile was created, the underlying stream's close method is called as well.
@@ -390,31 +482,37 @@ class ResourceFile(collections.abc.Mapping):
 		if self._close_stream:
 			self._stream.close()
 	
-	def __enter__(self):
-		pass
+	def __enter__(self) -> "ResourceFile":
+		return self
 	
-	def __exit__(self, exc_type, exc_val, exc_tb):
+	def __exit__(
+		self,
+		exc_type: typing.Optional[typing.Type[BaseException]],
+		exc_val: typing.Optional[BaseException],
+		exc_tb: typing.Optional[types.TracebackType]
+	) -> typing.Optional[bool]:
 		self.close()
+		return None
 	
-	def __len__(self):
+	def __len__(self) -> int:
 		"""Get the number of resource types in this ResourceFile."""
 		
 		return len(self._references)
 	
-	def __iter__(self):
+	def __iter__(self) -> typing.Iterator[bytes]:
 		"""Iterate over all resource types in this ResourceFile."""
 		
 		return iter(self._references)
 	
-	def __contains__(self, key: bytes):
+	def __contains__(self, key: object) -> bool:
 		"""Check whether this ResourceFile contains any resources of the given type."""
 		
 		return key in self._references
 	
-	def __getitem__(self, key: bytes) -> "ResourceFile._LazyResourceMap":
+	def __getitem__(self, key: bytes) -> "_LazyResourceMap":
 		"""Get a lazy mapping of all resources with the given type in this ResourceFile."""
 		
-		return ResourceFile._LazyResourceMap(self, key)
+		return _LazyResourceMap(key, self._references[key])
 	
-	def __repr__(self):
+	def __repr__(self) -> str:
 		return f"<{type(self).__module__}.{type(self).__qualname__} at {id(self):#x}, attributes {self.file_attributes}, containing {len(self)} resource types: {list(self)}>"

rsrcfork/compress/__init__.py

@@ -1,97 +1,68 @@
-import struct
+import io
+import typing
 
 from . import dcmp0
 from . import dcmp1
 from . import dcmp2
 
-from .common import DecompressError
+from .common import DecompressError, CompressedHeaderInfo, CompressedType8HeaderInfo, CompressedType9HeaderInfo
 
 __all__ = [
+	"CompressedHeaderInfo",
+	"CompressedType8HeaderInfo",
+	"CompressedType9HeaderInfo",
 	"DecompressError",
 	"decompress",
+	"decompress_parsed",
+	"decompress_stream",
+	"decompress_stream_parsed",
 ]
 
-# The signature of all compressed resource data, 0xa89f6572 in hex, or "®üer" in MacRoman.
-COMPRESSED_SIGNATURE = b"\xa8\x9fer"
-# The compression type commonly used for application resources.
-COMPRESSED_TYPE_APPLICATION = 0x0801
-# The compression type commonly used for System file resources.
-COMPRESSED_TYPE_SYSTEM = 0x0901
 
-# Common header for compressed resources of all types.
-# 4 bytes: Signature (see above).
-# 2 bytes: Length of the complete header (this common part and the type-specific part that follows it). (This meaning is just a guess - the field's value is always 0x0012, so there's no way to know for certain what it means.)
-# 2 bytes: Compression type. Known so far: 0x0901 is used in the System file's resources. 0x0801 is used in other files' resources.
-# 4 bytes: Length of the data after decompression.
-STRUCT_COMPRESSED_HEADER = struct.Struct(">4sHHI")
-
-# Header continuation part for an "application" compressed resource.
-# 1 byte: "Working buffer fractional size" - the ratio of the compressed data size to the uncompressed data size, times 256.
-# 1 byte: "Expansion buffer size" - the maximum number of bytes that the data might grow during decompression.
-# 2 bytes: The ID of the 'dcmp' resource that can decompress this resource. Currently only ID 0 is supported.
-# 2 bytes: Reserved (always zero).
-STRUCT_COMPRESSED_APPLICATION_HEADER = struct.Struct(">BBhH")
-
-# Header continuation part for a "system" compressed resource.
-# 2 bytes: The ID of the 'dcmp' resource that can decompress this resource. Currently only ID 2 is supported.
-# 4 bytes: Decompressor-specific parameters.
-STRUCT_COMPRESSED_SYSTEM_HEADER = struct.Struct(">h4s")
+# Maps 'dcmp' IDs to their corresponding Python implementations.
+# Each decompressor has the signature (header_info: CompressedHeaderInfo, stream: typing.BinaryIO, *, debug: bool=False) -> typing.Iterator[bytes].
+DECOMPRESSORS = {
+	0: dcmp0.decompress_stream,
+	1: dcmp1.decompress_stream,
+	2: dcmp2.decompress_stream,
+}
 
 
-def _decompress_application(data: bytes, decompressed_length: int, *, debug: bool=False) -> bytes:
-	working_buffer_fractional_size, expansion_buffer_size, dcmp_id, reserved = STRUCT_COMPRESSED_APPLICATION_HEADER.unpack_from(data)
-	
-	if debug:
-		print(f"Working buffer fractional size: {working_buffer_fractional_size} (=> {len(data) * 256 / working_buffer_fractional_size})")
-		print(f"Expansion buffer size: {expansion_buffer_size}")
-	
-	if dcmp_id == 0:
-		decompress_func = dcmp0.decompress
-	elif dcmp_id == 1:
-		decompress_func = dcmp1.decompress
-	else:
-		raise DecompressError(f"Unsupported 'dcmp' ID: {dcmp_id}, expected 0 or 1")
-	
-	if reserved != 0:
-		raise DecompressError(f"Reserved field should be 0, not 0x{reserved:>04x}")
-	
-	return decompress_func(data[STRUCT_COMPRESSED_APPLICATION_HEADER.size:], decompressed_length, debug=debug)
-
-
-def _decompress_system(data: bytes, decompressed_length: int, *, debug: bool=False) -> bytes:
-	dcmp_id, params = STRUCT_COMPRESSED_SYSTEM_HEADER.unpack_from(data)
-	
-	if dcmp_id == 2:
-		decompress_func = dcmp2.decompress
-	else:
-		raise DecompressError(f"Unsupported 'dcmp' ID: {dcmp_id}, expected 2")
-	
-	return decompress_func(data[STRUCT_COMPRESSED_SYSTEM_HEADER.size:], decompressed_length, params, debug=debug)
-
-
-def decompress(data: bytes, *, debug: bool=False) -> bytes:
-	"""Decompress the given compressed resource data."""
+def decompress_stream_parsed(header_info: CompressedHeaderInfo, stream: typing.BinaryIO, *, debug: bool = False) -> typing.Iterator[bytes]:
+	"""Decompress compressed resource data from a stream, whose header has already been read and parsed into a CompressedHeaderInfo object."""
 	
 	try:
-		signature, header_length, compression_type, decompressed_length = STRUCT_COMPRESSED_HEADER.unpack_from(data)
-	except struct.error:
-		raise DecompressError(f"Invalid header")
-	if signature != COMPRESSED_SIGNATURE:
-		raise DecompressError(f"Invalid signature: {signature!r}, expected {COMPRESSED_SIGNATURE}")
-	if header_length != 0x12:
-		raise DecompressError(f"Unsupported header length: 0x{header_length:>04x}, expected 0x12")
+		decompress_func = DECOMPRESSORS[header_info.dcmp_id]
+	except KeyError:
+		raise DecompressError(f"Unsupported 'dcmp' ID: {header_info.dcmp_id}")
 	
-	if compression_type == COMPRESSED_TYPE_APPLICATION:
-		decompress_func = _decompress_application
-	elif compression_type == COMPRESSED_TYPE_SYSTEM:
-		decompress_func = _decompress_system
-	else:
-		raise DecompressError(f"Unsupported compression type: 0x{compression_type:>04x}")
+	decompressed_length = 0
+	for chunk in decompress_func(header_info, stream, debug=debug):
+		decompressed_length += len(chunk)
+		yield chunk
+	
+	if decompressed_length != header_info.decompressed_length:
+		raise DecompressError(f"Actual length of decompressed data ({decompressed_length}) does not match length stored in resource ({header_info.decompressed_length})")
+
+
+def decompress_parsed(header_info: CompressedHeaderInfo, data: bytes, *, debug: bool = False) -> bytes:
+	"""Decompress the given compressed resource data, whose header has already been removed and parsed into a CompressedHeaderInfo object."""
+	
+	return b"".join(decompress_stream_parsed(header_info, io.BytesIO(data), debug=debug))
+
+
+def decompress_stream(stream: typing.BinaryIO, *, debug: bool = False) -> typing.Iterator[bytes]:
+	"""Decompress compressed resource data from a stream."""
+	
+	header_info = CompressedHeaderInfo.parse_stream(stream)
 	
 	if debug:
-		print(f"Decompressed length: {decompressed_length}")
+		print(f"Compressed resource data header: {header_info}")
 	
-	decompressed = decompress_func(data[STRUCT_COMPRESSED_HEADER.size:], decompressed_length, debug=debug)
-	if len(decompressed) != decompressed_length:
-		raise DecompressError(f"Actual length of decompressed data ({len(decompressed)}) does not match length stored in resource ({decompressed_length})")
-	return decompressed
+	yield from decompress_stream_parsed(header_info, stream, debug=debug)
+
+
+def decompress(data: bytes, *, debug: bool = False) -> bytes:
+	"""Decompress the given compressed resource data."""
+	
+	return b"".join(decompress_stream(io.BytesIO(data), debug=debug))

rsrcfork/compress/common.py

@@ -1,3 +1,5 @@
+import io
+import struct
 import typing
 
 
@@ -5,19 +7,199 @@ class DecompressError(Exception):
 	"""Raised when resource data decompression fails, because the data is invalid or the compression type is not supported."""
 
 
-def _read_variable_length_integer(data: bytes, position: int) -> typing.Tuple[int, int]:
-	"""Read a variable-length integer starting at the given position in the data, and return the integer as well as the number of bytes consumed.
+# The signature of all compressed resource data, 0xa89f6572 in hex, or "®üer" in MacRoman.
+COMPRESSED_SIGNATURE = b"\xa8\x9fer"
+# The number of the "type 8" compression type. This type is used in the Finder, ResEdit, and some other system files.
+COMPRESSED_TYPE_8 = 0x0801
+# The number of the "type 9" compression type. This type is used in the System file and System 7.5's Installer.
+COMPRESSED_TYPE_9 = 0x0901
+
+# Common header for compressed resources of all types.
+# 4 bytes: Signature (see above).
+# 2 bytes: Length of the complete header (this common part and the type-specific part that follows it). (This meaning is just a guess - the field's value is always 0x0012, so there's no way to know for certain what it means.)
+# 2 bytes: Compression type. Known so far: 0x0801 ("type 8") and 0x0901 ("type 9").
+# 4 bytes: Length of the data after decompression.
+# 6 bytes: Remainder of the header. The exact format varies depending on the compression type.
+STRUCT_COMPRESSED_HEADER = struct.Struct(">4sHHI6s")
+
+# Remainder of header for a "type 8" compressed resource.
+# 1 byte: "Working buffer fractional size" - the ratio of the compressed data size to the uncompressed data size, times 256.
+# 1 byte: "Expansion buffer size" - the maximum number of bytes that the data might grow during decompression.
+# 2 bytes: The ID of the 'dcmp' resource that can decompress this resource. Currently only ID 0 is supported.
+# 2 bytes: Reserved (always zero).
+STRUCT_COMPRESSED_TYPE_8_HEADER = struct.Struct(">BBhH")
+
+# Remainder of header for a "type 9" compressed resource.
+# 2 bytes: The ID of the 'dcmp' resource that can decompress this resource. Currently only ID 2 is supported.
+# 4 bytes: Decompressor-specific parameters.
+STRUCT_COMPRESSED_TYPE_9_HEADER = struct.Struct(">h4s")
+
+
+class CompressedHeaderInfo(object):
+	@classmethod
+	def parse_stream(cls, stream: typing.BinaryIO) -> "CompressedHeaderInfo":
+		try:
+			signature, header_length, compression_type, decompressed_length, remainder = STRUCT_COMPRESSED_HEADER.unpack(stream.read(STRUCT_COMPRESSED_HEADER.size))
+		except struct.error:
+			raise DecompressError("Invalid header")
+		if signature != COMPRESSED_SIGNATURE:
+			raise DecompressError(f"Invalid signature: {signature!r}, expected {COMPRESSED_SIGNATURE!r}")
+		if header_length != 0x12:
+			raise DecompressError(f"Unsupported header length: 0x{header_length:>04x}, expected 0x12")
+		
+		if compression_type == COMPRESSED_TYPE_8:
+			working_buffer_fractional_size, expansion_buffer_size, dcmp_id, reserved = STRUCT_COMPRESSED_TYPE_8_HEADER.unpack(remainder)
+			
+			if reserved != 0:
+				raise DecompressError(f"Reserved field should be 0, not 0x{reserved:>04x}")
+			
+			return CompressedType8HeaderInfo(header_length, compression_type, decompressed_length, dcmp_id, working_buffer_fractional_size, expansion_buffer_size)
+		elif compression_type == COMPRESSED_TYPE_9:
+			dcmp_id, parameters = STRUCT_COMPRESSED_TYPE_9_HEADER.unpack(remainder)
+			
+			return CompressedType9HeaderInfo(header_length, compression_type, decompressed_length, dcmp_id, parameters)
+		else:
+			raise DecompressError(f"Unsupported compression type: 0x{compression_type:>04x}")
+	
+	@classmethod
+	def parse(cls, data: bytes) -> "CompressedHeaderInfo":
+		return cls.parse_stream(io.BytesIO(data))
+	
+	header_length: int
+	compression_type: int
+	decompressed_length: int
+	dcmp_id: int
+	
+	def __init__(self, header_length: int, compression_type: int, decompressed_length: int, dcmp_id: int) -> None:
+		super().__init__()
+		
+		self.header_length = header_length
+		self.compression_type = compression_type
+		self.decompressed_length = decompressed_length
+		self.dcmp_id = dcmp_id
+
+
+class CompressedType8HeaderInfo(CompressedHeaderInfo):
+	working_buffer_fractional_size: int
+	expansion_buffer_size: int
+	
+	def __init__(self, header_length: int, compression_type: int, decompressed_length: int, dcmp_id: int, working_buffer_fractional_size: int, expansion_buffer_size: int) -> None:
+		super().__init__(header_length, compression_type, decompressed_length, dcmp_id)
+		
+		self.working_buffer_fractional_size = working_buffer_fractional_size
+		self.expansion_buffer_size = expansion_buffer_size
+	
+	def __repr__(self) -> str:
+		return f"{type(self).__qualname__}(header_length={self.header_length}, compression_type=0x{self.compression_type:>04x}, decompressed_length={self.decompressed_length}, dcmp_id={self.dcmp_id}, working_buffer_fractional_size={self.working_buffer_fractional_size}, expansion_buffer_size={self.expansion_buffer_size})"
+
+
+class CompressedType9HeaderInfo(CompressedHeaderInfo):
+	parameters: bytes
+	
+	def __init__(self, header_length: int, compression_type: int, decompressed_length: int, dcmp_id: int, parameters: bytes) -> None:
+		super().__init__(header_length, compression_type, decompressed_length, dcmp_id)
+		
+		self.parameters = parameters
+	
+	def __repr__(self) -> str:
+		return f"{type(self).__qualname__}(header_length={self.header_length}, compression_type=0x{self.compression_type:>04x}, decompressed_length={self.decompressed_length}, dcmp_id={self.dcmp_id}, parameters={self.parameters!r})"
+
+
+if typing.TYPE_CHECKING:
+	class PeekableIO(typing.Protocol):
+		"""Minimal protocol for binary IO streams that support the peek method.
+		
+		The peek method is supported by various standard Python binary IO streams, such as io.BufferedReader. If a stream does not natively support the peek method, it may be wrapped using the custom helper function make_peekable.
+		"""
+		
+		def readable(self) -> bool:
+			...
+		
+		def read(self, size: typing.Optional[int] = ...) -> bytes:
+			...
+		
+		def peek(self, size: int = ...) -> bytes:
+			...
+
+
+class _PeekableIOWrapper(object):
+	"""Wrapper class to add peek support to an existing stream. Do not instantiate this class directly, use the make_peekable function instead.
+	
+	Python provides a standard io.BufferedReader class, which supports the peek method. However, according to its documentation, it only supports wrapping io.RawIOBase subclasses, and not streams which are already otherwise buffered.
+	
+	Warning: this class does not perform any buffering of its own, outside of what is required to make peek work. It is strongly recommended to only wrap streams that are already buffered or otherwise fast to read from. In particular, raw streams (io.RawIOBase subclasses) should be wrapped using io.BufferedReader instead.
+	"""
+	
+	_wrapped: typing.BinaryIO
+	_readahead: bytes
+	
+	def __init__(self, wrapped: typing.BinaryIO) -> None:
+		super().__init__()
+		
+		self._wrapped = wrapped
+		self._readahead = b""
+	
+	def readable(self) -> bool:
+		return self._wrapped.readable()
+	
+	def read(self, size: typing.Optional[int] = None) -> bytes:
+		if size is None or size < 0:
+			ret = self._readahead + self._wrapped.read()
+			self._readahead = b""
+		elif size <= len(self._readahead):
+			ret = self._readahead[:size]
+			self._readahead = self._readahead[size:]
+		else:
+			ret = self._readahead + self._wrapped.read(size - len(self._readahead))
+			self._readahead = b""
+		
+		return ret
+	
+	def peek(self, size: int = -1) -> bytes:
+		if not self._readahead:
+			self._readahead = self._wrapped.read(io.DEFAULT_BUFFER_SIZE if size < 0 else size)
+		return self._readahead
+
+
+def make_peekable(stream: typing.BinaryIO) -> "PeekableIO":
+	"""Wrap an arbitrary binary IO stream so that it supports the peek method.
+	
+	The stream is wrapped as efficiently as possible (or not at all if it already supports the peek method). However, in the worst case a custom wrapper class needs to be used, which may not be particularly efficient and only supports a very minimal interface. The only methods that are guaranteed to exist on the returned stream are readable, read, and peek.
+	"""
+	
+	if hasattr(stream, "peek"):
+		# Stream is already peekable, nothing to be done.
+		return typing.cast("PeekableIO", stream)
+	elif not typing.TYPE_CHECKING and isinstance(stream, io.RawIOBase):
+		# This branch is skipped when type checking - mypy incorrectly warns about this code being unreachable, because it thinks that a typing.BinaryIO cannot be an instance of io.RawIOBase.
+		# Raw IO streams can be wrapped efficiently using BufferedReader.
+		return io.BufferedReader(stream)
+	else:
+		# Other streams need to be wrapped using our custom wrapper class.
+		return _PeekableIOWrapper(stream)
+
+
+def read_exact(stream: typing.BinaryIO, byte_count: int) -> bytes:
+	"""Read byte_count bytes from the stream and raise an exception if too few bytes are read (i. e. if EOF was hit prematurely)."""
+	
+	data = stream.read(byte_count)
+	if len(data) != byte_count:
+		raise DecompressError(f"Attempted to read {byte_count} bytes of data, but only got {len(data)} bytes")
+	return data
+
+
+def read_variable_length_integer(stream: typing.BinaryIO) -> int:
+	"""Read a variable-length integer from the stream.
 	
 	This variable-length integer format is used by the 0xfe codes in the compression formats used by 'dcmp' (0) and 'dcmp' (1).
 	"""
 	
-	assert len(data) > position
-	if data[position] == 0xff:
-		assert len(data) > position + 4
-		return int.from_bytes(data[position+1:position+5], "big", signed=True), 5
-	elif data[position] >= 0x80:
-		assert len(data) > position + 1
-		data_modified = bytes([(data[position] - 0xc0) & 0xff, data[position+1]])
-		return int.from_bytes(data_modified, "big", signed=True), 2
+	head = read_exact(stream, 1)
+	
+	if head[0] == 0xff:
+		return int.from_bytes(read_exact(stream, 4), "big", signed=True)
+	elif head[0] >= 0x80:
+		data_modified = bytes([(head[0] - 0xc0) & 0xff]) + read_exact(stream, 1)
+		return int.from_bytes(data_modified, "big", signed=True)
 	else:
-		return int.from_bytes(data[position:position+1], "big", signed=True), 1
+		return int.from_bytes(head, "big", signed=True)

rsrcfork/compress/dcmp0.py

@@ -1,3 +1,5 @@
+import typing
+
 from . import common
 
 # Lookup table for codes in range(0x4b, 0xfe).
@@ -36,133 +38,103 @@ TABLE = [TABLE_DATA[i:i + 2] for i in range(0, len(TABLE_DATA), 2)]
 assert len(TABLE) == len(range(0x4b, 0xfe))
 
 
-def decompress(data: bytes, decompressed_length: int, *, debug: bool=False) -> bytes:
-	"""Decompress compressed data in the format used by 'dcmp' (0)."""
+def decompress_stream_inner(header_info: common.CompressedHeaderInfo, stream: typing.BinaryIO, *, debug: bool = False) -> typing.Iterator[bytes]:
+	"""Internal helper function, implements the main decompression algorithm. Only called from decompress_stream, which performs some extra checks and debug logging."""
 	
-	prev_literals = []
-	decompressed = b""
+	if not isinstance(header_info, common.CompressedType8HeaderInfo):
+		raise common.DecompressError(f"Incorrect header type: {type(header_info).__qualname__}")
 	
-	i = 0
+	prev_literals: typing.List[bytes] = []
 	
-	while i < len(data):
-		byte = data[i]
+	while True: # Loop is terminated when the EOF marker (0xff) is encountered
+		(byte,) = common.read_exact(stream, 1)
 		if debug:
-			print(f"Tag byte 0x{byte:>02x}, at 0x{i:x}, decompressing to 0x{len(decompressed):x}")
+			print(f"Tag byte 0x{byte:>02x}")
 		
 		if byte in range(0x00, 0x20):
 			# Literal byte sequence.
 			if byte in (0x00, 0x10):
 				# The length of the literal data is stored in the next byte.
-				count_div2 = data[i+1]
-				begin = i + 2
+				(count_div2,) = common.read_exact(stream, 1)
 			else:
 				# The length of the literal data is stored in the low nibble of the tag byte.
 				count_div2 = byte >> 0 & 0xf
-				begin = i + 1
-			end = begin + 2*count_div2
+			count = 2 * count_div2
 			# Controls whether or not the literal is stored so that it can be referenced again later.
 			do_store = byte >= 0x10
-			literal = data[begin:end]
+			literal = common.read_exact(stream, count)
 			if debug:
 				print(f"Literal (storing: {do_store})")
-				print(f"\t-> {literal}")
-			decompressed += literal
 			if do_store:
 				if debug:
-					print(f"\t-> stored as literal number 0x{len(prev_literals):x}")
+					print(f"\t-> storing as literal number 0x{len(prev_literals):x}")
 				prev_literals.append(literal)
-			i = end
+			yield literal
 		elif byte in (0x20, 0x21):
 			# Backreference to a previous literal, 2-byte form.
 			# This can reference literals with index in range(0x28, 0x228).
-			table_index = 0x28 + ((byte - 0x20) << 8 | data[i+1])
-			i += 2
+			(next_byte,) = common.read_exact(stream, 1)
+			table_index = 0x28 + ((byte - 0x20) << 8 | next_byte)
 			if debug:
 				print(f"Backreference (2-byte form) to 0x{table_index:>02x}")
-			literal = prev_literals[table_index]
-			if debug:
-				print(f"\t-> {literal}")
-			decompressed += literal
+			yield prev_literals[table_index]
 		elif byte == 0x22:
 			# Backreference to a previous literal, 3-byte form.
 			# This can reference any literal with index 0x28 and higher, but is only necessary for literals with index 0x228 and higher.
-			table_index = 0x28 + int.from_bytes(data[i+1:i+3], "big", signed=False)
-			i += 3
+			table_index = 0x28 + int.from_bytes(common.read_exact(stream, 2), "big", signed=False)
 			if debug:
 				print(f"Backreference (3-byte form) to 0x{table_index:>02x}")
-			literal = prev_literals[table_index]
-			if debug:
-				print(f"\t-> {literal}")
-			decompressed += literal
+			yield prev_literals[table_index]
 		elif byte in range(0x23, 0x4b):
 			# Backreference to a previous literal, 1-byte form.
 			# This can reference literals with indices in range(0x28).
 			table_index = byte - 0x23
-			i += 1
 			if debug:
 				print(f"Backreference (1-byte form) to 0x{table_index:>02x}")
-			literal = prev_literals[table_index]
-			if debug:
-				print(f"\t-> {literal}")
-			decompressed += literal
+			yield prev_literals[table_index]
 		elif byte in range(0x4b, 0xfe):
 			# Reference into a fixed table of two-byte literals.
 			# All compressed resources use the same table.
 			table_index = byte - 0x4b
-			i += 1
 			if debug:
 				print(f"Fixed table reference to 0x{table_index:>02x}")
-			entry = TABLE[table_index]
-			if debug:
-				print(f"\t-> {entry}")
-			decompressed += entry
+			yield TABLE[table_index]
 		elif byte == 0xfe:
 			# Extended code, whose meaning is controlled by the following byte.
 			
-			i += 1
-			kind = data[i]
+			(kind,) = common.read_exact(stream, 1)
 			if debug:
 				print(f"Extended code: 0x{kind:>02x}")
-			i += 1
 			
 			if kind == 0x00:
 				# Compact representation of (part of) a segment loader jump table, as used in 'CODE' (0) resources.
 				
 				if debug:
-					print(f"Segment loader jump table entries")
+					print("Segment loader jump table entries")
 				
 				# All generated jump table entries have the same segment number.
-				segment_number_int, length = common._read_variable_length_integer(data, i)
-				i += length
+				segment_number_int = common.read_variable_length_integer(stream)
 				if debug:
 					print(f"\t-> segment number: {segment_number_int:#x}")
 				
 				# The tail part of all jump table entries (i. e. everything except for the address).
-				entry_tail = b"?<" + segment_number_int.to_bytes(2, "big", signed=True) + b"\xa9\xf0"
-				if debug:
-					print(f"\t-> tail of first entry: {entry_tail}")
+				entry_tail = b"?<" + segment_number_int.to_bytes(2, "big", signed=False) + b"\xa9\xf0"
 				# The tail is output once *without* an address in front, i. e. the first entry's address must be generated manually by a previous code.
-				decompressed += entry_tail
+				yield entry_tail
 				
-				count, length = common._read_variable_length_integer(data, i)
-				i += length
+				count = common.read_variable_length_integer(stream)
 				if count <= 0:
 					raise common.DecompressError(f"Jump table entry count must be greater than 0, not {count}")
 				
 				# The second entry's address is stored explicitly.
-				current_int, length = common._read_variable_length_integer(data, i)
-				i += length
+				current_int = common.read_variable_length_integer(stream)
 				if debug:
-					print(f"-> address of second entry: {current_int:#x}")
-				entry = current_int.to_bytes(2, "big", signed=False) + entry_tail
-				if debug:
-					print(f"-> second entry: {entry}")
-				decompressed += entry
+					print(f"\t-> address of second entry: {current_int:#x}")
+				yield current_int.to_bytes(2, "big", signed=False) + entry_tail
 				
 				for _ in range(1, count):
 					# All further entries' addresses are stored as differences relative to the previous entry's address.
-					diff, length = common._read_variable_length_integer(data, i)
-					i += length
+					diff = common.read_variable_length_integer(stream)
 					# For some reason, each difference is 6 higher than it should be.
 					diff -= 6
 					
@@ -170,10 +142,7 @@ def decompress(data: bytes, decompressed_length: int, *, debug: bool=False) -> b
 					current_int = (current_int + diff) & 0xffff
 					if debug:
 						print(f"\t-> difference {diff:#x}: {current_int:#x}")
-					entry = current_int.to_bytes(2, "big", signed=False) + entry_tail
-					if debug:
-						print(f"\t-> {entry}")
-					decompressed += entry
+					yield current_int.to_bytes(2, "big", signed=False) + entry_tail
 			elif kind in (0x02, 0x03):
 				# Repeat 1 or 2 bytes a certain number of times.
 				
@@ -188,42 +157,36 @@ def decompress(data: bytes, decompressed_length: int, *, debug: bool=False) -> b
 					print(f"Repeat {byte_count}-byte value")
 				
 				# The byte(s) to repeat, stored as a variable-length integer. The value is treated as unsigned, i. e. the integer is never negative.
-				to_repeat_int, length = common._read_variable_length_integer(data, i)
-				i += length
+				to_repeat_int = common.read_variable_length_integer(stream)
 				try:
 					to_repeat = to_repeat_int.to_bytes(byte_count, "big", signed=False)
 				except OverflowError:
 					raise common.DecompressError(f"Value to repeat out of range for {byte_count}-byte repeat: {to_repeat_int:#x}")
 				
-				count_m1, length = common._read_variable_length_integer(data, i)
-				i += length
-				count = count_m1 + 1
+				count = common.read_variable_length_integer(stream) + 1
 				if count <= 0:
 					raise common.DecompressError(f"Repeat count must be positive: {count}")
 				
-				repeated = to_repeat * count
 				if debug:
-					print(f"\t-> {to_repeat} * {count}: {repeated}")
-				decompressed += repeated
+					print(f"\t-> {to_repeat!r} * {count}")
+				yield to_repeat * count
 			elif kind == 0x04:
 				# A sequence of 16-bit signed integers, with each integer encoded as a difference relative to the previous integer. The first integer is stored explicitly.
 				
 				if debug:
-					print(f"Difference-encoded 16-bit integers")
+					print("Difference-encoded 16-bit integers")
 				
 				# The first integer is stored explicitly, as a signed value.
-				initial_int, length = common._read_variable_length_integer(data, i)
-				i += length
+				initial_int = common.read_variable_length_integer(stream)
 				try:
 					initial = initial_int.to_bytes(2, "big", signed=True)
 				except OverflowError:
 					raise common.DecompressError(f"Initial value out of range for 16-bit integer difference encoding: {initial_int:#x}")
 				if debug:
-					print(f"\t-> initial: {initial}")
-				decompressed += initial
+					print(f"\t-> initial: 0x{initial_int:>04x}")
+				yield initial
 				
-				count, length = common._read_variable_length_integer(data, i)
-				i += length
+				count = common.read_variable_length_integer(stream)
 				if count < 0:
 					raise common.DecompressError(f"Count cannot be negative: {count}")
 				
@@ -232,64 +195,75 @@ def decompress(data: bytes, decompressed_length: int, *, debug: bool=False) -> b
 				for _ in range(count):
 					# The difference to the previous integer is stored as an 8-bit signed integer.
 					# The usual variable-length integer format is *not* used here.
-					diff = int.from_bytes(data[i:i+1], "big", signed=True)
-					i += 1
+					diff = int.from_bytes(common.read_exact(stream, 1), "big", signed=True)
 					
 					# Simulate 16-bit integer wraparound.
 					current_int = (current_int + diff) & 0xffff
-					current = current_int.to_bytes(2, "big", signed=False)
 					if debug:
-						print(f"\t-> difference {diff:#x}: {current}")
-					decompressed += current
+						print(f"\t-> difference {diff:#x}: 0x{current_int:>04x}")
+					yield current_int.to_bytes(2, "big", signed=False)
 			elif kind == 0x06:
 				# A sequence of 32-bit signed integers, with each integer encoded as a difference relative to the previous integer. The first integer is stored explicitly.
 				
 				if debug:
-					print(f"Difference-encoded 16-bit integers")
+					print("Difference-encoded 32-bit integers")
 				
 				# The first integer is stored explicitly, as a signed value.
-				initial_int, length = common._read_variable_length_integer(data, i)
-				i += length
+				initial_int = common.read_variable_length_integer(stream)
 				try:
 					initial = initial_int.to_bytes(4, "big", signed=True)
 				except OverflowError:
 					raise common.DecompressError(f"Initial value out of range for 32-bit integer difference encoding: {initial_int:#x}")
 				if debug:
-					print(f"\t-> initial: {initial}")
-				decompressed += initial
+					print(f"\t-> initial: 0x{initial_int:>08x}")
+				yield initial
 				
-				count, length = common._read_variable_length_integer(data, i)
-				i += length
+				count = common.read_variable_length_integer(stream)
 				assert count >= 0
 				
 				# To make the following calculations simpler, the signed initial_int value is converted to unsigned.
 				current_int = initial_int & 0xffffffff
 				for _ in range(count):
 					# The difference to the previous integer is stored as a variable-length integer, whose value may be negative.
-					diff, length = common._read_variable_length_integer(data, i)
-					i += length
+					diff = common.read_variable_length_integer(stream)
 					
 					# Simulate 32-bit integer wraparound.
 					current_int = (current_int + diff) & 0xffffffff
-					current = current_int.to_bytes(4, "big", signed=False)
 					if debug:
-						print(f"\t-> difference {diff:#x}: {current}")
-					decompressed += current
+						print(f"\t-> difference {diff:#x}: 0x{current_int:>08x}")
+					yield current_int.to_bytes(4, "big", signed=False)
 			else:
 				raise common.DecompressError(f"Unknown extended code: 0x{kind:>02x}")
 		elif byte == 0xff:
 			# End of data marker, always occurs exactly once as the last byte of the compressed data.
 			if debug:
 				print("End marker")
-			if i != len(data) - 1:
-				raise common.DecompressError(f"End marker reached at {i}, before the expected end of data at {len(data) - 1}")
-			i += 1
+			
+			# Check that there really is no more data left.
+			extra = stream.read(1)
+			if extra:
+				raise common.DecompressError(f"Extra data encountered after end of data marker (first extra byte: {extra!r})")
+			break
 		else:
-			raise common.DecompressError(f"Unknown tag byte: 0x{data[i]:>02x}")
+			raise common.DecompressError(f"Unknown tag byte: 0x{byte:>02x}")
+
+
+def decompress_stream(header_info: common.CompressedHeaderInfo, stream: typing.BinaryIO, *, debug: bool = False) -> typing.Iterator[bytes]:
+	"""Decompress compressed data in the format used by 'dcmp' (0)."""
 	
-	if decompressed_length % 2 != 0 and len(decompressed) == decompressed_length + 1:
-		# Special case: if the decompressed data length stored in the header is odd and one less than the length of the actual decompressed data, drop the last byte.
-		# This is necessary because nearly all codes generate data in groups of 2 or 4 bytes, so it is basically impossible to represent data with an odd length using this compression format.
-		decompressed = decompressed[:-1]
-	
-	return decompressed
+	decompressed_length = 0
+	for chunk in decompress_stream_inner(header_info, stream, debug=debug):
+		if debug:
+			print(f"\t-> {chunk!r}")
+		
+		if header_info.decompressed_length % 2 != 0 and decompressed_length + len(chunk) == header_info.decompressed_length + 1:
+			# Special case: if the decompressed data length stored in the header is odd and one less than the length of the actual decompressed data, drop the last byte.
+			# This is necessary because nearly all codes generate data in groups of 2 or 4 bytes, so it is basically impossible to represent data with an odd length using this compression format.
+			decompressed_length += len(chunk) - 1
+			yield chunk[:-1]
+		else:
+			decompressed_length += len(chunk)
+			yield chunk
+		
+		if debug:
+			print(f"Decompressed {decompressed_length:#x} bytes so far")

rsrcfork/compress/dcmp1.py

@@ -1,3 +1,5 @@
+import typing
+
 from . import common
 
 # Lookup table for codes in range(0xd5, 0xfe).
@@ -19,96 +21,75 @@ TABLE = [TABLE_DATA[i:i + 2] for i in range(0, len(TABLE_DATA), 2)]
 assert len(TABLE) == len(range(0xd5, 0xfe))
 
 
-def decompress(data: bytes, decompressed_length: int, *, debug: bool=False) -> bytes:
-	"""Decompress compressed data in the format used by 'dcmp' (1)."""
+def decompress_stream_inner(header_info: common.CompressedHeaderInfo, stream: typing.BinaryIO, *, debug: bool = False) -> typing.Iterator[bytes]:
+	"""Internal helper function, implements the main decompression algorithm. Only called from decompress_stream, which performs some extra checks and debug logging."""
 	
-	prev_literals = []
-	decompressed = b""
+	if not isinstance(header_info, common.CompressedType8HeaderInfo):
+		raise common.DecompressError(f"Incorrect header type: {type(header_info).__qualname__}")
 	
-	i = 0
+	prev_literals: typing.List[bytes] = []
 	
-	while i < len(data):
-		byte = data[i]
+	while True: # Loop is terminated when the EOF marker (0xff) is encountered
+		(byte,) = common.read_exact(stream, 1)
 		if debug:
-			print(f"Tag byte 0x{byte:>02x}, at 0x{i:x}, decompressing to 0x{len(decompressed):x}")
+			print(f"Tag byte 0x{byte:>02x}")
 		
 		if byte in range(0x00, 0x20):
 			# Literal byte sequence, 1-byte header.
 			# The length of the literal data is stored in the low nibble of the tag byte.
 			count = (byte >> 0 & 0xf) + 1
-			begin = i + 1
-			end = begin + count
 			# Controls whether or not the literal is stored so that it can be referenced again later.
 			do_store = byte >= 0x10
-			literal = data[begin:end]
+			literal = common.read_exact(stream, count)
 			if debug:
 				print(f"Literal (1-byte header, storing: {do_store})")
-				print(f"\t-> {literal}")
-			decompressed += literal
 			if do_store:
 				if debug:
-					print(f"\t-> stored as literal number 0x{len(prev_literals):x}")
+					print(f"\t-> storing as literal number 0x{len(prev_literals):x}")
 				prev_literals.append(literal)
-			i = end
+			yield literal
 		elif byte in range(0x20, 0xd0):
 			# Backreference to a previous literal, 1-byte form.
 			# This can reference literals with indices in range(0xb0).
 			table_index = byte - 0x20
-			i += 1
 			if debug:
 				print(f"Backreference (1-byte form) to 0x{table_index:>02x}")
-			literal = prev_literals[table_index]
-			if debug:
-				print(f"\t-> {literal}")
-			decompressed += literal
+			yield prev_literals[table_index]
 		elif byte in (0xd0, 0xd1):
 			# Literal byte sequence, 2-byte header.
 			# The length of the literal data is stored in the following byte.
-			count = data[i+1]
-			begin = i + 2
-			end = begin + count
+			(count,) = common.read_exact(stream, 1)
 			# Controls whether or not the literal is stored so that it can be referenced again later.
 			do_store = byte == 0xd1
-			literal = data[begin:end]
+			literal = common.read_exact(stream, count)
 			if debug:
 				print(f"Literal (2-byte header, storing: {do_store})")
-				print(f"\t-> {literal}")
-			decompressed += literal
 			if do_store:
 				if debug:
-					print(f"\t-> stored as literal number 0x{len(prev_literals):x}")
+					print(f"\t-> storing as literal number 0x{len(prev_literals):x}")
 				prev_literals.append(literal)
-			i = end
+			yield literal
 		elif byte == 0xd2:
 			# Backreference to a previous literal, 2-byte form.
 			# This can reference literals with indices in range(0xb0, 0x1b0).
-			table_index = data[i+1] + 0xb0
-			i += 2
+			(next_byte,) = common.read_exact(stream, 1)
+			table_index = next_byte + 0xb0
 			if debug:
 				print(f"Backreference (2-byte form) to 0x{table_index:>02x}")
-			literal = prev_literals[table_index]
-			if debug:
-				print(f"\t-> {literal}")
-			decompressed += literal
+			yield prev_literals[table_index]
 		elif byte in range(0xd5, 0xfe):
 			# Reference into a fixed table of two-byte literals.
 			# All compressed resources use the same table.
 			table_index = byte - 0xd5
-			i += 1
 			if debug:
 				print(f"Fixed table reference to 0x{table_index:>02x}")
-			entry = TABLE[table_index]
-			if debug:
-				print(f"\t-> {entry}")
-			decompressed += entry
+			yield TABLE[table_index]
 		elif byte == 0xfe:
 			# Extended code, whose meaning is controlled by the following byte.
 			
-			i += 1
-			kind = data[i]
+			(kind,) = common.read_exact(stream, 1)
 			if debug:
 				print(f"Extended code: 0x{kind:>02x}")
-			i += 1
 			
 			if kind == 0x02:
 				# Repeat 1 byte a certain number of times.
@@ -119,33 +100,45 @@ def decompress(data: bytes, decompressed_length: int, *, debug: bool=False) -> b
 					print(f"Repeat {byte_count}-byte value")
 				
 				# The byte(s) to repeat, stored as a variable-length integer. The value is treated as unsigned, i. e. the integer is never negative.
-				to_repeat_int, length = common._read_variable_length_integer(data, i)
-				i += length
+				to_repeat_int = common.read_variable_length_integer(stream)
 				try:
 					to_repeat = to_repeat_int.to_bytes(byte_count, "big", signed=False)
 				except OverflowError:
 					raise common.DecompressError(f"Value to repeat out of range for {byte_count}-byte repeat: {to_repeat_int:#x}")
 				
-				count_m1, length = common._read_variable_length_integer(data, i)
-				i += length
-				count = count_m1 + 1
+				count = common.read_variable_length_integer(stream) + 1
 				if count <= 0:
 					raise common.DecompressError(f"Repeat count must be positive: {count}")
 				
-				repeated = to_repeat * count
 				if debug:
-					print(f"\t-> {to_repeat} * {count}: {repeated}")
-				decompressed += repeated
+					print(f"\t-> {to_repeat!r} * {count}")
+				yield to_repeat * count
 			else:
 				raise common.DecompressError(f"Unknown extended code: 0x{kind:>02x}")
 		elif byte == 0xff:
 			# End of data marker, always occurs exactly once as the last byte of the compressed data.
 			if debug:
 				print("End marker")
-			if i != len(data) - 1:
-				raise common.DecompressError(f"End marker reached at {i}, before the expected end of data at {len(data) - 1}")
-			i += 1
+			
+			# Check that there really is no more data left.
+			extra = stream.read(1)
+			if extra:
+				raise common.DecompressError(f"Extra data encountered after end of data marker (first extra byte: {extra!r})")
+			break
 		else:
-			raise common.DecompressError(f"Unknown tag byte: 0x{data[i]:>02x}")
+			raise common.DecompressError(f"Unknown tag byte: 0x{byte:>02x}")
+
+
+def decompress_stream(header_info: common.CompressedHeaderInfo, stream: typing.BinaryIO, *, debug: bool = False) -> typing.Iterator[bytes]:
+	"""Decompress compressed data in the format used by 'dcmp' (1)."""
 	
-	return decompressed
+	decompressed_length = 0
+	for chunk in decompress_stream_inner(header_info, stream, debug=debug):
+		if debug:
+			print(f"\t-> {chunk!r}")
+		
+		decompressed_length += len(chunk)
+		yield chunk
+		
+		if debug:
+			print(f"Decompressed {decompressed_length:#x} bytes so far")

rsrcfork/compress/dcmp2.py

@@ -73,68 +73,73 @@ def _split_bits(i: int) -> typing.Tuple[bool, bool, bool, bool, bool, bool, bool
 	)
 
 
-def _decompress_system_untagged(data: bytes, decompressed_length: int, table: typing.Sequence[bytes], *, debug: bool=False) -> bytes:
-	parts = []
-	i = 0
-	while i < len(data):
-		if i == len(data) - 1 and decompressed_length % 2 != 0:
+def _decompress_untagged(stream: "common.PeekableIO", decompressed_length: int, table: typing.Sequence[bytes], *, debug: bool = False) -> typing.Iterator[bytes]:
+	while True: # Loop is terminated when EOF is reached.
+		table_index_data = stream.read(1)
+		if not table_index_data:
+			# End of compressed data.
+			break
+		elif not stream.peek(1) and decompressed_length % 2 != 0:
 			# Special case: if we are at the last byte of the compressed data, and the decompressed data has an odd length, the last byte is a single literal byte, and not a table reference.
 			if debug:
-				print(f"Last byte: {data[-1:]}")
-			parts.append(data[-1:])
+				print(f"Last byte: {table_index_data!r}")
+			yield table_index_data
 			break
 		
 		# Compressed data is untagged, every byte is a table reference.
+		(table_index,) = table_index_data
 		if debug:
-			print(f"Reference: {data[i]} -> {table[data[i]]}")
-		parts.append(table[data[i]])
-		i += 1
-	
-	return b"".join(parts)
+			print(f"Reference: {table_index} -> {table[table_index]!r}")
+		yield table[table_index]
 
-def _decompress_system_tagged(data: bytes, decompressed_length: int, table: typing.Sequence[bytes], *, debug: bool=False) -> bytes:
-	parts = []
-	i = 0
-	while i < len(data):
-		if i == len(data) - 1 and decompressed_length % 2 != 0:
+
+def _decompress_tagged(stream: "common.PeekableIO", decompressed_length: int, table: typing.Sequence[bytes], *, debug: bool = False) -> typing.Iterator[bytes]:
+	while True: # Loop is terminated when EOF is reached.
+		tag_data = stream.read(1)
+		if not tag_data:
+			# End of compressed data.
+			break
+		elif not stream.peek(1) and decompressed_length % 2 != 0:
 			# Special case: if we are at the last byte of the compressed data, and the decompressed data has an odd length, the last byte is a single literal byte, and not a tag or a table reference.
 			if debug:
-				print(f"Last byte: {data[-1:]}")
-			parts.append(data[-1:])
+				print(f"Last byte: {tag_data!r}")
+			yield tag_data
 			break
 		
 		# Compressed data is tagged, each tag byte is followed by 8 table references and/or literals.
-		tag = data[i]
+		(tag,) = tag_data
 		if debug:
 			print(f"Tag: 0b{tag:>08b}")
-		i += 1
 		for is_ref in _split_bits(tag):
 			if is_ref:
 				# This is a table reference (a single byte that is an index into the table).
+				table_index_data = stream.read(1)
+				if not table_index_data:
+					# End of compressed data.
+					break
+				(table_index,) = table_index_data
 				if debug:
-					print(f"Reference: {data[i]} -> {table[data[i]]}")
-				parts.append(table[data[i]])
-				i += 1
+					print(f"Reference: {table_index} -> {table[table_index]!r}")
+				yield table[table_index]
 			else:
 				# This is a literal (two uncompressed bytes that are literally copied into the output).
-				# Note: if i == len(data)-1, the literal is actually only a single byte long.
-				# This case is handled automatically - the slice extends one byte past the end of the data, and only one byte is returned.
+				literal = stream.read(2)
+				if not literal:
+					# End of compressed data.
+					break
+				# Note: the literal may be only a single byte long if it is located exactly at EOF. This is intended and expected - the 1-byte literal is yielded normally, and on the next iteration, decompression is terminated as EOF is detected.
 				if debug:
-					print(f"Literal: {data[i:i+2]}")
-				parts.append(data[i:i + 2])
-				i += 2
-			
-			# If the end of the compressed data is reached in the middle of a chunk, all further tag bits are ignored (they should be zero) and decompression ends.
-			if i >= len(data):
-				break
-	
-	return b"".join(parts)
+					print(f"Literal: {literal!r}")
+				yield literal
 
 
-def decompress(data: bytes, decompressed_length: int, parameters: bytes, *, debug: bool=False) -> bytes:
+def decompress_stream(header_info: common.CompressedHeaderInfo, stream: typing.BinaryIO, *, debug: bool = False) -> typing.Iterator[bytes]:
 	"""Decompress compressed data in the format used by 'dcmp' (2)."""
 	
-	unknown, table_count_m1, flags_raw = STRUCT_PARAMETERS.unpack(parameters)
+	if not isinstance(header_info, common.CompressedType9HeaderInfo):
+		raise common.DecompressError(f"Incorrect header type: {type(header_info).__qualname__}")
+	
+	unknown, table_count_m1, flags_raw = STRUCT_PARAMETERS.unpack(header_info.parameters)
 	
 	if debug:
 		print(f"Value of unknown parameter field: 0x{unknown:>04x}")
@@ -152,24 +157,21 @@ def decompress(data: bytes, decompressed_length: int, parameters: bytes, *, debu
 		print(f"Flags: {flags}")
 	
 	if ParameterFlags.CUSTOM_TABLE in flags:
-		table_start = 0
-		data_start = table_start + table_count * 2
 		table = []
-		for i in range(table_start, data_start, 2):
-			table.append(data[i:i + 2])
+		for _ in range(table_count):
+			table.append(common.read_exact(stream, 2))
 		if debug:
 			print(f"Using custom table: {table}")
 	else:
 		if table_count_m1 != 0:
 			raise common.DecompressError(f"table_count_m1 field is {table_count_m1}, but must be zero when the default table is used")
 		table = DEFAULT_TABLE
-		data_start = 0
 		if debug:
 			print("Using default table")
 	
 	if ParameterFlags.TAGGED in flags:
-		decompress_func = _decompress_system_tagged
+		decompress_func = _decompress_tagged
 	else:
-		decompress_func = _decompress_system_untagged
+		decompress_func = _decompress_untagged
 	
-	return decompress_func(data[data_start:], decompressed_length, table, debug=debug)
+	yield from decompress_func(common.make_peekable(stream), header_info.decompressed_length, table, debug=debug)

setup.cfg

@@ -18,8 +18,10 @@ classifiers =
 	Programming Language :: Python :: 3 :: Only
 	Programming Language :: Python :: 3.6
 	Programming Language :: Python :: 3.7
+	Programming Language :: Python :: 3.8
 license = MIT
-license_file = LICENSE
+license_files =
+	LICENSE
 description = A pure Python, cross-platform library/tool for reading Macintosh resource data, as stored in resource forks and ``.rsrc`` files
 long_description = file: README.rst
 long_description_content_type = text/x-rst
@@ -33,12 +35,55 @@ keywords =
 	macos
 
 [options]
-setup_requires =
-	setuptools>=39.2.0
+# mypy can only find type hints in the package if zip_safe is set to False,
+# see https://mypy.readthedocs.io/en/latest/installed_packages.html#making-pep-561-compatible-packages
+zip_safe = False
 python_requires = >=3.6
-packages =
+packages = find:
+
+[options.package_data]
+rsrcfork =
+	py.typed
+
+[options.packages.find]
+include =
 	rsrcfork
+	rsrcfork.*
 
 [options.entry_points]
 console_scripts =
 	rsrcfork = rsrcfork.__main__:main
+
+[flake8]
+extend-exclude =
+	.mypy_cache/,
+	build/,
+	dist/,
+
+# The following issues are ignored because they do not match our code style:
+ignore =
+	E226, # missing whitespace around arithmetic operator
+	E261, # at least two spaces before inline comment
+	E501, # line too long
+	W293, # blank line contains whitespace
+	W503, # line break before binary operator
+
+# flake8-tabs configuration
+use-flake8-tabs = true
+blank-lines-indent = always
+indent-tabs-def = 1
+
+[mypy]
+files=rsrcfork/**/*.py
+python_version = 3.6
+
+disallow_untyped_calls = True
+disallow_untyped_defs = True
+disallow_untyped_decorators = True
+
+no_implicit_optional = True
+
+warn_unused_ignores = True
+warn_unreachable = True
+
+warn_redundant_casts = True

tests/test_rsrcfork.py

@@ -0,0 +1,290 @@
+import collections
+import io
+import pathlib
+import shutil
+import sys
+import tempfile
+import typing
+import unittest
+
+import rsrcfork
+
+RESOURCE_FORKS_SUPPORTED = sys.platform.startswith("darwin")
+RESOURCE_FORKS_NOT_SUPPORTED_MESSAGE = "Resource forks are only supported on Mac"
+
+DATA_DIR = pathlib.Path(__file__).parent / "data"
+EMPTY_RSRC_FILE = DATA_DIR / "empty.rsrc"
+TEXTCLIPPING_RSRC_FILE = DATA_DIR / "unicode.textClipping.rsrc"
+TESTFILE_RSRC_FILE = DATA_DIR / "testfile.rsrc"
+
+COMPRESS_DATA_DIR = DATA_DIR / "compress"
+COMPRESSED_DIR = COMPRESS_DATA_DIR / "compressed"
+UNCOMPRESSED_DIR = COMPRESS_DATA_DIR / "uncompressed"
+COMPRESS_RSRC_FILE_NAMES = [
+	"Finder.rsrc",
+	"Finder Help.rsrc",
+	# "Install.rsrc", # Commented out for performance - this file contains a lot of small resources.
+	"System.rsrc",
+]
+
+
+def make_pascal_string(s):
+	return bytes([len(s)]) + s
+
+
+UNICODE_TEXT = "Here is some text, including Üñïçø∂é!"
+DRAG_DATA = (
+	b"\x00\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x03"
+	b"utxt\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00"
+	b"utf8\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00"
+	b"TEXT\x00\x00\x01\x00\x00\x00\x00\x00\x00\x00\x00\x00"
+)
+TEXTCLIPPING_RESOURCES = collections.OrderedDict([
+	(b"utxt", collections.OrderedDict([
+		(256, UNICODE_TEXT.encode("utf-16-be")),
+	])),
+	(b"utf8", collections.OrderedDict([
+		(256, UNICODE_TEXT.encode("utf-8")),
+	])),
+	(b"TEXT", collections.OrderedDict([
+		(256, UNICODE_TEXT.encode("macroman")),
+	])),
+	(b"drag", collections.OrderedDict([
+		(128, DRAG_DATA),
+	]))
+])
+
+TESTFILE_HEADER_SYSTEM_DATA = (
+	b"\xa7F$\x08 <\x00\x00\xab\x03\xa7F <\x00\x00"
+	b"\x01\x00\xb4\x88f\x06`\np\x00`\x06 <\x00\x00"
+	b"\x08testfile\x00\x02\x00\x02\x00rs"
+	b"rcRSED\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00"
+	b"\x02\x00rsrcRSED\x00\x00\x00\x00\x00\x00"
+	b"\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00"
+	b"\x00\x00\xdaIp~\x00\x00\x00\x00\x00\x00\x02.\xfe\x84"
+)
+TESTFILE_HEADER_APPLICATION_DATA = b"This is the application-specific header data section. Apparently I can write whatever nonsense I want here. A few more bytes...."
+TESTFILE_RESOURCES = collections.OrderedDict([
+	(b"STR ", collections.OrderedDict([
+		(128, (
+			None, rsrcfork.ResourceAttrs(0),
+			make_pascal_string(b"The String, without name or attributes"),
+		)),
+		(129, (
+			b"The Name", rsrcfork.ResourceAttrs(0),
+			make_pascal_string(b"The String, with name and no attributes"),
+		)),
+		(130, (
+			None, rsrcfork.ResourceAttrs.resProtected | rsrcfork.ResourceAttrs.resPreload,
+			make_pascal_string(b"The String, without name but with attributes"),
+		)),
+		(131, (
+			b"The Name with Attributes", rsrcfork.ResourceAttrs.resSysHeap,
+			make_pascal_string(b"The String, with both name and attributes"),
+		)),
+	])),
+])
+
+
+class UnseekableStreamWrapper(io.BufferedIOBase):
+	_wrapped: typing.BinaryIO
+	
+	def __init__(self, wrapped: typing.BinaryIO) -> None:
+		super().__init__()
+		
+		self._wrapped = wrapped
+	
+	def read(self, size: typing.Optional[int] = -1) -> bytes:
+		return self._wrapped.read(size)
+
+
+def open_resource_fork(path: pathlib.Path, mode: str) -> typing.BinaryIO:
+	return (path / "..namedfork" / "rsrc").open(mode)
+
+
+class ResourceFileReadTests(unittest.TestCase):
+	def test_empty(self) -> None:
+		with rsrcfork.open(EMPTY_RSRC_FILE, fork="data") as rf:
+			self.assertEqual(rf.header_system_data, bytes(112))
+			self.assertEqual(rf.header_application_data, bytes(128))
+			self.assertEqual(rf.file_attributes, rsrcfork.ResourceFileAttrs(0))
+			self.assertEqual(list(rf), [])
+	
+	def internal_test_textclipping(self, rf: rsrcfork.ResourceFile) -> None:
+		self.assertEqual(rf.header_system_data, bytes(112))
+		self.assertEqual(rf.header_application_data, bytes(128))
+		self.assertEqual(rf.file_attributes, rsrcfork.ResourceFileAttrs(0))
+		self.assertEqual(list(rf), list(TEXTCLIPPING_RESOURCES))
+		
+		for (actual_type, actual_reses), (expected_type, expected_reses) in zip(rf.items(), TEXTCLIPPING_RESOURCES.items()):
+			with self.subTest(type=expected_type):
+				self.assertEqual(actual_type, expected_type)
+				self.assertEqual(list(actual_reses), list(expected_reses))
+				
+				for (actual_id, actual_res), (expected_id, expected_data) in zip(actual_reses.items(), expected_reses.items()):
+					with self.subTest(id=expected_id):
+						self.assertEqual(actual_res.type, expected_type)
+						self.assertEqual(actual_id, expected_id)
+						self.assertEqual(actual_res.id, expected_id)
+						self.assertEqual(actual_res.name, None)
+						self.assertEqual(actual_res.attributes, rsrcfork.ResourceAttrs(0))
+						self.assertEqual(actual_res.data, expected_data)
+						self.assertEqual(actual_res.compressed_info, None)
+	
+	def test_textclipping_seekable_stream(self) -> None:
+		with TEXTCLIPPING_RSRC_FILE.open("rb") as f:
+			with rsrcfork.ResourceFile(f) as rf:
+				self.internal_test_textclipping(rf)
+	
+	def test_textclipping_unseekable_stream(self) -> None:
+		with TEXTCLIPPING_RSRC_FILE.open("rb") as f:
+			with UnseekableStreamWrapper(f) as usf:
+				with rsrcfork.ResourceFile(usf) as rf:
+					self.internal_test_textclipping(rf)
+	
+	def test_textclipping_path_data_fork(self) -> None:
+		with rsrcfork.open(TEXTCLIPPING_RSRC_FILE, fork="data") as rf:
+			self.internal_test_textclipping(rf)
+	
+	@unittest.skipUnless(RESOURCE_FORKS_SUPPORTED, RESOURCE_FORKS_NOT_SUPPORTED_MESSAGE)
+	def test_textclipping_path_resource_fork(self) -> None:
+		with tempfile.NamedTemporaryFile() as tempf:
+			with TEXTCLIPPING_RSRC_FILE.open("rb") as dataf:
+				with open_resource_fork(pathlib.Path(tempf.name), "wb") as rsrcf:
+					shutil.copyfileobj(dataf, rsrcf)
+			
+			with rsrcfork.open(tempf.name, fork="rsrc") as rf:
+				self.internal_test_textclipping(rf)
+	
+	@unittest.skipUnless(RESOURCE_FORKS_SUPPORTED, RESOURCE_FORKS_NOT_SUPPORTED_MESSAGE)
+	def test_textclipping_path_auto_resource_fork(self) -> None:
+		with tempfile.NamedTemporaryFile() as temp_data_fork:
+			with TEXTCLIPPING_RSRC_FILE.open("rb") as source_file:
+				with open_resource_fork(pathlib.Path(temp_data_fork.name), "wb") as temp_rsrc_fork:
+					shutil.copyfileobj(source_file, temp_rsrc_fork)
+			
+			with self.subTest(data_fork="empty"):
+				# Resource fork is selected when data fork is empty.
+				
+				with rsrcfork.open(temp_data_fork.name) as rf:
+					self.internal_test_textclipping(rf)
+			
+			with self.subTest(data_fork="non-resource data"):
+				# Resource fork is selected when data fork contains non-resource data.
+				
+				temp_data_fork.write(b"This is the file's data fork. It should not be read, as the file has a resource fork.")
+				
+				with rsrcfork.open(temp_data_fork.name) as rf:
+					self.internal_test_textclipping(rf)
+			
+			with self.subTest(data_fork="valid resource data"):
+				# Resource fork is selected even when data fork contains valid resource data.
+				
+				with EMPTY_RSRC_FILE.open("rb") as source_file:
+					shutil.copyfileobj(source_file, temp_data_fork)
+				
+				with rsrcfork.open(temp_data_fork.name) as rf:
+					self.internal_test_textclipping(rf)
+	
+	@unittest.skipUnless(RESOURCE_FORKS_SUPPORTED, RESOURCE_FORKS_NOT_SUPPORTED_MESSAGE)
+	def test_textclipping_path_auto_data_fork(self) -> None:
+		with tempfile.NamedTemporaryFile() as temp_data_fork:
+			with TEXTCLIPPING_RSRC_FILE.open("rb") as source_file:
+				shutil.copyfileobj(source_file, temp_data_fork)
+				# Have to flush the temporary file manually so that the data is visible to the other reads below.
+				# Normally this happens automatically as part of the close method, but that would also delete the temporary file, which we don't want.
+				temp_data_fork.flush()
+			
+			with self.subTest(rsrc_fork="nonexistant"):
+				# Data fork is selected when resource fork does not exist.
+				
+				with rsrcfork.open(temp_data_fork.name) as rf:
+					self.internal_test_textclipping(rf)
+			
+			with self.subTest(rsrc_fork="empty"):
+				# Data fork is selected when resource fork exists, but is empty.
+				
+				with open_resource_fork(pathlib.Path(temp_data_fork.name), "wb") as temp_rsrc_fork:
+					temp_rsrc_fork.write(b"")
+				
+				with rsrcfork.open(temp_data_fork.name) as rf:
+					self.internal_test_textclipping(rf)
+			
+			with self.subTest(rsrc_fork="non-resource data"):
+				# Data fork is selected when resource fork contains non-resource data.
+				
+				with open_resource_fork(pathlib.Path(temp_data_fork.name), "wb") as temp_rsrc_fork:
+					temp_rsrc_fork.write(b"This is the file's resource fork. It contains junk, so it should be ignored in favor of the data fork.")
+				
+				with rsrcfork.open(temp_data_fork.name) as rf:
+					self.internal_test_textclipping(rf)
+	
+	def test_testfile(self) -> None:
+		with rsrcfork.open(TESTFILE_RSRC_FILE, fork="data") as rf:
+			self.assertEqual(rf.header_system_data, TESTFILE_HEADER_SYSTEM_DATA)
+			self.assertEqual(rf.header_application_data, TESTFILE_HEADER_APPLICATION_DATA)
+			self.assertEqual(rf.file_attributes, rsrcfork.ResourceFileAttrs.mapPrinterDriverMultiFinderCompatible | rsrcfork.ResourceFileAttrs.mapReadOnly)
+			self.assertEqual(list(rf), list(TESTFILE_RESOURCES))
+			
+			for (actual_type, actual_reses), (expected_type, expected_reses) in zip(rf.items(), TESTFILE_RESOURCES.items()):
+				with self.subTest(type=expected_type):
+					self.assertEqual(actual_type, expected_type)
+					self.assertEqual(list(actual_reses), list(expected_reses))
+					
+					for (actual_id, actual_res), (expected_id, (expected_name, expected_attrs, expected_data)) in zip(actual_reses.items(), expected_reses.items()):
+						with self.subTest(id=expected_id):
+							self.assertEqual(actual_res.type, expected_type)
+							self.assertEqual(actual_id, expected_id)
+							self.assertEqual(actual_res.id, expected_id)
+							self.assertEqual(actual_res.name, expected_name)
+							self.assertEqual(actual_res.attributes, expected_attrs)
+							self.assertEqual(actual_res.data, expected_data)
+							self.assertEqual(actual_res.compressed_info, None)
+	
+	def test_compress_compare(self) -> None:
+		# This test goes through pairs of resource files: one original file with both compressed and uncompressed resources, and one modified file where all compressed resources have been decompressed (using ResEdit on System 7.5.5).
+		# It checks that the rsrcfork library performs automatic decompression on the compressed resources, so that the compressed resource file appears to the user like the uncompressed resource file (ignoring resource order, which was lost during decompression using ResEdit).
+		
+		for name in COMPRESS_RSRC_FILE_NAMES:
+			with self.subTest(name=name):
+				with rsrcfork.open(COMPRESSED_DIR / name, fork="data") as compressed_rf, rsrcfork.open(UNCOMPRESSED_DIR / name, fork="data") as uncompressed_rf:
+					self.assertEqual(sorted(compressed_rf), sorted(uncompressed_rf))
+					
+					for (compressed_type, compressed_reses), (uncompressed_type, uncompressed_reses) in zip(sorted(compressed_rf.items()), sorted(uncompressed_rf.items())):
+						with self.subTest(type=compressed_type):
+							self.assertEqual(compressed_type, uncompressed_type)
+							self.assertEqual(sorted(compressed_reses), sorted(uncompressed_reses))
+							
+							for (compressed_id, compressed_res), (uncompressed_id, uncompressed_res) in zip(sorted(compressed_reses.items()), sorted(uncompressed_reses.items())):
+								with self.subTest(id=compressed_id):
+									# The metadata of the compressed and uncompressed resources must match.
+									self.assertEqual(compressed_res.type, uncompressed_res.type)
+									self.assertEqual(compressed_id, uncompressed_id)
+									self.assertEqual(compressed_res.id, compressed_id)
+									self.assertEqual(compressed_res.id, uncompressed_res.id)
+									self.assertEqual(compressed_res.name, uncompressed_res.name)
+									self.assertEqual(compressed_res.attributes & ~rsrcfork.ResourceAttrs.resCompressed, uncompressed_res.attributes)
+									
+									# The uncompressed resource really has to be not compressed.
+									self.assertNotIn(rsrcfork.ResourceAttrs.resCompressed, uncompressed_res.attributes)
+									self.assertEqual(uncompressed_res.compressed_info, None)
+									self.assertEqual(uncompressed_res.data, uncompressed_res.data_raw)
+									self.assertEqual(uncompressed_res.length, uncompressed_res.length_raw)
+									
+									# The compressed resource's (automatically decompressed) data must match the uncompressed data.
+									self.assertEqual(compressed_res.data, uncompressed_res.data)
+									self.assertEqual(compressed_res.length, uncompressed_res.length)
+									
+									if rsrcfork.ResourceAttrs.resCompressed in compressed_res.attributes:
+										# Resources with the compressed attribute must expose correct compression metadata.
+										self.assertNotEqual(compressed_res.compressed_info, None)
+										self.assertEqual(compressed_res.compressed_info.decompressed_length, compressed_res.length)
+									else:
+										# Some resources in the "compressed" files are not actually compressed, in which case there is no compression metadata.
+										self.assertEqual(compressed_res.compressed_info, None)
+										self.assertEqual(compressed_res.data, compressed_res.data_raw)
+										self.assertEqual(compressed_res.length, compressed_res.length_raw)
+
+
+if __name__ == "__main__":
+	unittest.main()

tox.ini

@@ -0,0 +1,27 @@
+[tox]
+# When adding a new Python version here, please also update the list of Python versions called by the GitHub Actions workflow (.github/workflows/ci.yml).
+envlist = py{36,37,38},flake8,mypy,package
+
+[testenv]
+commands = python -m unittest discover --start-directory ./tests
+
+[testenv:flake8]
+deps =
+	flake8 >= 3.8.0
+	flake8-bugbear
+	flake8-tabs
+commands = flake8
+
+[testenv:mypy]
+deps =
+	mypy
+commands = mypy
+
+[testenv:package]
+deps =
+	twine
+	wheel >= 0.32.0
+
+commands =
+	python setup.py sdist bdist_wheel
+	twine check dist/*