Merge pull request #115 from ArcanaFramework/vendor-splitout

split out microsoft vendor types to fileformats.vendor namespace

Author: tclose

.github/workflows/ci-cd.yml

@@ -46,7 +46,7 @@ jobs:
     - name: Install Package
       run: python3 -m pip install -e .[test]
     - name: Install Extras Package
-      run: python3 -m pip install -e ./extras[test]
+      run: python3 -m pip install -e ./extras[test,application,image,vendor_openxmlformats_officedocument]
     - name: MyPy
       run: mypy --install-types --non-interactive --no-warn-unused-ignores .
     - name: Pytest

extras/fileformats/extras/vendor/openxmlformats_officedocument/application.py

@@ -0,0 +1,19 @@
+import typing as ty
+
+from docx import Document
+from docx.document import Document as DocumentObject
+
+from fileformats.core import FileSet, extra_implementation
+from fileformats.vendor.openxmlformats_officedocument.application import (
+    Wordprocessingml_Document as MswordX,
+)
+
+
+@extra_implementation(FileSet.load)
+def load_docx(doc: MswordX, **kwargs: ty.Any) -> DocumentObject:
+    return Document(str(doc))  # type: ignore[no-any-return]
+
+
+@extra_implementation(FileSet.save)
+def save_docx(doc: MswordX, data: DocumentObject, **kwargs: ty.Any) -> None:
+    data.save(str(doc))

extras/pyproject.toml

@@ -9,11 +9,7 @@ readme = "README.rst"
 requires-python = ">=3.11"
 dependencies = [
     "fileformats",
-    "imageio >=2.24.0",
     "pydra >=1.0a",
-    "PyYAML>=6.0",
-    "pydicom >=2.3",
-    "medimages4tests",
 ]
 license = { file = "LICENSE" }
 authors = [{ name = "Thomas G. Close", email = "[email protected]" }]
@@ -43,6 +39,18 @@ test = [
     "codecov",
     "medimages4tests",
 ]
+application = [
+    "PyYAML>=6.0",
+    "pydicom >=2.3",
+    "medimages4tests",
+]
+image = [
+    "imageio >=2.24.0",
+]
+vendor_openxmlformats_officedocument = [
+    "python-docx >= 1.2.0",
+]
+
 
 [project.urls]
 repository = "https://github.com/ArcanaFramework/fileformats"

fileformats/application/__init__.py

@@ -1,7 +1,7 @@
 from fileformats.core import __version__
 
 from .archive import Archive, Bzip, Gzip, Tar, TarGzip, Zip
-from .document import Document, Msword, MswordX, Pdf, Postscript
+from .document import Document, Msword, Pdf, Postscript
 from .medical import Dicom
 from .misc import (
     H224,
@@ -492,6 +492,7 @@
     _3gpphalforms__Json,
     _3gppIms__Xml,
 )
+from .presentation import Presentation
 from .serialization import (
     InformalSchema,
     Json,
@@ -503,8 +504,12 @@
     XmlSchema,
     Yaml,
 )
+from .spreadsheet import Spreadsheet
 
 from fileformats.text import Javascript  # isort: skip
+from fileformats.vendor.openxmlformats_officedocument.application import (  # isort:skip
+    Wordprocessingml_Document as MswordX,
+)
 
 __all__ = [
     "__version__",
@@ -833,6 +838,7 @@
     "Pls__Xml",
     "PocSettings__Xml",
     "PpspTracker__Json",
+    "Presentation",
     "Problem__Json",
     "Problem__Xml",
     "Provenance__Xml",
@@ -922,6 +928,7 @@
     "Spdx__Json",
     "SparqlResults__Xml",
     "SpiritsEvent__Xml",
+    "Spreadsheet",
     "Sql",
     "Srgs",
     "Srgs__Xml",

fileformats/application/document.py

@@ -1,8 +1,6 @@
 from fileformats.core.mixin import WithMagicNumber
 from fileformats.generic import BinaryFile
 
-from .archive import Zip
-
 
 # Document formats
 class Document(BinaryFile):
@@ -25,16 +23,11 @@ class Msword(Document):
     iana_mime = "application/msword"
 
 
-class MswordX(Zip, Document):
-    ext = ".docx"
-
-    iana_mime = (
-        "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
-    )
-
-
 class Postscript(WithMagicNumber, Document):
     ext = ".eps"
     alternate_exts = (".ps",)
     magic_number = b"%!"
     iana_mime = "application/postscript"
+
+
+__all__ = ["Document", "Pdf", "Msword", "Postscript"]

fileformats/application/presentation.py

@@ -0,0 +1,7 @@
+from fileformats.generic import BinaryFile
+
+
+# Document formats
+class Presentation(BinaryFile):
+    # iana_mime = None
+    pass

fileformats/application/spreadsheet.py

@@ -0,0 +1,7 @@
+from fileformats.generic import BinaryFile
+
+
+# Document formats
+class Spreadsheet(BinaryFile):
+    # iana_mime = None
+    pass

fileformats/core/classifier.py

@@ -20,10 +20,26 @@ def namespace(cls) -> ty.Optional[str]:
         module_parts = cls.__module__.split(".")
         if module_parts[0] != "fileformats":
             raise FormatDefinitionError(
-                f"Cannot create reversible MIME type for {cls} as it is not in the "
-                "fileformats namespace"
+                f"Cannot determine namespace for {cls} format as it is not in the "
+                "fileformats namespace package"
             )
-        return module_parts[1].replace("_", "-")
+        namespace = module_parts[1]
+        if namespace == "vendor":
+            if len(module_parts) < 4:
+                raise FormatDefinitionError(
+                    f"Cannot determine namespace for vendor-specific format, {cls} it needs "
+                    "to be in a subpackage of the form `fileformats.vendor.<vendor-name>.<namespace>`,"
+                    f"found `{'.'.join(module_parts)}`"
+                )
+            namespace = module_parts[3]
+        return namespace.replace("_", "-")
+
+    @classproperty
+    def vendor(cls) -> ty.Optional[str]:
+        module_parts = cls.__module__.split(".")
+        if module_parts[0] != "fileformats" or module_parts[1] != "vendor":
+            return None
+        return module_parts[2].replace("_", "-")
 
     def dummy(self) -> float:
 

fileformats/core/datatype.py

@@ -117,7 +117,11 @@ def mime_like(cls) -> str:
         return a MIME-like identifier, e.g. "text/plain" for fileformats.text.Plain.
         and "medimage/nifti" for fileformats.medimage.Nifti.
         """
-        return f"{cls.namespace}/{to_mime_format_name(cls.__name__)}"  # type: ignore
+        mime_like: str = cls.namespace + "/"
+        if cls.vendor:
+            mime_like += "vnd." + cls.vendor + "."
+        mime_like += to_mime_format_name(cls.__name__)  # type: ignore[attr-defined]
+        return mime_like
 
     @classmethod
     def from_mime(cls, mime_string: str) -> ty.Type[DataType]:
@@ -196,16 +200,25 @@ def from_mime(cls, mime_string: str) -> ty.Type[DataType]:
             else:
                 klass = next(iter(matching_name))
         else:
-            class_name = from_mime_format_name(format_name)
+            # Get the path to the module to load the class from
+            if format_name.startswith("vnd."):
+                name_parts = format_name.split(".")
+                vendor = name_parts[1]
+                format_name = ".".join(name_parts[2:])
+                module_path = f"fileformats.vendor.{vendor}.{namespace}"
+            else:
+                module_path = f"fileformats.{namespace}"
+            module_path = module_path.replace("-", "_")
             try:
-                module = importlib.import_module("fileformats." + namespace)
+                module = importlib.import_module(module_path)
             except ImportError:
                 raise FormatRecognitionError(
-                    f"Did not find fileformats namespace package corresponding to {namespace} "
+                    f"Did not find fileformats namespace package at '{module_path}' "
                     f"required to interpret '{mime_string}' MIME, or MIME-like, type. "
                     f"try installing the namespace package with "
                     f"'python3 -m pip install fileformats-{namespace}'."
                 ) from None
+            class_name = from_mime_format_name(format_name)
             try:
                 klass = getattr(module, class_name)
             except AttributeError:

fileformats/core/tests/test_classifiers.py

@@ -1,28 +1,31 @@
 from __future__ import annotations
+
 import decimal
+
 import pytest
 from pydra.compose import python
-from fileformats.core import from_mime, DataType, FileSet
-from fileformats.core import converter
+
 from fileformats.application import Zip
-from fileformats.generic import DirectoryOf
-from fileformats.field import Array, Integer, Decimal, Text, Boolean
+from fileformats.core import DataType, FileSet, converter, from_mime
 from fileformats.core.exceptions import (
-    FormatDefinitionError,
     FormatConversionError,
-    FormatRecognitionError,
+    FormatDefinitionError,
     FormatMismatchError,
+    FormatRecognitionError,
 )
+from fileformats.field import Array, Boolean, Decimal, Integer, Text
+from fileformats.generic import DirectoryOf
+from fileformats.testing import J  # Y,
 from fileformats.testing import (
     A,
     B,
     C,
+    Classified,
     D,
     E,
     F,
     G,
     H,
-    J,
     K,
     L,
     M,
@@ -31,16 +34,13 @@
     Q,
     R,
     TestField,
-    Classified,
     U,
     V,
     W,
     X,
-    # Y,
     Z,
 )
 
-
 SpecificDataType = DataType.type_var("SpecificDataType")
 SpecificFileSet = FileSet.type_var("SpecificFileSet")
 
@@ -242,9 +242,28 @@ def test_mime_roundtrips():
     assert from_mime("testing/b.a+k") is K[B, A]
     assert from_mime("testing/b.a+k") is not K[A, B]
 
-    with pytest.raises(FormatRecognitionError) as e:
+    with pytest.raises(
+        FormatRecognitionError, match="Cannot create reversible MIME type"
+    ):
         Array[TestField].mime_like
-    assert "Cannot create reversible MIME type for " in str(e)
+
+
+def test_mime_fail():
+    class BadFormat(DataType):
+        pass
+
+    with pytest.raises(FormatDefinitionError, match="Cannot determine namespace"):
+        BadFormat.namespace
+
+
+def test_vendor_mime_fail():
+    class BadVendorFormat(DataType):
+        pass
+
+    BadVendorFormat.__module__ = "fileformats.vendor.badnamespace"
+
+    with pytest.raises(FormatDefinitionError, match="Cannot determine namespace"):
+        BadVendorFormat.namespace
 
 
 def test_inherited_classifiers():
@@ -430,3 +449,4 @@ def test_classifier_categories6():
         match="Cannot have more than one occurrence of a classifier ",
     ):
         Classified[C, E]
+        Classified[C, E]