A Python implementation of the Open Packaging Conventions (OPC).
ECMA 376 Part 2 defines the “Open Packaging Conventions”, which is the packaging format to be used by the Office Open XML file formats. It specifies, how to represent multiple logical files (“Parts”) within a physical Package (as a ZIP container), how to express semantic relationships between those Parts (using accompanying XML Parts), and how to add meta data and cryptographic signatures to the Package. The format is defined in two steps: an abstract logic package model with Parts, Content Types and Relationships, and a physical mapping of this package model to PKZIP files.
This Python package aims to implement both, the logical model and physical mapping of OPC package files, to allow reading and writing such files. However, it does not provide functionality to deal with the packages' payload, i.e. there is not functionality included to parse MS Word Documents from .docx files etc.
-
reading OPC package files
- listing contained Parts (incl. Content Type)
- reading Parts as file-like objects (incl. interleaved Parts)
- parsing and following Relationships
- parsing package meta data (“Core Properties”)
-
writing OPC package files
- creating and writing Parts (via writable file-like objects, incl. interleaved Parts)
- adding Relationships (as simple Python objects)
- adding Content Type information
- composing and writing package meta data (“Core Properties”)
Modifying packages in-place is not supported.
- reading/verifying/creating cryptographic signatures
This package requires lxml
for XML reading and writing (with proper XML namespaces support).
Apart from that only the Python standard library is required.
The Python interpreter must support Python 3.6 or higher.
Short example of reading an OPC package file:
import pyecma376_2
with pyecma376_2.ZipPackageReader("document.docx") as reader:
# List parts in package
for part_name, content_type in reader.list_parts():
print(part_name)
# Get Relationship of type "…/officeDocument" from package-level Relationships
document_part_name = reader.get_related_parts_by_type("/")[
'http://schemas.openxmlformats.org/officeDocument/2006/relationships/officeDocument'][0]
# Read core properties (package meta data)
core_props = reader.get_core_properties()
print(core_props.creator)
# Open part as (binary) file-like object
with reader.open_part(document_part_name) as part:
# XML parsing and document interpretation goes here
print(part.read().decode())
Short example of creating and writing into an OPC package file:
import pyecma376_2
import datetime
with pyecma376_2.ZipPackageWriter("new_document.myx") as writer:
# Add a part
with writer.open_part("/example/document.txt", "text/plain") as part:
part.write("Lorem ipsum dolor sit amet.".encode())
# Write core properties (meta data)
# To make those work, we need to add the RELATIONSHIP_TYPE_CORE_PROPERTIES relationship below.
cp = pyecma376_2.OPCCoreProperties()
cp.created = datetime.datetime.now()
with writer.open_part(pyecma376_2.DEFAULT_CORE_PROPERTIES_NAME, "application/xml") as part:
cp.write_xml(part)
# Write the packages root relationships
writer.write_relationships([
pyecma376_2.OPCRelationship("r1", "http://example.com/my-package-relationship-id", "http://example.com",
pyecma376_2.OPCTargetMode.EXTERNAL),
pyecma376_2.OPCRelationship("r2", "http://example.com/my-document-rel", "example/document.txt",
pyecma376_2.OPCTargetMode.INTERNAL),
pyecma376_2.OPCRelationship("r3", pyecma376_2.RELATIONSHIP_TYPE_CORE_PROPERTIES,
pyecma376_2.DEFAULT_CORE_PROPERTIES_NAME,
pyecma376_2.OPCTargetMode.INTERNAL),
])
# The Content Types Stream with all parts' ContentTypes is automatically added when closing the package
# Modify `writer.content_types` to change Content Types representation and use `writer.write_content_types_stream()`
# for premature serialization/output.
The architecture of this package follows the logical concept of the ECMA standard:
The package_model
module defines abstract OPCPackageReader
and OPCPackageWriter
classes that implement all the logical package model functionality, but omit the physical mapping to ZIP files.
This mapping is reflected in the abstract methods list_items()
, open_item()
and create_item()
which are then implemented by the ZipPackageReader
and ZipPackageWriter
classes from the zip_package
module.
Auxiliary classes and functions like OPCRelationship
, part_realpath
and normalize_part_name
are also contained in the package_model
module.
This package has been developed by Michael Thies at the Chair of Information and Automation Systems for Process and Material Technology (PLT) at RWTH Aachen University.
It is published under the terms of Apache License v2. See LICENSE and NOTICE files for details.