zms.repositoryutil

Mid-level: Build the complete set of file records for one model object from its ZODB representation.

Given a ZODB model object dict o and its pre-serialized init file content (from get_init_py or get_init_yaml), this function produces the full filename-to-metadata mapping that represents the object on disk. The mapping includes:

• The init file itself (__init__.py or __init__.yaml).
• Sibling resource files (templates, scripts, images) extracted from the object's attribute lists.
• A readme.md sibling if the object carries a readme field.

Does not perform filesystem I/O — it operates purely on in-memory data. Called by get_modelfileset_from_zodb for each model object.

Get class from py-string.

This function extracts the class definition from a given Python code string and executes it to return the defined class object. It uses regular expressions to identify the class name and then executes the code to obtain the class object.

Undocumented

Mid-level: Read a single file from disk and return a metadata record.

Delegates to read_file_from_disk for the actual I/O, then packages the result into a dict with id, filename (relative to base), data, and version (file modification time).

Called by get_modelfileset_from_disk.

Generate a Python class representation of a given object.

This function takes an object `o` (typically a dictionary) and generates a Python class definition as a list of strings. The generated class includes attributes and nested classes based on the structure and content of the input object.

Serialize a Python object into a YAML-formatted string.

This method processes a given object `o` and converts it into a YAML representation. It handles attributes and keys in the object, ensuring that non-serializable elements (e.g., Acquisition-Wrappers) are excluded.

High-level: Collect all model files from a filesystem repository path.

Recursively traverses the directory tree rooted at basepath looking for repository object folders (identified by __init__.py or __init__.yaml). For each folder found, the init file and all sibling resource files are read into metadata records.

Delegates to:

• get_file_from_disk — to read each individual file and build its metadata record {id, filename, data, version}.
• parse_modelfile — to extract the revision from the init file's parsed object definition.

High-level: Export model files from ZODB via a repository provider.

Queries the provider for repository objects stored in the ZODB, then serializes each object into a set of filesystem-ready file records (init file + sibling resource files).

This is the ZODB counterpart to get_modelfileset_from_disk; both return the same {filename: {{id, filename, data, version, ...}}} dict structure so that results can be compared with get_diffs.

Delegates to:

• provider.provideRepository(ids) — to retrieve objects from ZODB.
• get_init_py / get_init_yaml — to serialize objects into init file content (Python class or YAML).
• create_modelfileset — to build the filename-to-metadata mapping from the serialized init content and sibling resource objects.

High-level: Load and parse all model definitions from a filesystem repository path.

Traverses the directory tree rooted at basepath, reads each init file (__init__.py / __init__.yaml), parses it into a structured model definition, and resolves embedded attribute data from sibling resource files.

Unlike get_modelfileset_from_disk (which returns raw file records), this function returns fully parsed model objects keyed by object ID.

Delegates to:

• read_file_from_disk — low-level file I/O for init and readme files.
• parse_modelfile — to parse the init file contents into a model dict.
• value_data — to resolve embedded attribute data from sibling files.

Returns list of repository-providers.

These are collected by traversing the ZMS object tree starting from the document element and checking for objects that provide the IZMSRepositoryProvider interface. The traversal also includes objects that provide the IZMSConfigurationProvider interface, allowing for nested configurations that may contain additional repository providers.

Returns system conf-basepath.

Low-level: Parse in-memory model file contents into a model definition dict.

Interprets the data string (previously read by read_file_from_disk) according to the file extension:

• .yaml: parsed via yamlutil.parse into a dict keyed by object ID.
• .py: executed via get_class to extract the class __dict__.

Does not perform any filesystem I/O — it operates solely on the data parameter.

Called by get_modelfileset_from_disk and get_models_from_disk.

Low-level: Read raw file contents from the filesystem.

Opens the file at path/filename in binary mode and attempts UTF-8 decoding. Returns str if decodable, raw bytes otherwise, or None if the file does not exist.

This is the lowest-level I/O helper in the repository utilities. Called by get_file_from_disk, get_models_from_disk, value_data, and any other function needing raw file contents.

Helper function to resolve embedded attribute data from sibling resource files.

For a given attribute value vv (which may be a dict representing an object reference), this function checks if it contains an 'id' field that matches any of the sibling files in the same directory. If a match is found, it reads the corresponding file content and assigns it to the 'data' field of vv. This allows repository definitions to reference external resource files by ID, which are then loaded and included in the final model representation.