medali stands for MEtaDAta LIbrary and is responsible for parsing, tagging and interpreting product metadata.
medali should be a place to establish version-controlled metadata definitions to guarantee homogeneous reading and writing of metadata within the scope of one product. It provides one module, the core module containing the class MetaData. This class offers a simple interface to actual metadata items, which are defined with respect to reference metadata. Such reference metadata can either be prepared manually as a dictionary or can be set in a configuration file. For mature products, the configuration files are stored in the folder in "src/medali/lib" under a product ID and a metadata version.
The Metadata class offers the following functionality:
- initialisation via a metadata dictionary (+ optionally a reference metadata dictionary), via a product ID and metadata version combination, or via a configuration file path
- setting and encoding (from a given data type to a string) the given metadata items
- returning and decoding (from a string to a given data type) the given metadata items
- checking if given metadata meets the criteria defined in the reference metadata (e.g., if the value is within a list of allowed values or if the value contains a specific string, ...)
medali is a very light-weight package and has no dependencies. It can be installed via pip:
pip install medaliA template configuration file for creating new reference metadata for a product can be found in the "templates" folder. The file should be named according to the metadata version and should be placed in a sub-folder in "src/medali/lib" named after the product ID. The reference metadata should have two sections:
- "Metadata": all needed tags and their data type (currently supported: string, boolean, datetime, integer, number)
- "Expected_value": should list metadata items that need to meet some specific criteria