Before the update, the MDTF was grabbing necessary information from a set filename template (i.e):
self.filename_template = ('{realm}.{time_range}.{variable_id}.{frequency}.nc' )
In an update, I converted this function into a base class (ppParserBase) that can be spun out depending on what DRS the user is using (i.e. ppParserGFDL) . Currently, the MDTF is acting in the former manner as it is implemented with the line:
entries = [e.cat_entry for e in list(map(util.catalog.ppParser['ppParser' + 'GFDL'], file_list))]
A final form of the refactor needs to be decided with templates for the varying DRS's. In the best case, this type of filename parsing might be able to be bypassed entirely. For example, the following might be needed to be included in the .jsonc file passed to the MDTF by the command-line:
"filename_template" : "{realm}.{time_range}.{variable_id}.{frequency}.nc"
jtmims
commented
5 months ago
Before the update, the MDTF was grabbing necessary information from a set filename template (i.e):
self.filename_template = ('{realm}.{time_range}.{variable_id}.{frequency}.nc' )In an update, I converted this function into a base class (ppParserBase) that can be spun out depending on what DRS the user is using (i.e. ppParserGFDL) . Currently, the MDTF is acting in the former manner as it is implemented with the line:
entries = [e.cat_entry for e in list(map(util.catalog.ppParser['ppParser' + 'GFDL'], file_list))]A final form of the refactor needs to be decided with templates for the varying DRS's. In the best case, this type of filename parsing might be able to be bypassed entirely. For example, the following might be needed to be included in the .jsonc file passed to the MDTF by the command-line:
"filename_template" : "{realm}.{time_range}.{variable_id}.{frequency}.nc"