Metadata templates store content that is later converted to EML. Most templating functions read a data object, to extract as much information as possible, then writes it to file for the user to validate the inferred content and add any missing info. Each function focuses on a data feature enabling a modular build of the metadata (not all data contain the same features). The current set of templating functions are (click function names for docs):
template_core_metadata() Describes
core information of a data package (abstract, methods, keywords,
personnel, license). Communicates what the data are, why and how they
were created, who was involved in their creation, and under what terms
the data may be used.template_table_attributes() Describes
columns of a data table (classes, units, datetime formats, missing value
codes).template_categorical_variables()
Describes categorical variables of a data table (if any columns are
classified as categorical in table attributes template).template_geographic_coverage()
Describes where the data were collected.template_taxonomic_coverage()
Describes biological organisms occuring in the data and helps resolve
them to authority systems. If matches can be made, then the full
taxonomic hierarchy of scientific and common names are automatically
rendered in the final EML metadata. This enables future users to search
on any taxonomic level of interest across data packages in
repositories.template_provenance() Describes source
datasets. Explicitly listing the DOIs and/or URLs of input data help
future users understand in greater detail how the derived data were
created and may some day be able to assign attribution to the creators
of referenced datasets.template_annotations() Adds semantic
meaning to metadata (variables, locations, persons, etc.) through links
to ontology terms. This enables greater human understanding and machine
actionability (linked data) and greatly improves the discoverability and
interoperability of data in general.NOTE: Data objects should be UTF-8 encoded so metadata extracted during the templating process will also be UTF-8, which is the required by the EML schema. Non-UTF-8 encoded data may result in metadata appearing as malformed character strings.