The content-connector-api defines the Java API that a content provider must implement in order to be included in OpenMinTeD. This API allows the OpenMinTeD platform to search for scientific publications, refine the results (based on a number of predefined facets) and then download the fulltext of the publications and also download the metadata of the publications in bulk mode.

The main class of the API is ContentConnector, which defined four methods (see the javadoc for more information):

• search(), which allows the users to perform queries on the metadata and get paged search results containing both a page of the metadata and the facets.
• fetchMetadata(), which returns an InputStream containing an xml file with the metadata of all the results of the query. The root element of the xml file is and its children are the metadata of the publications in the OMTD-SHARE schema.
• downloadFullText(), which returns the fulltext of a publication in an InputStream.

The two auxiliary classes that are used by the interaface are Query and SearchResult. The former describes a query in the metadata and contains the following information:

• keyword: the term for a generic, google like search.
• params: a list of extra constraints (e.g. publicationYear=2016, licence=CC-BY).
• facets: the list of metadata fields for which statistics will be returned.
• from and to: used for the paging of the results.

The latter class contains the results of a query:

• publications: a page of publications metadata.
• totalHits: the total number of results for the query
• from and to: the position of the page of results in the total results
• facets: statistics about the query results

In the Query class the search parameters do not belong in a well defined enumeration (coming in a next release) and the same applies for the possible facets that the user can ask for. The currently supported list of query parameters is the following:

• publicationtype the type of the publication (e.g. 'article', 'thesis'; see PublicationTypeEnum in OMTD-SHARE schema)
• publicationyear the year of the publication
• publisher: the publisher of the publication
• rights: the licence of the publication (see RightsStatementEnum in OMTD-SHARE schema)
• documentlanguage: the language of the publication (iso639-1 or iso639-3)
• documenttype: The type of fulltext of the publication (fulltext, abstract)
• keyword: one or more keyword characterising the publication.

In the next major version of the API we plan to include a REST API, so that content providers need not provide a Java implementation but they can provide a REST endpoint. Also, we'll define vocabularies and enumerations that will prevent the users for making illegal queries.