Home

GBIF Data Providers

How to

Connecting to a Database

Connecting to a Database

This is the most essential part of the installation because after the completion of it your data will be publicly available.

PHP Provider

The process of connecting a DiGIR installation to a database (resource in the DiGIR terminology) can be initiated through a button on the left-hand-side panel of the administration part of the provider. It comprises five steps:

1. Metadata: You have to define the metadata information of the resource, e.g. code, descriptive name of it, contact information, etc. Pay attention especially to the following fields:

   • Citation: Give the full name of the collection, if nothing else.

   • UseRestrictions: The standard text to enter here is "GBIF Data Use Agreement and GBIF Data Sharing Agreement apply." If you have additional requirements like "data is only available for non-profit non-commercial scientific research" you should also enter them here.

   • RecordIdentifier: This is basically a combination of the Code of the institution and another code for the collection. Your institution should assign the latter and publish them somewhere.

2. Data Source: You have to define the information regarding the connection to the database, i.e. the database name, the credentials for the connection to it, the type of your database and the driver that DiGIR should use in order to connect to it. You can find a thorough description for the values that you can use for a number of database in the Users' Guide manual of the package (one for Windows and one for Linux installations).

3. Tables: You have to define the table and the column of it, which provides a unique identifier for each row. In addition, in case that your data are divided in several tables you can join them so that the DiGIR software will be able to retrieve all the necessary information.

4. Filter: Additionally, filter(s) can be defined, which separate different resources that exist within the same database. For example, a possible scenario is that a database table contains information regarding N different collections that are identified by the collection_code column and where the same string appears to all rows of the same collection. Using the Filter feature you can define N different DiGIR resources, served by the same DiGIR installation, by simply setting a filter with the rule "column collection_code equals to text: collectionX", for instance.
   The advantages of this feature are that you can have only one database installation though you serve multiple collections (resources) and you can define dirrerent metadata information for each collection (resource) acknowledging the institutes/people that provide the data.

5. Mapping: You have to map the different elements of the Darwin Core schema (left-hand-side) to the columns of the table(s) that you defined at the above step. Only the first five elements are mandatory.

Python Provider

In the case of the Python provider you skip the 3, 4 and 5 steps above, since the apllication is doing the whole mapping automatically, but you have to upload your data creating a Report Document. The Report Documents can be either XML files formatted according to the schema http://provider.finsiel.ro/DC_DwC.xsd, or delimited text files exported from your databases or spreasheets.

The Users' guide of the GBIF Data Repository tool gives you the hints to proceed with the definition of Report Documents. In addition, you can find an example Report Document here.