Creating an import source database

The success or failure of an import performed by the Database Import Wizard is very often determined by the quality of the source database. Creating an accurate and powerful source database doesn’t require a computer science degree. You can create one in an application you probably already know, just by being prepared, being organized, and paying close attention to details. After you’ve created a working source database, you can reuse the basic structure of the file over and over again for future imports just by changing the data.

Note    The source database must include a field that contains the paths to the files to be imported so that the metadata and document content can be associated with the same document in the vault. Document metadata cannot be imported to existing documents in a vault.

Following are guidelines and suggestions for creating a source database for the Database Import Wizard:

• Choose the file format you know best. Practically any ODBC data file will work equally well with the Database Import Wizard; there are few advantages of any one over another. The important thing is to be able to easily create the necessary data structure in the file and type document metadata.

Tip    Importing from a Microsoft Excel worksheet is limited to 256 columns. If your source data includes more than 256 properties, choose a different file format.

If you regularly import documents from other departments, divisions, or external business partners, you might be able to get them to create the source database for you if you give them these guidelines, especially if they have the document data readily available and you do not. Alternatively, you might have to accept their document data in whatever format you can get from them and then convert or clean up the data yourself into a format usable by the Database Import Wizard.

• The source database must include fields with the following document metadata at a minimum, in order to successfully import documents into a vault:

Path: To the source file to be imported. Select this field for the Path field option in the Database Import Wizard, as described in Importing documents from a database. This path should be relative to the computer that will be running the Database Import Wizard. Use URLs to specify paths on other computers rather than mapped drive letters that may not exist or match. Any files not found in this location will not be imported, so this field is the most important of all. Confirm that every file exists before you start an import.

File name: In addition to the full path to the file in the preceding field. This value will be used to assign the document’s name in the vault. If you can, use a query inside your application to create this data for you if you don’t have it in a separate field already. Select this field to initialize the Name property in the Database Import Wizard, as described in Initializing document properties.

Revision number: This is another critical piece of metadata. Without it, Meridian will assign the first revision number calculated by the revision number options of the destination document type to your imported documents. This number may be lower than the actual revision number of the document. Without accurate revision numbers, you can never fully trust the documents to be accurate. For information on the fields required to correctly assign revision numbers, see Importing revisions.

Note    All revision numbers must be either a number or a letter. If the source database contains a mixture of numbers and letters in the same field, they must all be made one or the other. Otherwise, the data will not be imported.

Document type: This field is less critical than the others listed here because the document type of a document can be changed later, although you don’t want to have to change the type of many documents. Without this field, Meridian will assign the vault’s default document type to imported documents. This is a necessary field if you want to import a mix of document types in the same batch. For more information on the data required to assign document types, see Assigning document types to imported files.

• If you have additional metadata, or you can make it from other fields or sources, you will be glad if you import it so that it is available for reference and searching in Meridian. Additional data cannot be imported to existing documents later. Optional, but important, metadata fields include:

• Status: The workflow state of the documents being imported. Set this field to values that can be assigned to the Status property (Under Change, Released, Retired, In Review, and so on) as described in Initializing document properties.
• Created by/Modified by: These fields can either be operating system file data or data maintained in another data file. Either way, knowing who created and/or modified the document last will be valuable. Select these fields to initialize properties as described in Initializing document properties.

• If you have any other legacy metadata about the documents being imported, take time to carefully consider every piece. Any data that will help identify documents should be a candidate for import. Numbers, names, codes, document titles, departments, comments, sizes, and so on can all be valuable when it comes time to search for and revise the imported documents with confidence. Select these fields to initialize properties as described in Initializing document properties.

Tip    Consult a database expert within your organization to help you create the best possible source database. He or she may be able to save you a lot of time and effort and help produce a better result. When the documents are in the vault, the metadata will be more difficult to change.