Copying Schemas

Use the Copy Schema From dialog to insert an existing schema into the Fields grid of an Edit Schema tab in the Properties view, or in an Edit Named Schema or Edit Table Schema dialog. Copy Schema From is also available in the Properties view for arcs when you select the Declare option.

Use the Export Schema to Clipboard command to copy an existing schema to the system clipboard in JSON (default) or XML format. This feature, discussed below, is designed to help you work with external applications that need to interact with StreamBase or LiveView streams, or to paste a schema definition into a StreamBase JUnit test file, if required.

Select a CSV or TSV file from the workspace or file system. For these two file types, Studio can also import from an archive file that contains exactly one CSV or TSV file. Supported archive formats are Zip, BZip2, and Gnu Zip.

CSV files must use the comma field separators, TSV files must use tab separators, and both files must use the double quote character when quotes are required.

When a valid file is selected, Studio shows a preview of the file's schema. If Studio finds an error in the file, it reports an error and does not continue.

The only File Option for these two file types is whether to read the file's header row to determine the field names of the incoming schema. Without the header row, Studio generates field_n placeholder names.

When you click OK, Studio attempts to infer data types from the first few column values; you can change the inferred data type later. Auto-detection of int, double, boolean, string, timestamp, and tuples are supported, but not lists or functions.

Studio supports hierarchical CSV and TSV files with subfields imported as the StreamBase tuple data type. For this to work, the file must have a header row, which you cannot skip. Edit the header row to specify all tuple fields and their sub-fields using dot notation. For example, the following CSV file specifies seven fields, of which the Contact field has Agent and Location sub-fields:

Product,Contact.Agent,Contact.Location,Stock,Value,Price,Category
"HEPA Air Purifier","Carlos Soltero,Northwest",515,30.94,21.78,Appliances

This imports with the following schema:

Select this option if you have a CSV or TSV file with non-standard features. With this file type, the File Options section becomes active, allowing you to specify Space or another character as field delimiter, or to specify single quotes when quotes are required, or no quote symbol.

For XML files, Studio attempts to infer data types from attribute values for elements.

For XSD files, Studio uses the XSD data type definitions to generate the schema. XSD files are used by several TIBCO products to save schema structure information. Use this feature to migrate schemas in XSD format from other TIBCO products.

Use the Copy Schema dialog with JSON files that specify a consistent schema, including those with nested arrays. The following excerpt shows a portion of a retail product record in JSON format:

The file that contains this excerpt is imported with the tuple field correctly recognized, but with the fields out of order compared to the JSON original:

If field order is important to your application, after import, you can use the up and down arrows in the field grid to rearrange the fields in the JSON file's order:

Instead of a JSON data file, you can select a JSON schema file. This has the benefit of the schema being translated exactly as written, with fields in order and having correct data types, which may not be possible with a JSON data file. Additionally, this can be used if there is no data present yet and only the schema is known.

Spotfire Binary Data File (SBDF) format files are written by one or more Spotfire products. The following table shows the mapping between SBDF and StreamBase data types:

Predictive Model Markup Language files are an XML-based predictive model interchange format.

Protocol Buffers (protobuf) are a binary data interchange format created by Google. Since protobuf data is binary, the protobuf schema file (ending in .proto) must be compiled first using the protoc command.

For example, if you have a protobuf file called company.proto with the following contents:

syntax = "proto3";

message Company {
    string Category = 1;
    string Product = 2;
    double Price = 3;
    double Value = 4;
    ProductManager ProductMgr = 5;
    int32 Stock = 6;
}

message ProductManager {
    string Contact = 1;
    string Location = 2;
}
              

You can run the Shell command protoc --descriptor_set_out=company.desc company.proto to compile company.proto into company.desc, which can be imported by this wizard. After import, you will see the schema converted to StreamBase tuple format.

Copy from the schema of one of the streams emitted by the global system container that runs in every EventFlow engine process. For example, you can copy the schema of the standard StreamBase error tuple by selecting error:schema from the dropdown list after selecting the From system container schemas radio button.

Copy from the schema of any component in any EventFlow application or interface in any project in the current Studio workspace. You can copy the schemas from the named schemas defined in the current module, or from any intermediate stream in any module.