Adapter SDK Schema

The Adapter SDK Schema is a representation of the data that an adapter retrieves from a repository. The schema defines a set of fields that the adapters are allowed to set.

This repository has code that auto generates schema definition libraries in multiple languages based on this schema descriptor.

These libraries provide classes that expose the schema fields to Adapter implementations.

The libraries also include builder classes to make it easy to set fields in custom Adapters. The builder classes have type safe methods to set or clear fields and a validate method to ensure all mandatory fields have been set.

Usage

Here are some ways to use the AdapterSdkSchema to get details about a Field

To use the Java version of the schema add the following dependencies:

<dependency>
  <groupId>io.github.fileanalysissuite.adaptersdk.schema</groupId>
  <artifactId>adaptersdk-schema-java-interfaces</artifactId>
</dependency>
<dependency>
  <groupId>io.github.fileanalysissuite.adaptersdk.schema</groupId>
  <artifactId>adaptersdk-schema-java-lib</artifactId>
</dependency>

• Access a field:

  AdapterSdkSchema.TITLE;

• Look up a field by its name:

  AdapterSdkSchema.getField("ADDRESS_CC");

• Check if the field is multivalued:

  AdapterSdkSchema.getField("ADDRESS_CC").isMultivalue();

• Check how the field's value is encoded, which could either by json or flattened:

  AdapterSdkSchema.getField("OCR").getObjectEncoding();

To use the .NET version of the schema add the MicroFocus.FAS.AdapterSdkSchema.dll package.

• Access a field:

  AdapterSdkSchema.TITLE;

• Look up a field by its name:

  AdapterSdkSchema.GetField("ADDRESS_CC");

• Check if the field is multivalued:

  AdapterSdkSchema.GetField("ADDRESS_CC").IsMultivalue;

• Check the field's type:

  AdapterSdkSchema.GetField("TITLE").FieldType;

Create a Java AdapterSdkSchemaObject using the AdapterSdkSchemaObjectBuilder

• Create a AdapterSdkSchemaObjectBuilder passing in an instance of a SchemaObjectBuilder

    final Map<String, Object> document = new HashMap<String, Object>();
    final AdapterSdkSchemaObjectBuilder documentBuilder
        = new AdapterSdkSchemaObjectBuilder(new MapSchemaObjectBuilder(document));

• Set field values by calling the appropriate setter methods

    documentBuilder.setFileName("test.doc");
    documentBuilder.setMimetype("msword");

    documentBuilder.setAddressDisplayBcc("person1@abc.com", "person2@abc.com");
    documentBuilder.setAddressDisplayTo("johnDoe@abc.com");

• Clear a field

    documentBuilder.clearAddressDisplayBcc();

• Add more values to a field

    documentBuilder.addAddressDisplayTo("janeDoe@abc.com");

• Check if all mandatory fields have been set and a valid AdapterSdkSchemaObject has been created

    documentBuilder.validate();

Create a .NET AdapterSdkSchemaObject using the AdapterSdkSchemaObjectBuilder

• Create a AdapterSdkSchemaObjectBuilder passing in an instance of a ISchemaObjectBuilder

    var document = new Dictionary<string, object>();
    AdapterSdkSchemaObjectBuilder documentBuilder = new(new DictionarySchemaObjectBuilder(document));

• Set field values by calling the appropriate setter methods

    documentBuilder.SetFileName("test.doc");
    documentBuilder.SetMimetype("msword");

    documentBuilder.SetAddressDisplayBcc("person1@abc.com", "person2@abc.com");
    documentBuilder.SetAddressDisplayTo("johnDoe@abc.com");

• Clear a field

    documentBuilder.ClearAddressDisplayBcc();

• Add more values to a field

    documentBuilder.AddAddressDisplayTo("janeDoe@abc.com");

• Add a nested field values

    documentBuilder.SetColumns(
        builder => {
            builder.SetName("ColumnA");
            builder.SetReviewedBy("somebody");
            builder.SetReviewerComments("LGTM");
            builder.SetSelectedGrammarRule("GrammarRuleA");
            builder.SetStatus("REVIEWED");
            builder.SetType("TypeA");
            builder.SetGrammarMatches(
                mbuilder => {
                    mbuilder.SetRules("RuleA-1", "RuleA-2");
                    mbuilder.SetWeight(50);
                }
            );
        },
        builder => {
            builder.SetName("ColumnB");
            builder.SetReviewedBy("somebody");
            builder.SetReviewerComments("Needs changes");
            builder.SetSelectedGrammarRule("GrammarRuleB");
            builder.SetStatus("REVIEWED");
            builder.SetType("TypeB");
            builder.SetGrammarMatches(
                mbuilder => {
                    mbuilder.SetRules("RuleB-1", "RuleB-2");
                    mbuilder.SetWeight(30);
                }
            );
        }
    );

• Set multiple values for single-dimensional multi-valued nested object

    documentBuilder.SetMetadataFiles(
        builder => {
            builder.SetContent("mdf-content1");
            builder.SetExtension("abc");
        },
        builder => {
            builder.SetContent("mdf-content2");
            builder.SetExtension("xyz");
        }
    );

• Set values for multi-dimensional multi-valued nested object

    documentBuilder.SetOcr(
        listBuilder => {
            listBuilder.Set(
                builder => {
                    builder.SetName("Reporter");
                    builder.SetType("Name");
                    builder.SetValue("John Doe");
                    builder.SetConfidence(41);
                },
                builder => {
                    builder.SetConfidence(42);
                    builder.SetValue("free form text from a non-templated region");
                }
            );
        },
        listBuilder => {
            listBuilder.Set(
                builder => {
                    builder.SetName("Reporter");
                    builder.SetType("Name");
                    builder.SetValue("Jane Doe");
                    builder.SetConfidence(41);
                },
                builder => {
                    builder.SetConfidence(44);
                    builder.SetValue("free form text from a non-templated region");
                }
            );
        }
    );

• Check if all mandatory fields have been set and a valid AdapterSdkSchemaObject has been created

    documentBuilder.Validate();

The SchemaObjectBuilder encodes field values as json or flattened

Here is an example of a json encoded field:

COLUMNS =
{
   "GRAMMAR_MATCHES":[
      {
         "RULES":[
            "RuleA-1",
            "RuleA-2"
         ],
         "WEIGHT":50
      }
   ],
   "NAME":"ColumnA",
   "REVIEWED_BY":"somebody",
   "REVIEWER_COMMENTS":"LGTM",
   "SELECTED_GRAMMAR_RULE":"GrammarRuleA",
   "STATUS":"REVIEWED",
   "TYPE":"TypeA"
},
{
   "GRAMMAR_MATCHES":[
      {
         "RULES":[
            "RuleB-1",
            "RuleB-2"
         ],
         "WEIGHT":30
      }
   ],
   "NAME":"ColumnB",
   "REVIEWED_BY":"somebody",
   "REVIEWER_COMMENTS":"Needs changes",
   "SELECTED_GRAMMAR_RULE":"GrammarRuleB",
   "STATUS":"REVIEWED",
   "TYPE":"TypeB"
}

Here is an example of a single dimensional flattened field:

METADATA_FILES_0_CONTENT = mdf-content1
METADATA_FILES_0_EXTENSION = abc
METADATA_FILES_1_CONTENT = mdf-content2
METADATA_FILES_1_EXTENSION = xyz

Here is an example of a two dimensional flattened field:

OCR_0_0_CONFIDENCE = 41
OCR_0_0_NAME = Reporter
OCR_0_0_TYPE = Name
OCR_0_0_VALUE = John Doe
OCR_0_1_CONFIDENCE = 42
OCR_0_1_VALUE = free form text from a non-templated region
OCR_1_0_CONFIDENCE = 41
OCR_1_0_NAME = Reporter
OCR_1_0_TYPE = Name
OCR_1_0_VALUE = Jane Doe
OCR_1_1_CONFIDENCE = 44
OCR_1_1_VALUE = free form text from a non-templated region