hibernate/hibernate-orm · error · IllegalArgumentException

Unrecognized schema generation source type: '%s'

Error message

Unrecognized schema generation source type: '%s'

What it means

The schema generation source type, interpreted by SourceType (used for jakarta.persistence.schema-generation create-source/drop-source style settings), must name one of metadata, script, metadata-then-script, or script-then-metadata. Hyphens are normalized to underscores and an empty value defaults to METADATA; every other value throws this IllegalArgumentException.

Source

Thrown at hibernate-core/src/main/java/org/hibernate/tool/schema/SourceType.java:85

	public static SourceType interpret(Object value, SourceType defaultValue) {
		if ( value == null ) {
			return defaultValue;
		}

		if ( value instanceof SourceType sourceType ) {
			return sourceType;
		}

		final String name = value.toString().trim().replace('-', '_');
		if ( name.isEmpty() ) {
			return METADATA;
		}
		for ( var sourceType: values() ) {
			if ( sourceType.toString().equalsIgnoreCase(name) ) {
				return sourceType;
			}
		}
		throw new IllegalArgumentException( "Unrecognized schema generation source type: '" + value + "'");
	}
}

View on GitHub (pinned to fad1729dce)

Solutions

1. Use one of the exact values: metadata, script, metadata-then-script, script-then-metadata.
2. Omit the property to fall back to the default (metadata-only generation).

Example fix

# before
<property name="jakarta.persistence.schema-generation.create-source" value="database"/>

# after
<property name="jakarta.persistence.schema-generation.create-source" value="metadata-then-script"/>

private static final Set<String> VALID_SOURCES = Set.of(
        "metadata", "script", "metadata-then-script", "script-then-metadata");

String token = source == null ? "" : source.trim().toLowerCase(Locale.ROOT);
if (!token.isEmpty() && !VALID_SOURCES.contains(token)) {
    throw new IllegalStateException("Bad schema generation source type: " + source);
}

When it happens

Trigger: SourceType.interpret receiving "database", "scripts", "both", or a misspelled token: no enum constant's toString matches case-insensitively after normalization, so interpretation fails.

Common situations: Confusing the source type with the schema action; hand-copying JPA schema-generation properties with wrong tokens; unclear which of script/metadata ordering tokens the provider expects.

Related errors

• Unrecognized JPA schema management action setting: '%s'
• Duplicate generator name '%s'; you will likely want to set t
• Duplicate generator name %s; you will likely want to set the
• Can't load class: {}
• The provided {} setting value [{}] is not supported

AI-assisted analysis of hibernate/hibernate-orm@fad1729dce (2026-08-22). Data as JSON: /api/errors/1681ca09fe617374. Report an issue: GitHub.