Skip to main content

Fory v1.7.5 Released

· 3 min read
Shawn Yang
Shawn Yang
Apache Fory PMC Chair

Apache Fory 1.7.5 is now available. This release includes 15 pull requests from 4 contributors, with more control over JSON output, improved Scala and Kotlin JSON handling, and broader Python collection support. See Getting Started for installation instructions.

Highlights​

  • JSON output supports per-call pretty printing, non-ASCII escaping, configurable byte-array formats including Base16, and explicit default-value omission.
  • Scala and Kotlin JSON handling improves codec coverage and missing-field defaults, with additional GraalVM fixes for Java JSON accessors.
  • Python adds collection-subclass support and fixes named-tuple and temporal-array-buffer handling.

JSON Output Options​

Fory JSON is a Java JSON serialization framework with optional Kotlin and Scala modules. Applications can now choose pretty output for individual calls while reusing the same instance:

import java.util.Collections;
import org.apache.fory.json.ForyJson;

ForyJson json = ForyJson.builder().build();
String pretty = json.toPrettyJson(Collections.singletonMap("name", "Fory"));
byte[] prettyBytes = json.toPrettyJsonBytes(Collections.singletonMap("name", "Fory"));
String compact = json.toJson(Collections.singletonMap("name", "Fory"));

escapeNonAscii(true) writes non-ASCII string values and names as JSON Unicode escapes. The builder's byteArrayFormat selects Base64 strings, Base16 strings, or signed numeric arrays for ordinary byte arrays, including nested values. Property-level JsonByteArray annotations can override that default:

import org.apache.fory.json.ForyJson;
import org.apache.fory.json.annotation.JsonByteArray;

ForyJson hexJson = ForyJson.builder()
    .byteArrayFormat(JsonByteArray.Format.BASE16)
    .build();
String text = hexJson.toJson(new byte[] {1, -2, 3}); // "\"01fe03\""
byte[] bytes = hexJson.fromJson(text, byte[].class);

Default-value omission is available through explicit NON_DEFAULT policies on a property or its class with JsonProperty or JsonInclude. Applications must authorize only defaults that are stable, safe to evaluate, and restored consistently when a field is missing. Global NON_DEFAULT is rejected. Scala NON_EMPTY also recognizes None and supported empty strict collections and maps.

See JSON object mapping for output options and default omission for supported models and configuration examples.

Scala and Kotlin JSON Support​

Scala gains broader support for generic case classes, Unit, enumeration type tokens, and explicit string codecs for sealed hierarchies of singleton cases. Scala and Kotlin also refine missing-field defaults: Scala uses declared constructor defaults or type defaults; Kotlin uses declared defaults first, then zero or false for non-null numeric and Boolean parameters, or null for nullable parameters. Other non-null Kotlin parameters remain required.

See the Scala and Kotlin guides for precise type and null behavior. This release also fixes Java JSON accessor handling in GraalVM Native Image.

Python Collections and Buffers​

Xlang mode accepts Mapping, Sequence, and Set implementations, including mutable variants, and restores their values as built-in dict, list, and set objects. Python native mode preserves registered list, dict, and set subclasses together with their instance state.

Native mode also fixes named-tuple reconstruction and supports out-of-band buffers for NumPy datetime and timedelta arrays and Pandas datetime columns and indexes. See Python collection interfaces, native mode, and out-of-band serialization for examples.

The complete release notes are available on GitHub.

Full Changelog: v1.7.4...v1.7.5