private void writeRdfs()

in src/main/java/com/amazonaws/services/neptune/profiles/neptune_ml/v2/RdfTrainingDataConfigWriter.java [69:187]
94 lines of code
17 McCabe index (conditional complexity)

    private void writeRdfs() throws IOException {
        generator.writeArrayFieldStart("rdfs");

        Collection<LabelConfigV2> classificationSpecifications = config.nodeConfig().getAllClassificationSpecifications();

        if (classificationSpecifications.isEmpty()) {
            for (String filename : filenames) {
                generator.writeStartObject();
                generator.writeStringField("file_name", filename);

                generator.writeObjectFieldStart("label");
                generator.writeStringField("task_type", EdgeTaskTypeV2.link_prediction.name());

                generator.writeArrayFieldStart("targets");

                generator.writeStartObject();
                generator.writeArrayFieldStart("split_rate");
                for (Double splitRate : config.defaultSplitRates()) {
                    generator.writeNumber(splitRate);
                }
                generator.writeEndArray();
                generator.writeEndObject();

                generator.writeEndArray();
                generator.writeEndObject();
                generator.writeEndObject();
            }
        } else {

            for (RdfTaskTypeV2 taskType : RdfTaskTypeV2.values()) {
                List<LabelConfigV2> taskSpecificConfigs = classificationSpecifications.stream().filter(c -> c.taskType().equals(taskType.name())).collect(Collectors.toList());

                if (taskSpecificConfigs.isEmpty()) {
                    continue;
                }

                if (taskType == RdfTaskTypeV2.link_prediction) {
                    for (String filename : filenames) {
                        generator.writeStartObject();
                        generator.writeStringField("file_name", filename);

                        generator.writeObjectFieldStart("label");
                        generator.writeStringField("task_type", taskType.name());

                        generator.writeArrayFieldStart("targets");

                        for (LabelConfigV2 taskSpecificConfig : taskSpecificConfigs) {
                            generator.writeStartObject();

                            if (StringUtils.isNotEmpty(taskSpecificConfig.subject())) {
                                generator.writeStringField("subject", taskSpecificConfig.subject());
                            } else {
                                warnings.add("'subject' field is missing for link_prediction task, so all edges will be treated as the training target.");
                            }

                            if (StringUtils.isNotEmpty(taskSpecificConfig.property())) {
                                generator.writeStringField("predicate", taskSpecificConfig.property());
                            }else {
                                warnings.add("'predicate' field is missing for link_prediction task, so all edges will be treated as the training target.");
                            }

                            if (StringUtils.isNotEmpty(taskSpecificConfig.object())) {
                                generator.writeStringField("object", taskSpecificConfig.object());
                            }else {
                                warnings.add("'object' field is missing for link_prediction task, so all edges will be treated as the training target.");
                            }

                            generator.writeArrayFieldStart("split_rate");
                            for (Double splitRate : taskSpecificConfig.splitRates()) {
                                generator.writeNumber(splitRate);
                            }
                            generator.writeEndArray();
                            generator.writeEndObject();
                        }

                        generator.writeEndArray();
                        generator.writeEndObject();
                        generator.writeEndObject();
                    }
                } else {
                    for (String filename : filenames) {
                        generator.writeStartObject();
                        generator.writeStringField("file_name", filename);

                        generator.writeObjectFieldStart("label");
                        generator.writeStringField("task_type", taskType.name());

                        generator.writeArrayFieldStart("targets");

                        for (LabelConfigV2 taskSpecificConfig : taskSpecificConfigs) {
                            generator.writeStartObject();
                            generator.writeStringField("node", taskSpecificConfig.label().labelsAsString());
                            String property = taskSpecificConfig.property();

                            if (StringUtils.isNotEmpty(property)){
                                generator.writeStringField("predicate", property);
                            } else {
                                warnings.add(String.format("'predicate' field is missing for %s task. If the target nodes have more than one predicate defining the target node feature, the training task will fail with an error.", taskType));
                            }

                            generator.writeArrayFieldStart("split_rate");
                            for (Double splitRate : taskSpecificConfig.splitRates()) {
                                generator.writeNumber(splitRate);
                            }
                            generator.writeEndArray();
                            generator.writeEndObject();
                        }

                        generator.writeEndArray();
                        generator.writeEndObject();
                        generator.writeEndObject();
                    }
                }
            }
        }


        generator.writeEndArray();
    }