diff --git a/src/main/kotlin/dev/dnpm/etl/processor/pseudonym/PseudonymizeService.kt b/src/main/kotlin/dev/dnpm/etl/processor/pseudonym/PseudonymizeService.kt index 1a79850..ab8ce2f 100644 --- a/src/main/kotlin/dev/dnpm/etl/processor/pseudonym/PseudonymizeService.kt +++ b/src/main/kotlin/dev/dnpm/etl/processor/pseudonym/PseudonymizeService.kt @@ -19,7 +19,6 @@ package dev.dnpm.etl.processor.pseudonym -import de.ukw.ccc.bwhc.dto.MtbFile import dev.dnpm.etl.processor.config.PseudonymizeConfigProperties class PseudonymizeService( @@ -27,38 +26,11 @@ class PseudonymizeService( private val configProperties: PseudonymizeConfigProperties ) { - fun pseudonymize(mtbFile: MtbFile): MtbFile { - val patientPseudonym = patientPseudonym(mtbFile.patient.id) - - mtbFile.episode.patient = patientPseudonym - mtbFile.carePlans.forEach { it.patient = patientPseudonym } - mtbFile.patient.id = patientPseudonym - mtbFile.claims.forEach { it.patient = patientPseudonym } - mtbFile.consent.patient = patientPseudonym - mtbFile.claimResponses.forEach { it.patient = patientPseudonym } - mtbFile.diagnoses.forEach { it.patient = patientPseudonym } - mtbFile.ecogStatus.forEach { it.patient = patientPseudonym } - mtbFile.familyMemberDiagnoses.forEach { it.patient = patientPseudonym } - mtbFile.geneticCounsellingRequests.forEach { it.patient = patientPseudonym } - mtbFile.histologyReevaluationRequests.forEach { it.patient = patientPseudonym } - mtbFile.histologyReports.forEach { it.patient = patientPseudonym } - mtbFile.lastGuidelineTherapies.forEach { it.patient = patientPseudonym } - mtbFile.molecularPathologyFindings.forEach { it.patient = patientPseudonym } - mtbFile.molecularTherapies.forEach { it.history.forEach { it.patient = patientPseudonym } } - mtbFile.ngsReports.forEach { it.patient = patientPseudonym } - mtbFile.previousGuidelineTherapies.forEach { it.patient = patientPseudonym } - mtbFile.rebiopsyRequests.forEach { it.patient = patientPseudonym } - mtbFile.recommendations.forEach { it.patient = patientPseudonym } - mtbFile.recommendations.forEach { it.patient = patientPseudonym } - mtbFile.responses.forEach { it.patient = patientPseudonym } - mtbFile.specimens.forEach { it.patient = patientPseudonym } - mtbFile.specimens.forEach { it.patient = patientPseudonym } - - return mtbFile - } - fun patientPseudonym(patientId: String): String { - return "${configProperties.prefix}_${generator.generate(patientId)}" + return when (generator) { + is GpasPseudonymGenerator -> generator.generate(patientId) + else -> "${configProperties.prefix}_${generator.generate(patientId)}" + } } } \ No newline at end of file diff --git a/src/main/kotlin/dev/dnpm/etl/processor/pseudonym/extensions.kt b/src/main/kotlin/dev/dnpm/etl/processor/pseudonym/extensions.kt new file mode 100644 index 0000000..580785d --- /dev/null +++ b/src/main/kotlin/dev/dnpm/etl/processor/pseudonym/extensions.kt @@ -0,0 +1,50 @@ +/* + * This file is part of ETL-Processor + * + * Copyright (c) 2023 Comprehensive Cancer Center Mainfranken, Datenintegrationszentrum Philipps-Universität Marburg and Contributors + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU Affero General Public License as published + * by the Free Software Foundation, either version 3 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU Affero General Public License for more details. + * + * You should have received a copy of the GNU Affero General Public License + * along with this program. If not, see . + */ + +package dev.dnpm.etl.processor.pseudonym + +import de.ukw.ccc.bwhc.dto.MtbFile + +infix fun MtbFile.pseudonymizeWith(pseudonymizeService: PseudonymizeService) { + val patientPseudonym = pseudonymizeService.patientPseudonym(this.patient.id) + + this.episode.patient = patientPseudonym + this.carePlans.forEach { it.patient = patientPseudonym } + this.patient.id = patientPseudonym + this.claims.forEach { it.patient = patientPseudonym } + this.consent.patient = patientPseudonym + this.claimResponses.forEach { it.patient = patientPseudonym } + this.diagnoses.forEach { it.patient = patientPseudonym } + this.ecogStatus.forEach { it.patient = patientPseudonym } + this.familyMemberDiagnoses.forEach { it.patient = patientPseudonym } + this.geneticCounsellingRequests.forEach { it.patient = patientPseudonym } + this.histologyReevaluationRequests.forEach { it.patient = patientPseudonym } + this.histologyReports.forEach { it.patient = patientPseudonym } + this.lastGuidelineTherapies.forEach { it.patient = patientPseudonym } + this.molecularPathologyFindings.forEach { it.patient = patientPseudonym } + this.molecularTherapies.forEach { it.history.forEach { it.patient = patientPseudonym } } + this.ngsReports.forEach { it.patient = patientPseudonym } + this.previousGuidelineTherapies.forEach { it.patient = patientPseudonym } + this.rebiopsyRequests.forEach { it.patient = patientPseudonym } + this.recommendations.forEach { it.patient = patientPseudonym } + this.recommendations.forEach { it.patient = patientPseudonym } + this.responses.forEach { it.patient = patientPseudonym } + this.specimens.forEach { it.patient = patientPseudonym } + this.specimens.forEach { it.patient = patientPseudonym } +} \ No newline at end of file diff --git a/src/main/kotlin/dev/dnpm/etl/processor/services/RequestProcessor.kt b/src/main/kotlin/dev/dnpm/etl/processor/services/RequestProcessor.kt index 936c1bf..6465e82 100644 --- a/src/main/kotlin/dev/dnpm/etl/processor/services/RequestProcessor.kt +++ b/src/main/kotlin/dev/dnpm/etl/processor/services/RequestProcessor.kt @@ -27,6 +27,7 @@ import dev.dnpm.etl.processor.monitoring.RequestStatus import dev.dnpm.etl.processor.monitoring.RequestType import dev.dnpm.etl.processor.output.MtbFileSender import dev.dnpm.etl.processor.pseudonym.PseudonymizeService +import dev.dnpm.etl.processor.pseudonym.pseudonymizeWith import org.apache.commons.codec.binary.Base32 import org.apache.commons.codec.digest.DigestUtils import org.slf4j.LoggerFactory @@ -47,12 +48,13 @@ class RequestProcessor( fun processMtbFile(mtbFile: MtbFile) { val pid = mtbFile.patient.id - val pseudonymized = pseudonymizeService.pseudonymize(mtbFile) - if (isDuplication(pseudonymized)) { + mtbFile pseudonymizeWith pseudonymizeService + + if (isDuplication(mtbFile)) { requestService.save( Request( - patientId = pseudonymized.patient.id, + patientId = mtbFile.patient.id, pid = pid, fingerprint = fingerprint(mtbFile), status = RequestStatus.DUPLICATION, @@ -64,19 +66,19 @@ class RequestProcessor( return } - val request = MtbFileSender.MtbFileRequest(UUID.randomUUID().toString(), pseudonymized) + val request = MtbFileSender.MtbFileRequest(UUID.randomUUID().toString(), mtbFile) val responseStatus = sender.send(request) if (responseStatus.status == MtbFileSender.ResponseStatus.SUCCESS || responseStatus.status == MtbFileSender.ResponseStatus.WARNING) { logger.info( "Sent file for Patient '{}' using '{}'", - pseudonymized.patient.id, + mtbFile.patient.id, sender.javaClass.simpleName ) } else { logger.error( "Error sending file for Patient '{}' using '{}'", - pseudonymized.patient.id, + mtbFile.patient.id, sender.javaClass.simpleName ) } diff --git a/src/test/kotlin/dev/dnpm/etl/processor/pseudonym/PseudonymizeServiceTest.kt b/src/test/kotlin/dev/dnpm/etl/processor/pseudonym/PseudonymizeServiceTest.kt new file mode 100644 index 0000000..a30a328 --- /dev/null +++ b/src/test/kotlin/dev/dnpm/etl/processor/pseudonym/PseudonymizeServiceTest.kt @@ -0,0 +1,86 @@ +/* + * This file is part of ETL-Processor + * + * Copyright (c) 2023 Comprehensive Cancer Center Mainfranken, Datenintegrationszentrum Philipps-Universität Marburg and Contributors + * + * This program is free software: you can redistribute it and/or modify + * it under the terms of the GNU Affero General Public License as published + * by the Free Software Foundation, either version 3 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU Affero General Public License for more details. + * + * You should have received a copy of the GNU Affero General Public License + * along with this program. If not, see . + */ + +package dev.dnpm.etl.processor.pseudonym + +import de.ukw.ccc.bwhc.dto.* +import dev.dnpm.etl.processor.config.PseudonymizeConfigProperties +import org.assertj.core.api.Assertions.assertThat +import org.junit.jupiter.api.Test +import org.junit.jupiter.api.extension.ExtendWith +import org.mockito.ArgumentMatchers.anyString +import org.mockito.Mock +import org.mockito.junit.jupiter.MockitoExtension +import org.mockito.kotlin.doAnswer +import org.mockito.kotlin.whenever + +@ExtendWith(MockitoExtension::class) +class PseudonymizeServiceTest { + + private val mtbFile = MtbFile.builder() + .withPatient( + Patient.builder() + .withId("123") + .withBirthDate("2000-08-08") + .withGender(Patient.Gender.MALE) + .build() + ) + .withConsent( + Consent.builder() + .withId("1") + .withStatus(Consent.Status.ACTIVE) + .withPatient("123") + .build() + ) + .withEpisode( + Episode.builder() + .withId("1") + .withPatient("123") + .withPeriod(PeriodStart("2023-08-08")) + .build() + ) + .build() + + @Test + fun shouldNotUsePseudonymPrefixForGpas(@Mock generator: GpasPseudonymGenerator) { + doAnswer { + it.arguments[0] + }.whenever(generator).generate(anyString()) + + val pseudonymizeService = PseudonymizeService(generator, PseudonymizeConfigProperties()) + + mtbFile.pseudonymizeWith(pseudonymizeService) + + assertThat(mtbFile.patient.id).isEqualTo("123") + } + + @Test + fun shouldUsePseudonymPrefixForBuiltin(@Mock generator: AnonymizingGenerator) { + doAnswer { + it.arguments[0] + }.whenever(generator).generate(anyString()) + + val pseudonymizeService = PseudonymizeService(generator, PseudonymizeConfigProperties()) + + mtbFile.pseudonymizeWith(pseudonymizeService) + + assertThat(mtbFile.patient.id).isEqualTo("UNKNOWN_123") + } + +} \ No newline at end of file diff --git a/src/test/kotlin/dev/dnpm/etl/processor/services/RequestProcessorTest.kt b/src/test/kotlin/dev/dnpm/etl/processor/services/RequestProcessorTest.kt index 6e97343..8552bbb 100644 --- a/src/test/kotlin/dev/dnpm/etl/processor/services/RequestProcessorTest.kt +++ b/src/test/kotlin/dev/dnpm/etl/processor/services/RequestProcessorTest.kt @@ -82,7 +82,7 @@ class RequestProcessorTest { uuid = UUID.randomUUID().toString(), patientId = "TEST_12345678901", pid = "P1", - fingerprint = "cwaxsvectyfj4qcw4hiwzx5fwwo7lekyagpzd2ayuf36jlvi6msa", + fingerprint = "xrysxpozhbs2lnrjgf3yq4fzj33kxr7xr5c2cbuskmelfdmckl3a", type = RequestType.MTB_FILE, status = RequestStatus.SUCCESS, processedAt = Instant.parse("2023-08-08T02:00:00Z") @@ -94,8 +94,8 @@ class RequestProcessorTest { }.`when`(requestService).isLastRequestDeletion(anyString()) doAnswer { - it.arguments[0] as MtbFile - }.`when`(pseudonymizeService).pseudonymize(any()) + it.arguments[0] as String + }.`when`(pseudonymizeService).patientPseudonym(any()) val mtbFile = MtbFile.builder() .withPatient( @@ -153,8 +153,8 @@ class RequestProcessorTest { }.`when`(sender).send(any()) doAnswer { - it.arguments[0] as MtbFile - }.`when`(pseudonymizeService).pseudonymize(any()) + it.arguments[0] as String + }.`when`(pseudonymizeService).patientPseudonym(any()) val mtbFile = MtbFile.builder() .withPatient( @@ -212,8 +212,8 @@ class RequestProcessorTest { }.`when`(sender).send(any()) doAnswer { - it.arguments[0] as MtbFile - }.`when`(pseudonymizeService).pseudonymize(any()) + it.arguments[0] as String + }.`when`(pseudonymizeService).patientPseudonym(any()) val mtbFile = MtbFile.builder() .withPatient(