Skip to content

GitLab

Verified Commit c09b783f authored by Sebastian Schüpbach's avatar Sebastian Schüpbach
Browse files

fetch mime type dynamically from input data 


Signed-off-by: Sebastian Schüpbach's avatarSebastian Schüpbach <sebastian.schuepbach@unibas.ch>
parent 5f3e0310
Pipeline #11983 failed with stages
in 1 minute and 51 seconds
Hide whitespace changes
Inline Side-by-side
.gitignore
View file @ c09b783f
......@@ -2,5 +2,6 @@
!.idea/copyright
build
out/
.gradle
\ No newline at end of file
.gradle
build.gradle
View file @ c09b783f
......@@ -46,7 +46,7 @@ dependencies {
implementation 'org.apache.jena:apache-jena:3.14.0'
implementation 'org.memobase:memobase-service-utilities:1.6.0'
implementation 'org.memobase:memobase-service-utilities:1.10.0'
implementation 'org.memobase:fedora-client:0.4.1'
// KOTLIN IMPORTS
......
src/main/kotlin/Ingester.kt
View file @ c09b783f
......@@ -29,7 +29,11 @@ import java.io.File
import java.io.StringWriter
import java.net.URI
class Ingester(private val producer: Producer, private val sftpClient: SftpClient, private val fedoraClient: FedoraClient) : Closeable {
class Ingester(
private val producer: Producer,
private val sftpClient: SftpClient,
private val fedoraClient: FedoraClient
) : Closeable {
private val log = LogManager.getLogger("FedoraIngester")
......@@ -54,11 +58,13 @@ class Ingester(private val producer: Producer, private val sftpClient: SftpClien
fedoraClient.createOrUpdateRdfResource(URI(recordPair.first), data, RdfContentTypes.NTRIPLES)
log.info("Ingested record ${recordPair.first}.")
producer.sendReport(
Report(recordPair.first, ReportStatus.success, "Ingested record in fedora."))
Report(recordPair.first, ReportStatus.success, "Ingested record in fedora.")
)
} catch (ex: FcrepoOperationFailedException) {
log.error("Ingestion of record ${recordPair.first} failed: ${ex.localizedMessage}.")
producer.sendReport(
Report(recordPair.first, ReportStatus.failure, ex.localizedMessage))
Report(recordPair.first, ReportStatus.failure, ex.localizedMessage)
)
}
rdfHandler.getInstantiations().forEach { instantiationPair ->
......@@ -69,11 +75,13 @@ class Ingester(private val producer: Producer, private val sftpClient: SftpClien
fedoraClient.createOrUpdateRdfResource(URI(instantiationPair.first), data, RdfContentTypes.NTRIPLES)
log.info("Ingested instantiation ${instantiationPair.first}.")
producer.sendReport(
Report(instantiationPair.first, ReportStatus.success, "Ingested an instantiation in fedora."))
Report(instantiationPair.first, ReportStatus.success, "Ingested an instantiation in fedora.")
)
} catch (ex: FcrepoOperationFailedException) {
log.error("Ingestion of instantiation ${instantiationPair.first} failed: ${ex.localizedMessage}.")
producer.sendReport(
Report(instantiationPair.first, ReportStatus.failure, ex.localizedMessage))
Report(instantiationPair.first, ReportStatus.failure, ex.localizedMessage)
)
}
}
......@@ -81,16 +89,22 @@ class Ingester(private val producer: Producer, private val sftpClient: SftpClien
it.second.let { path ->
if (path != null) {
sftpClient.open(File(path)).use { stream ->
val binaryUri = "${it.first}/binary"
val digitalInstantationUri = it.first
val binaryUri = "$digitalInstantationUri/binary"
try {
fedoraClient.createOrUpdateBinaryResource(URI(binaryUri), stream, "image/jpeg")
val mimeType = rdfHandler.getMimeType().first { mT ->
mT.first == digitalInstantationUri
}.second
fedoraClient.createOrUpdateBinaryResource(URI(binaryUri), stream, mimeType)
log.info("Ingested binary $binaryUri.")
producer.sendReport(
Report(binaryUri, ReportStatus.success, "Ingested a binary in fedora."))
Report(binaryUri, ReportStatus.success, "Ingested a binary in fedora.")
)
} catch (ex: FcrepoOperationFailedException) {
log.error("Ingestion of binary $binaryUri failed: ${ex.localizedMessage}.")
producer.sendReport(
Report(binaryUri, ReportStatus.failure, ex.localizedMessage))
Report(binaryUri, ReportStatus.failure, ex.localizedMessage)
)
}
}
}
......@@ -110,15 +124,15 @@ class Ingester(private val producer: Producer, private val sftpClient: SftpClien
return try {
ingest(record.key(), record.value())
Report(
id = record.key(),
status = ReportStatus.success,
message = ReportMessages.ingestedRecord(record.key())
id = record.key(),
status = ReportStatus.success,
message = ReportMessages.ingestedRecord(record.key())
)
} catch (e: FcrepoOperationFailedException) {
Report(
id = record.key(),
status = ReportStatus.failure,
message = ReportMessages.ingestFailed(record.key())
id = record.key(),
status = ReportStatus.failure,
message = ReportMessages.ingestFailed(record.key())
)
}
}
......
src/main/kotlin/RdfHandler.kt
View file @ c09b783f
......@@ -92,4 +92,12 @@ class RdfHandler(data: String) {
}
}.toList()
}
fun getMimeType(): List<Pair<String, String?>> {
return model.listSubjectsWithProperty(RICO.type, "digitalObject").filterKeep {
it.hasProperty(EBUCORE.hasMimeType)
}.mapWith {
Pair(it.uri, it.getProperty(EBUCORE.hasMimeType).`object`.asLiteral().string)
}.toList()
}
}
\ No newline at end of file
src/test/kotlin/TestRdfHandler.kt
View file @ c09b783f
package org.memobase
import org.apache.jena.rdf.model.Model
import org.apache.jena.rdf.model.ModelFactory
import org.apache.jena.rdf.model.impl.SelectorImpl
import org.apache.jena.riot.Lang
import org.apache.jena.riot.RDFDataMgr
import org.assertj.core.api.Assertions.`in`
import org.assertj.core.api.Assertions.assertThat
import org.assertj.core.api.Condition
import org.junit.jupiter.api.Test
import org.junit.jupiter.api.TestInstance
import java.io.File
import java.io.FileInputStream
import java.io.StringWriter
import java.nio.charset.Charset
import java.util.function.Predicate
@TestInstance(TestInstance.Lifecycle.PER_CLASS)
......@@ -113,6 +106,12 @@ class TestRdfHandler {
}
}
@Test
fun `test mime type extraction for binaries`() {
val rdfHandler = RdfHandler(readFile("input.nt"))
assertThat(rdfHandler.getMimeType().first().second).isEqualTo("video/mpeg")
}
private val uris2 = listOf(
"https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1",
......
src/test/resources/rdf/input.nt
View file @ c09b783f
......@@ -107,6 +107,7 @@ _:Bcb5da6d9X2Dfec6X2D4dd8X2D9b01X2D1c158a0464c7 <http://www.w3.org/1999/02/22-rd
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <https://www.ica.org/standards/RiC/ontology#type> "digitalObject" .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <https://www.ica.org/standards/RiC/ontology#identifiedBy> _:Bdd8afd91375dbb381b51dad3dd9af500 .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <http://www.ebu.ch/metadata/ontologies/ebucore/ebucore#locator> "http://digitalisate.afz.ethz.ch/?&guid=0a9aef73bc2845eabc3c20db7c141bea" .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <http://www.ebu.ch/metadata/ontologies/ebucore/ebucore#hasMimeType> "video/mpeg" .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <https://www.ica.org/standards/RiC/ontology#regulatedBy> _:Bc3f1507debc5b4ef586c9d1adc9a4e45 .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <https://www.ica.org/standards/RiC/ontology#regulatedBy> _:B9e9ebd9742b512fd44e26136d2fef230 .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <https://www.ica.org/standards/RiC/ontology#instantiates> <https://memobase.ch/record/AFZ-IB_Becker_Audiovisuals_63> .
......
src/test/resources/rdf/inputSftp.nt
View file @ c09b783f
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <https://www.ica.org/standards/RiC/ontology#Instantiation> .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <https://www.ica.org/standards/RiC/ontology#type> "digitalObject" .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <http://www.ebu.ch/metadata/ontologies/ebucore/ebucore#locator> "http://digitalisate.afz.ethz.ch/?&guid=0a9aef73bc2845eabc3c20db7c141bea" .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-1> <http://www.ebu.ch/metadata/ontologies/ebucore/ebucore#hasMimeType> "video/mpeg" .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-2> <http://www.w3.org/1999/02/22-rdf-syntax-ns#type> <https://www.ica.org/standards/RiC/ontology#Instantiation> .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-2> <https://www.ica.org/standards/RiC/ontology#type> "digitalObject" .
<https://memobase.ch/instantiation/digital/AFZ-IB_Becker_Audiovisuals_63-2> <http://www.ebu.ch/metadata/ontologies/ebucore/ebucore#locator> "sftp:/path/to/file/filename.jpg" .
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment