Fix mimeType and title extraction with tika (#79)

KSIUJ · May 15, 2020 · a4ee123 · a4ee123
1 parent cc63298
commit a4ee123
Showing 1 changed file with 7 additions and 4 deletions.
diff --git a/src/main/kotlin/pl/edu/uj/ii/ksi/mordor/services/TikaMetadataExtractor.kt b/src/main/kotlin/pl/edu/uj/ii/ksi/mordor/services/TikaMetadataExtractor.kt
@@ -4,6 +4,7 @@ import java.io.File
 import java.io.IOException
 import org.apache.tika.Tika
 import org.apache.tika.metadata.Metadata
+import org.apache.tika.metadata.TikaCoreProperties
 import org.slf4j.LoggerFactory
 import org.springframework.stereotype.Service
 import pl.edu.uj.ii.ksi.mordor.persistence.entities.FileMetadata
@@ -18,14 +19,16 @@ class TikaMetadataExtractor(private val tika: Tika, private val hashProvider: Fi
 
     override fun extract(file: File): FileMetadata? {
         return try {
+            val mimeType = tika.detect(file)
             val metadata = getTikaMetadata(file)
+            val title = metadata.get(TikaCoreProperties.TITLE) ?: file.nameWithoutExtension
 
             FileMetadata(
-                    author = metadata.get("Author"),
-                    description = metadata.get("Subject"),
+                    author = metadata.get(TikaCoreProperties.CREATOR),
+                    description = metadata.get(TikaCoreProperties.DESCRIPTION),
                     fileHash = hashProvider.calculate(file),
-                    title = metadata.get("Author"),
-                    mimeType = metadata.get("Author"),
+                    title = title,
+                    mimeType = mimeType,
                     thumbnail = null,
                     crawledContent = null
             )