From 6d5d69532ad452f81371ed147e0bd5aa5b4cd603 Mon Sep 17 00:00:00 2001
From: Gauthier Roebroeck <gauthier.roebroeck@gmail.com>
Date: Thu, 19 Oct 2023 15:53:50 +0800
Subject: [PATCH] fix(komga): do not cache PDF documents

this can cause some issues when multiple tasks are running in parallel
---
 .../mediacontainer/PdfExtractor.kt            | 36 ++++++++-----------
 1 file changed, 15 insertions(+), 21 deletions(-)

diff --git a/komga/src/main/kotlin/org/gotson/komga/infrastructure/mediacontainer/PdfExtractor.kt b/komga/src/main/kotlin/org/gotson/komga/infrastructure/mediacontainer/PdfExtractor.kt
index f0aaeca7..e72d6e6c 100644
--- a/komga/src/main/kotlin/org/gotson/komga/infrastructure/mediacontainer/PdfExtractor.kt
+++ b/komga/src/main/kotlin/org/gotson/komga/infrastructure/mediacontainer/PdfExtractor.kt
@@ -1,6 +1,5 @@
 package org.gotson.komga.infrastructure.mediacontainer
 
-import com.github.benmanes.caffeine.cache.Caffeine
 import mu.KotlinLogging
 import org.apache.pdfbox.io.MemoryUsageSetting
 import org.apache.pdfbox.multipdf.PageExtractor
@@ -15,7 +14,6 @@ import org.gotson.komga.domain.model.MediaType
 import org.springframework.stereotype.Service
 import java.io.ByteArrayOutputStream
 import java.nio.file.Path
-import java.util.concurrent.TimeUnit
 import javax.imageio.ImageIO
 import kotlin.math.roundToInt
 
@@ -28,12 +26,6 @@ class PdfExtractor : MediaContainerRawExtractor {
   private val imageIOFormat = "jpeg"
   private val resolution = 1536F
 
-  private val cache = Caffeine.newBuilder()
-    .maximumSize(20)
-    .expireAfterAccess(1, TimeUnit.MINUTES)
-    .evictionListener { _: Path?, pdf: PDDocument?, _ -> pdf?.close() }
-    .build<Path, PDDocument>()
-
   override fun mediaTypes(): List<String> = listOf(MediaType.PDF.type)
 
   override fun getEntries(path: Path, analyzeDimensions: Boolean): List<MediaContainerEntry> =
@@ -47,24 +39,26 @@ class PdfExtractor : MediaContainerRawExtractor {
     }
 
   override fun getEntryStream(path: Path, entryName: String): ByteArray {
-    val pdf = cache.get(path) { PDDocument.load(path.toFile(), MemoryUsageSetting.setupTempFileOnly()) }!!
-    val pageNumber = entryName.toInt()
-    val page = pdf.getPage(pageNumber)
-    val image = PDFRenderer(pdf).renderImage(pageNumber, page.getScale(), ImageType.RGB)
-    return ByteArrayOutputStream().use { out ->
-      ImageIO.write(image, imageIOFormat, out)
-      out.toByteArray()
+    PDDocument.load(path.toFile(), MemoryUsageSetting.setupTempFileOnly()).use { pdf ->
+      val pageNumber = entryName.toInt()
+      val page = pdf.getPage(pageNumber)
+      val image = PDFRenderer(pdf).renderImage(pageNumber, page.getScale(), ImageType.RGB)
+      return ByteArrayOutputStream().use { out ->
+        ImageIO.write(image, imageIOFormat, out)
+        out.toByteArray()
+      }
     }
   }
 
   override fun getRawEntryStream(path: Path, entryName: String): BookPageContent {
-    val pdf = cache.get(path) { PDDocument.load(path.toFile(), MemoryUsageSetting.setupTempFileOnly()) }!!
-    val pageNumber = entryName.toInt() + 1
-    val bytes = ByteArrayOutputStream().use { out ->
-      PageExtractor(pdf, pageNumber, pageNumber).extract().save(out)
-      out.toByteArray()
+    PDDocument.load(path.toFile(), MemoryUsageSetting.setupTempFileOnly()).use { pdf ->
+      val pageNumber = entryName.toInt() + 1
+      val bytes = ByteArrayOutputStream().use { out ->
+        PageExtractor(pdf, pageNumber, pageNumber).extract().save(out)
+        out.toByteArray()
+      }
+      return BookPageContent(bytes, MediaType.PDF.type)
     }
-    return BookPageContent(bytes, MediaType.PDF.type)
   }
 
   private fun PDPage.getScale() = resolution / minOf(cropBox.width, cropBox.height)