Add new convert module and sketch its integration

modules/convert/src/main/scala/docspell/convert/Conversion.scala

@@ -0,0 +1,24 @@
+package docspell.convert
+
+import fs2._
+import cats.effect._
+import docspell.common._
+
+trait Conversion[F[_]] {
+
+  def toPDF(inType: MimeType): Pipe[F, Byte, Byte]
+
+}
+
+object Conversion {
+
+  def create[F[_]: Sync](cfg: ConvertConfig): Resource[F, Conversion[F]] =
+    Resource.pure(new Conversion[F] {
+
+      def toPDF(inType: MimeType): Pipe[F, Byte, Byte] = {
+        println(cfg)
+        ???
+      }
+
+    })
+}

modules/convert/src/main/scala/docspell/convert/ConvertConfig.scala

@@ -0,0 +1,3 @@
+package docspell.convert
+
+case class ConvertConfig()

modules/joex/src/main/scala/docspell/joex/Config.scala

@@ -4,6 +4,7 @@ import docspell.common.{Ident, LenientUri}
 import docspell.joex.scheduler.SchedulerConfig
 import docspell.store.JdbcConfig
 import docspell.text.ocr.{Config => OcrConfig}
+import docspell.convert.ConvertConfig
 
 case class Config(
     appId: Ident,
@@ -11,17 +12,10 @@ case class Config(
     bind: Config.Bind,
     jdbc: JdbcConfig,
     scheduler: SchedulerConfig,
-    extraction: OcrConfig
+    extraction: OcrConfig,
+    convert: ConvertConfig
 )
 
 object Config {
-  val postgres =
-    JdbcConfig(LenientUri.unsafe("jdbc:postgresql://localhost:5432/docspelldev"), "dev", "dev")
-  val h2 = JdbcConfig(
-    LenientUri.unsafe("jdbc:h2:./target/docspelldev.db;MODE=PostgreSQL;DATABASE_TO_LOWER=TRUE"),
-    "sa",
-    ""
-  )
-
   case class Bind(address: String, port: Int)
 }

modules/joex/src/main/scala/docspell/joex/JoexAppImpl.scala

@@ -55,7 +55,7 @@ object JoexAppImpl {
               .withTask(
                 JobTask.json(
                   ProcessItemArgs.taskName,
-                  ItemHandler[F](cfg.extraction),
+                  ItemHandler[F](cfg),
                   ItemHandler.onCancel[F]
                 )
               )

modules/joex/src/main/scala/docspell/joex/process/ConvertPdf.scala

@@ -7,6 +7,7 @@ import cats.effect._
 import cats.data.OptionT
 
 import docspell.common._
+import docspell.convert._
 import docspell.joex.scheduler._
 import docspell.store.records._
 
@@ -27,17 +28,17 @@ import docspell.store.records._
 object ConvertPdf {
 
   def apply[F[_]: Sync: ContextShift](
+      cfg: ConvertConfig,
       item: ItemData
   ): Task[F, ProcessItemArgs, ItemData] =
     Task { ctx =>
-
       // get mimetype
       //   try to convert
       //   save to db
       //   update file_id of RAttachment
 
       def convert(ra: RAttachment) =
-        findMime(ctx)(ra).flatMap(m => convertSafe(ctx)(ra, m))
+        findMime(ctx)(ra).flatMap(m => convertSafe(cfg, ctx)(ra, m))
 
       for {
         ras <- item.attachments.traverse(convert)
@@ -51,10 +52,12 @@ object ConvertPdf {
       .getOrElse(Mimetype.`application/octet-stream`)
 
   def convertSafe[F[_]: Sync](
+      cfg: ConvertConfig,
       ctx: Context[F, ProcessItemArgs]
-  )(ra: RAttachment, mime: Mimetype): F[RAttachment] = {
-
-    ctx.logger.info(s"File ${ra.name} has mime ${mime.asString}").
-      map(_ => ra)
-  }
+  )(ra: RAttachment, mime: Mimetype): F[RAttachment] =
+    Conversion.create[F](cfg).use { conv =>
+      ctx.logger
+        .info(s"File ${ra.name} has mime ${mime.asString}. conv=$conv")
+        .map(_ => ra)
+    }
 }

modules/joex/src/main/scala/docspell/joex/process/ItemHandler.scala

@@ -3,16 +3,16 @@ package docspell.joex.process
 import cats.implicits._
 import cats.effect.{ContextShift, Sync}
 import docspell.common.{ItemState, ProcessItemArgs}
+import docspell.joex.Config
 import docspell.joex.scheduler.{Context, Task}
 import docspell.store.queries.QItem
 import docspell.store.records.{RItem, RJob}
-import docspell.text.ocr.{Config => OcrConfig}
 
 object ItemHandler {
   def onCancel[F[_]: Sync: ContextShift]: Task[F, ProcessItemArgs, Unit] =
     logWarn("Now cancelling. Deleting potentially created data.").flatMap(_ => deleteByFileIds)
 
-  def apply[F[_]: Sync: ContextShift](cfg: OcrConfig): Task[F, ProcessItemArgs, Unit] =
+  def apply[F[_]: Sync: ContextShift](cfg: Config): Task[F, ProcessItemArgs, Unit] =
     CreateItem[F]
       .flatMap(itemStateTask(ItemState.Processing))
       .flatMap(safeProcess[F](cfg))
@@ -30,7 +30,7 @@ object ItemHandler {
     } yield last
 
   def safeProcess[F[_]: Sync: ContextShift](
-      cfg: OcrConfig
+      cfg: Config
   )(data: ItemData): Task[F, ProcessItemArgs, ItemData] =
     Task(isLastRetry[F, ProcessItemArgs] _).flatMap {
       case true =>

modules/joex/src/main/scala/docspell/joex/process/ProcessItem.scala

@@ -3,15 +3,15 @@ package docspell.joex.process
 import cats.effect.{ContextShift, Sync}
 import docspell.common.ProcessItemArgs
 import docspell.joex.scheduler.Task
-import docspell.text.ocr.{Config => OcrConfig}
+import docspell.joex.Config
 
 object ProcessItem {
 
   def apply[F[_]: Sync: ContextShift](
-      cfg: OcrConfig
+      cfg: Config
   )(item: ItemData): Task[F, ProcessItemArgs, ItemData] =
-    ConvertPdf(item)
-      .flatMap(TextExtraction(cfg, _))
+    ConvertPdf(cfg.convert, item)
+      .flatMap(TextExtraction(cfg.extraction, _))
       .flatMap(Task.setProgress(25))
       .flatMap(TextAnalysis[F])
       .flatMap(Task.setProgress(50))