Reprocessing now sets metadata to an item if not in state confirmed

When reprocessing an item, the metadat of all *files* are replaced.
This change now also sets some metadat to an item, but only if the
item is not in state "confirmed". Confirmed items are not touched, but
the metadata of the files is updated.

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

@@ -113,7 +113,7 @@ object JoexAppImpl {
         .withTask(
           JobTask.json(
             ReProcessItemArgs.taskName,
-            ReProcessItem[F](cfg, fts, analyser, regexNer),
+            ReProcessItem[F](cfg, fts, itemOps, analyser, regexNer),
             ReProcessItem.onCancel[F]
           )
         )

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

@@ -10,11 +10,19 @@ import docspell.store.records.RItem
 
 object LinkProposal {
 
-  def apply[F[_]: Sync](data: ItemData): Task[F, ProcessItemArgs, ItemData] =
+  def onlyNew[F[_]: Sync](data: ItemData): Task[F, ProcessItemArgs, ItemData] =
     if (data.item.state.isValid)
       Task
         .log[F, ProcessItemArgs](_.debug(s"Not linking proposals on existing item"))
         .map(_ => data)
+    else
+      LinkProposal[F](data)
+
+  def apply[F[_]: Sync](data: ItemData): Task[F, ProcessItemArgs, ItemData] =
+    if (data.item.state == ItemState.Confirmed)
+      Task
+        .log[F, ProcessItemArgs](_.debug(s"Not linking proposals on confirmed item"))
+        .map(_ => data)
     else
       Task { ctx =>
         val proposals = data.finalProposals

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

@@ -22,8 +22,8 @@ object ProcessItem {
     ExtractArchive(item)
       .flatMap(Task.setProgress(20))
       .flatMap(processAttachments0(cfg, fts, analyser, regexNer, (40, 60, 80)))
-      .flatMap(LinkProposal[F])
-      .flatMap(SetGivenData[F](itemOps))
+      .flatMap(LinkProposal.onlyNew[F])
+      .flatMap(SetGivenData.onlyNew[F](itemOps))
       .flatMap(Task.setProgress(99))
       .flatMap(RemoveEmptyItem(itemOps))
 

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

@@ -5,6 +5,7 @@ import cats.effect._
 import cats.implicits._
 
 import docspell.analysis.TextAnalyser
+import docspell.backend.ops.OItem
 import docspell.common._
 import docspell.ftsclient.FtsClient
 import docspell.joex.Config
@@ -22,12 +23,17 @@ object ReProcessItem {
   def apply[F[_]: ConcurrentEffect: ContextShift](
       cfg: Config,
       fts: FtsClient[F],
+      itemOps: OItem[F],
       analyser: TextAnalyser[F],
       regexNer: RegexNerFile[F]
   ): Task[F, Args, Unit] =
-    loadItem[F]
-      .flatMap(safeProcess[F](cfg, fts, analyser, regexNer))
-      .map(_ => ())
+    Task
+      .log[F, Args](_.info("===== Start reprocessing ======"))
+      .flatMap(_ =>
+        loadItem[F]
+          .flatMap(safeProcess[F](cfg, fts, itemOps, analyser, regexNer))
+          .map(_ => ())
+      )
 
   def onCancel[F[_]]: Task[F, Args, Unit] =
     logWarn("Now cancelling re-processing.")
@@ -58,6 +64,11 @@ object ReProcessItem {
                 a.copy(fileId = src.fileId, name = src.name)
               }
             )
+        _ <- OptionT.liftF(
+          ctx.logger.debug(
+            s"Loaded item and ${attachSrc.size} attachments to reprocess"
+          )
+        )
       } yield ItemData(
         item,
         attachSrc,
@@ -76,6 +87,7 @@ object ReProcessItem {
   def processFiles[F[_]: ConcurrentEffect: ContextShift](
       cfg: Config,
       fts: FtsClient[F],
+      itemOps: OItem[F],
       analyser: TextAnalyser[F],
       regexNer: RegexNerFile[F],
       data: ItemData
@@ -89,9 +101,9 @@ object ReProcessItem {
               data.item.cid,
               args.itemId.some,
               lang,
-              None, //direction
-              "",   //source-id
-              None, //folder
+              None,             //direction
+              data.item.source, //source-id
+              None,             //folder
               Seq.empty,
               false,
               None,
@@ -103,6 +115,8 @@ object ReProcessItem {
     getLanguage[F].flatMap { lang =>
       ProcessItem
         .processAttachments[F](cfg, fts, analyser, regexNer)(data)
+        .flatMap(LinkProposal[F])
+        .flatMap(SetGivenData[F](itemOps))
         .contramap[Args](convertArgs(lang))
     }
   }
@@ -121,12 +135,13 @@ object ReProcessItem {
   def safeProcess[F[_]: ConcurrentEffect: ContextShift](
       cfg: Config,
       fts: FtsClient[F],
+      itemOps: OItem[F],
       analyser: TextAnalyser[F],
       regexNer: RegexNerFile[F]
   )(data: ItemData): Task[F, Args, ItemData] =
     isLastRetry[F].flatMap {
       case true =>
-        processFiles[F](cfg, fts, analyser, regexNer, data).attempt
+        processFiles[F](cfg, fts, itemOps, analyser, regexNer, data).attempt
           .flatMap({
             case Right(d) =>
               Task.pure(d)
@@ -136,7 +151,7 @@ object ReProcessItem {
               ).andThen(_ => Sync[F].raiseError(ex))
           })
       case false =>
-        processFiles[F](cfg, fts, analyser, regexNer, data)
+        processFiles[F](cfg, fts, itemOps, analyser, regexNer, data)
     }
 
   private def logWarn[F[_]](msg: => String): Task[F, Args, Unit] =

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

@@ -8,13 +8,20 @@ import docspell.common._
 import docspell.joex.scheduler.Task
 
 object SetGivenData {
+  type Args = ProcessItemArgs
 
-  def apply[F[_]: Sync](
-      ops: OItem[F]
-  )(data: ItemData): Task[F, ProcessItemArgs, ItemData] =
+  def onlyNew[F[_]: Sync](ops: OItem[F])(data: ItemData): Task[F, Args, ItemData] =
     if (data.item.state.isValid)
       Task
-        .log[F, ProcessItemArgs](_.debug(s"Not setting data on existing item"))
+        .log[F, Args](_.debug(s"Not setting data on existing item"))
+        .map(_ => data)
+    else
+      SetGivenData[F](ops)(data)
+
+  def apply[F[_]: Sync](ops: OItem[F])(data: ItemData): Task[F, Args, ItemData] =
+    if (data.item.state == ItemState.Confirmed)
+      Task
+        .log[F, Args](_.debug(s"Not setting data on confirmed item"))
         .map(_ => data)
     else
       setFolder(data, ops).flatMap(d => setTags[F](d, ops))
@@ -22,7 +29,7 @@ object SetGivenData {
   private def setFolder[F[_]: Sync](
       data: ItemData,
       ops: OItem[F]
-  ): Task[F, ProcessItemArgs, ItemData] =
+  ): Task[F, Args, ItemData] =
     Task { ctx =>
       val itemId     = data.item.id
       val folderId   = ctx.args.meta.folderId
@@ -41,7 +48,7 @@ object SetGivenData {
   private def setTags[F[_]: Sync](
       data: ItemData,
       ops: OItem[F]
-  ): Task[F, ProcessItemArgs, ItemData] =
+  ): Task[F, Args, ItemData] =
     Task { ctx =>
       val itemId     = data.item.id
       val collective = ctx.args.meta.collective

modules/restapi/src/main/resources/docspell-openapi.yml

@@ -2113,7 +2113,11 @@ paths:
       summary: Start reprocessing the files of the item.
       description: |
         This submits a job that will re-process the files (either all
-        or the ones specified) of the item and replace the metadata.
+        or the ones specified) of the item and replace their metadata.
+
+        If the item is not in "confirmed" state, its associated metada
+        is also updated. Otherwise only the file metadata is updated
+        (like extracted text etc).
       security:
         - authTokenHeader: []
       parameters: