Correctly count tag categories

If multiple tags of the same category are applied to the same item,
just summing tag counts will produce the wrong results as now items
are counted multiple times.

modules/backend/src/main/scala/docspell/backend/ops/OCollective.scala

@@ -71,6 +71,9 @@ object OCollective {
   type TagCount = docspell.store.queries.TagCount
   val TagCount = docspell.store.queries.TagCount
 
+  type CategoryCount = docspell.store.queries.CategoryCount
+  val CategoryCount = docspell.store.queries.CategoryCount
+
   type InsightData = QCollective.InsightData
   val insightData = QCollective.InsightData
 

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

@@ -4242,6 +4242,7 @@ components:
       required:
         - count
         - tagCloud
+        - tagCategoryCloud
         - fieldStats
         - folderStats
       properties:
@@ -4250,6 +4251,8 @@ components:
           format: int32
         tagCloud:
           $ref: "#/components/schemas/TagCloud"
+        tagCategoryCloud:
+          $ref: "#/components/schemas/NameCloud"
         fieldStats:
           type: array
           items:
@@ -4354,7 +4357,7 @@ components:
             $ref: "#/components/schemas/TagCount"
     TagCount:
       description: |
-        Generic structure for counting something.
+        Structure for counting tags.
       required:
         - tag
         - count
@@ -4364,6 +4367,30 @@ components:
         count:
           type: integer
           format: int32
+
+    NameCloud:
+      description: |
+        A set of counters.
+      required:
+        - items
+      properties:
+        items:
+          type: array
+          items:
+            $ref: "#/components/schemas/NameCount"
+    NameCount:
+      description: |
+        Generic structure for counting something.
+      required:
+        - name
+        - count
+      properties:
+        name:
+          type: string
+        count:
+          type: integer
+          format: int32
+
     AttachmentMeta:
       description: |
         Extracted meta data of an attachment.

modules/restserver/src/main/scala/docspell/restserver/conv/Conversions.scala

@@ -31,6 +31,7 @@ trait Conversions {
     SearchStats(
       sum.count,
       mkTagCloud(sum.tags),
+      mkTagCategoryCloud(sum.cats),
       sum.fields.map(mkFieldStats),
       sum.folders.map(mkFolderStats)
     )
@@ -63,6 +64,9 @@ trait Conversions {
   def mkTagCloud(tags: List[OCollective.TagCount]) =
     TagCloud(tags.map(tc => TagCount(mkTag(tc.tag), tc.count)))
 
+  def mkTagCategoryCloud(tags: List[OCollective.CategoryCount]) =
+    NameCloud(tags.map(tc => NameCount(tc.category, tc.count)))
+
   // attachment meta
   def mkAttachmentMeta(rm: RAttachmentMeta): AttachmentMeta =
     AttachmentMeta(

modules/store/src/main/scala/docspell/store/qb/DBFunction.scala

@@ -9,11 +9,11 @@ object DBFunction {
   val countAll: DBFunction = CountAll
 
   def countAs[A](column: Column[A]): DBFunction =
-    Count(column)
+    Count(column, false)
 
   case object CountAll extends DBFunction
 
-  case class Count(column: Column[_]) extends DBFunction
+  case class Count(column: Column[_], distinct: Boolean) extends DBFunction
 
   case class Max(expr: SelectExpr) extends DBFunction
 

modules/store/src/main/scala/docspell/store/qb/DSL.scala

@@ -63,7 +63,10 @@ trait DSL extends DoobieMeta {
     FromExpr.From(sel, alias)
 
   def count(c: Column[_]): DBFunction =
-    DBFunction.Count(c)
+    DBFunction.Count(c, false)
+
+  def countDistinct(c: Column[_]): DBFunction =
+    DBFunction.Count(c, true)
 
   def countAll: DBFunction =
     DBFunction.CountAll

modules/store/src/main/scala/docspell/store/qb/impl/DBFunctionBuilder.scala

@@ -13,8 +13,9 @@ object DBFunctionBuilder extends CommonBuilder {
       case DBFunction.CountAll =>
         sql"COUNT(*)"
 
-      case DBFunction.Count(col) =>
+      case DBFunction.Count(col, distinct) =>
-        sql"COUNT(" ++ column(col) ++ fr")"
+        if (distinct) sql"COUNT(DISTINCT " ++ column(col) ++ fr")"
+        else sql"COUNT(" ++ column(col) ++ fr")"
 
       case DBFunction.Max(expr) =>
         sql"MAX(" ++ SelectExprBuilder.build(expr) ++ fr")"

modules/store/src/main/scala/docspell/store/queries/CategoryCount.scala

@@ -0,0 +1,3 @@
+package docspell.store.queries
+
+final case class CategoryCount(category: String, count: Int)

modules/store/src/main/scala/docspell/store/queries/QItem.scala

@@ -190,9 +190,38 @@ object QItem {
     for {
       count   <- searchCountSummary(today)(q)
       tags    <- searchTagSummary(today)(q)
+      cats    <- searchTagCategorySummary(today)(q)
       fields  <- searchFieldSummary(today)(q)
       folders <- searchFolderSummary(today)(q)
-    } yield SearchSummary(count, tags, fields, folders)
+    } yield SearchSummary(count, tags, cats, fields, folders)
+
+  def searchTagCategorySummary(
+      today: LocalDate
+  )(q: Query): ConnectionIO[List[CategoryCount]] = {
+    val tagFrom =
+      from(ti)
+        .innerJoin(tag, tag.tid === ti.tagId)
+        .innerJoin(i, i.id === ti.itemId)
+
+    val tagCloud =
+      findItemsBase(q.fix, today, 0).unwrap
+        .withSelect(select(tag.category).append(countDistinct(i.id).as("num")))
+        .changeFrom(_.prepend(tagFrom))
+        .changeWhere(c => c && queryCondition(today, q.fix.account.collective, q.cond))
+        .groupBy(tag.category)
+        .build
+        .query[CategoryCount]
+        .to[List]
+
+    // the previous query starts from tags, so items with tag-count=0
+    // are not included they are fetched separately
+    for {
+      existing <- tagCloud
+      allCats  <- RTag.listCategories(q.fix.account.collective)
+      other = allCats.diff(existing.map(_.category))
+    } yield existing ++ other.map(CategoryCount(_, 0))
+
+  }
 
   def searchTagSummary(today: LocalDate)(q: Query): ConnectionIO[List[TagCount]] = {
     val tagFrom =

modules/store/src/main/scala/docspell/store/queries/SearchSummary.scala

@@ -3,6 +3,7 @@ package docspell.store.queries
 case class SearchSummary(
     count: Int,
     tags: List[TagCount],
+    cats: List[CategoryCount],
     fields: List[FieldStats],
     folders: List[FolderCount]
 )

modules/store/src/main/scala/docspell/store/queries/TagCount.scala

@@ -2,4 +2,4 @@ package docspell.store.queries
 
 import docspell.store.records.RTag
 
-case class TagCount(tag: RTag, count: Int)
+final case class TagCount(tag: RTag, count: Int)