stats

backend/src/main/kotlin/org/loculus/backend/api/StatisticsResponse.kt

@@ -0,0 +1,20 @@
+package org.loculus.backend.api
+
+/**
+ * Response model for pipeline statistics endpoint.
+ * Shows the count of sequences by organism and pipeline version.
+ */
+data class PipelineStatisticsResponse(
+    // Map of organism name to statistics for that organism
+    val statistics: Map<String, OrganismPipelineStatistics>
+)
+
+/**
+ * Statistics for a specific organism
+ */
+data class OrganismPipelineStatistics(
+    // Total number of sequences for this organism
+    val totalSequences: Int,
+    // Count of sequences by pipeline version
+    val sequencesByPipelineVersion: Map<Long, Int>
+)

backend/src/main/kotlin/org/loculus/backend/controller/StatisticsController.kt

@@ -0,0 +1,27 @@
+package org.loculus.backend.controller
+
+import io.swagger.v3.oas.annotations.Operation
+import io.swagger.v3.oas.annotations.security.SecurityRequirement
+import org.loculus.backend.api.PipelineStatisticsResponse
+import org.loculus.backend.service.submission.SubmissionDatabaseService
+import org.springframework.http.MediaType
+import org.springframework.web.bind.annotation.GetMapping
+import org.springframework.web.bind.annotation.RequestMapping
+import org.springframework.web.bind.annotation.RestController
+
+@RestController
+@RequestMapping("/statistics")
+@SecurityRequirement(name = "bearerAuth")
+class StatisticsController(
+    private val submissionDatabaseService: SubmissionDatabaseService,
+) {
+    @Operation(
+        description = "Get statistics on sequences processed by pipeline version for each organism",
+    )
+    @GetMapping("/pipeline-versions", produces = [MediaType.APPLICATION_JSON_VALUE])
+    fun getPipelineStatistics(): PipelineStatisticsResponse {
+        return PipelineStatisticsResponse(
+            statistics = submissionDatabaseService.getPipelineStatistics()
+        )
+    }
+}

backend/src/main/kotlin/org/loculus/backend/service/submission/SubmissionDatabaseService.kt

@@ -37,6 +37,7 @@ import org.jetbrains.exposed.sql.transactions.transaction
 import org.jetbrains.exposed.sql.update
 import org.jetbrains.exposed.sql.vendors.ForUpdateOption.PostgreSQL.ForUpdate
 import org.jetbrains.exposed.sql.vendors.ForUpdateOption.PostgreSQL.MODE
+import org.loculus.backend.api.OrganismPipelineStatistics
 import org.loculus.backend.api.AccessionVersion
 import org.loculus.backend.api.AccessionVersionInterface
 import org.loculus.backend.api.AccessionVersionOriginalMetadata
@@ -1121,6 +1122,43 @@ class SubmissionDatabaseService(
         SequenceEntriesTable.distinctOrganisms().map { organismName ->
             Pair(organismName, useNewerProcessingPipelineIfPossible(organismName))
         }.toMap()
+
+    /**
+     * Gets statistics on sequences grouped by organism and pipeline version
+     */
+    fun getPipelineStatistics(): Map<String, OrganismPipelineStatistics> {
+        val statistics = mutableMapOf<String, OrganismPipelineStatistics>()
+
+        SequenceEntriesTable.distinctOrganisms().forEach { organismName ->
+            val organism = Organism(organismName)
+
+            // Count total sequences for this organism
+            val totalSequences = SequenceEntriesView
+                .select(Count(stringLiteral("*")))
+                .where { SequenceEntriesView.organismIs(organism) }
+                .first()[Count(stringLiteral("*"))].toInt()
+
+            // Count sequences by pipeline version
+            val pipelineVersionColumn = SequenceEntriesView.pipelineVersionColumn
+            val countColumn = Count(stringLiteral("*"))
+
+            val pipelineVersionCounts = SequenceEntriesView
+                .select(pipelineVersionColumn, countColumn)
+                .where { 
+                    SequenceEntriesView.organismIs(organism) and
+                    (pipelineVersionColumn.isNotNull())
+                }
+                .groupBy(pipelineVersionColumn)
+                .associate { it[pipelineVersionColumn]!! to it[countColumn].toInt() }
+
+            statistics[organismName] = OrganismPipelineStatistics(
+                totalSequences = totalSequences,
+                sequencesByPipelineVersion = pipelineVersionCounts
+            )
+        }
+
+        return statistics
+    }
 
     /**
      * Looks for new preprocessing pipeline version with [findNewPreprocessingPipelineVersion];