Use scala-csv

CHANGELOG.md

@@ -1,5 +1,14 @@
 # Changelog
 
+## 3.8.0
+* Replace kantan.csv with scala-csv
+
+## 3.7.2
+* Parse barcodes from read IDs in demultiplexed mode
+
+## 3.7.1
+* Adjust handling of command-line arguments in demultiplexed FASTQ file case
+
 ## 3.7.0
 * Support for processing demultiplexed FASTQ files
 

build.sbt

@@ -13,22 +13,21 @@ lazy val versions = new {
   val acyclic = "0.2.1"
   val betterFiles = "3.9.2"
   val betterMonadicFor = "0.3.1"
-  val catsEffect3 = "3.5.1"
+  val catsEffect3 = "3.5.2"
   val cats = "2.10.0"
-  val commonsIo = "2.13.0"
-  val commonsText = "1.10.0"
+  val commonsIo = "2.15.1"
+  val commonsText = "1.11.0"
   val commonsMath3 = "3.6.1"
   val fastutil = "8.5.12"
-  val fs2 = "3.8.0"
-  val kantanCodecs = "0.5.3"
-  val kantanCsv = "0.7.0"
+  val fs2 = "3.9.3"
   val log4s = "1.10.0"
-  val logback = "1.2.11"
+  val logback = "1.2.13"
   val munit = "0.7.29"
   val munitCatsEffect3 = "1.0.7"
   val samTools = "3.0.5"
   val scalaCheck = "1.17.0"
-  val scalaTest = "3.2.16"
+  val scalaCsv = "1.3.10"
+  val scalaTest = "3.2.17"
   val scalaTestPlusScalaCheck = "3.2.2.0"
   val scopt = "4.1.0"
   val slf4j = "1.7.36"
@@ -45,12 +44,11 @@ lazy val libraries = new {
   val fastutil = "it.unimi.dsi" % "fastutil" % versions.fastutil
   val fs2Core = "co.fs2" %% "fs2-core" % versions.fs2
   val fs2Io = "co.fs2" %% "fs2-io" % versions.fs2
-  val kantanCodecs = "com.nrinaudo" %% "kantan.codecs" % versions.kantanCodecs
-  val kantanCsv = "com.nrinaudo" %% "kantan.csv" % versions.kantanCsv
   val log4s = "org.log4s" %% "log4s" % versions.log4s
   val logbackCore = "ch.qos.logback" % "logback-core" % versions.logback
   val logbackClassic = "ch.qos.logback" % "logback-classic" % versions.logback
   val samtools = "com.github.samtools" % "htsjdk" % versions.samTools
+  val scalaCsv = "com.github.tototoshi" %% "scala-csv" % versions.scalaCsv
   val scopt = "com.github.scopt" %% "scopt" % versions.scopt
   val slf4j = "org.slf4j" % "slf4j-api" % versions.slf4j
 
@@ -71,12 +69,11 @@ lazy val dependencies =
     libraries.commonsIo,
     libraries.commonsMath3,
     libraries.fastutil,
-    libraries.kantanCodecs,
-    libraries.kantanCsv,
     libraries.log4s,
     libraries.logbackCore % Runtime,
     libraries.logbackClassic % Runtime,
     libraries.samtools,
+    libraries.scalaCsv,
     libraries.scopt,
     libraries.slf4j,
     libraries.betterFiles % Test,
@@ -107,8 +104,11 @@ lazy val headerSettings = List(
 lazy val assemblySettings = List(
   assembly / assemblyJarName := "../bin/poolq3.jar",
   assembly / assemblyMergeStrategy := {
-    case "logback.xml"      => MergeStrategy.first
-    case "logback-test.xml" => MergeStrategy.discard
+    case "logback.xml"                                                 => MergeStrategy.first
+    case "logback-test.xml"                                            => MergeStrategy.discard
+    case PathList("module-info.class")                                 => MergeStrategy.discard
+    case PathList("META-INF", "versions", xs @ _, "module-info.class") => MergeStrategy.discard
+    case "module-info.class"                                           => MergeStrategy.first
     case x =>
       val old = (assembly / assemblyMergeStrategy).value
       old(x)

project/build.properties

@@ -1 +1 @@
-sbt.version=1.9.6
+sbt.version=1.9.8

src/main/scala/org/broadinstitute/gpp/poolq3/parser/ReferenceData.scala

@@ -10,8 +10,7 @@ import java.nio.file.Path
 
 import scala.util.Using
 
-import kantan.csv._
-import kantan.csv.ops._
+import com.github.tototoshi.csv._
 import org.apache.commons.io.ByteOrderMark
 import org.apache.commons.io.input.BOMInputStream
 import org.broadinstitute.gpp.poolq3.reports.{GctDialect, PoolQ2Dialect, ReportsDialect}
@@ -77,34 +76,34 @@ object ReferenceData {
         .setInclude(false)
         .get()
       val br = new BufferedReader(new InputStreamReader(in))
-      val delimiter = guessDelimiter(br)
-      val config =
-        CsvConfiguration(delimiter, quote, CsvConfiguration.QuotePolicy.WhenNeeded, CsvConfiguration.Header.None)
+      val guessedDelimiter = guessDelimiter(br)
+      implicit object CSVFormat extends DefaultCSVFormat {
+        override val delimiter = guessedDelimiter
+        override val quoteChar: Char = quote
+      }
       skipHeader(br, LineRegex)
-      val reader = br.asCsvReader[List[String]](config)
-      val barcodes = reader.map {
-        case Right(xs) =>
-          xs match {
-            case barcodeRaw :: idRaw :: _ =>
-              // if the CSV parser leaves spaces, we should remove them
-              val barcode = barcodeRaw.trim()
-              val id = idRaw.trim()
-
-              // N.B. empty IDs are commonly used and must be supported; as long as the barcode is a non-empty, valid
-              // DNA string, we must accept the row. However, sometimes Excel leaves empty lines in exported CSV; as
-              // long as *both* the barcode and ID are empty, it's safe to just skip the row. For now we'll be paranoid
-              // and reject cases where the barcode is empty but the ID is non-empty
-              if (barcode.isEmpty && id.isEmpty) None
-              else if (isReferenceBarcode(barcode)) Some(ReferenceEntry(barcode, id))
-              else throw InvalidFileException(file, s"Invalid DNA barcode '$barcode' for ID '$id'")
-            case _ =>
-              throw InvalidFileException(
-                file,
-                s"Incorrect number of columns. At least 2 required, got: ${xs.length}: $xs"
-              )
-          }
-        case Left(value) => throw InvalidFileException(file, s"Unable to parse data ${value.getMessage}")
-      }.toList
+      val rows = CSVReader.open(br).all()
+      val barcodes = rows.map { case xs =>
+        xs match {
+          case barcodeRaw :: idRaw :: _ =>
+            // if the CSV parser leaves spaces, we should remove them
+            val barcode = barcodeRaw.trim()
+            val id = idRaw.trim()
+
+            // N.B. empty IDs are commonly used and must be supported; as long as the barcode is a non-empty, valid
+            // DNA string, we must accept the row. However, sometimes Excel leaves empty lines in exported CSV; as
+            // long as *both* the barcode and ID are empty, it's safe to just skip the row. For now we'll be paranoid
+            // and reject cases where the barcode is empty but the ID is non-empty
+            if (barcode.isEmpty && id.isEmpty) None
+            else if (isReferenceBarcode(barcode)) Some(ReferenceEntry(barcode, id))
+            else throw InvalidFileException(file, s"Invalid DNA barcode '$barcode' for ID '$id'")
+          case _ =>
+            throw InvalidFileException(
+              file,
+              s"Incorrect number of columns. At least 2 required, got: ${xs.length}: $xs"
+            )
+        }
+      }
 
       if (barcodes.isEmpty) {
         throw InvalidFileException(file, "Empty reference file")

version.sbt

@@ -1 +1 @@
-ThisBuild / version := "3.7.3-SNAPSHOT"
+ThisBuild / version := "3.8.0-SNAPSHOT"