Adds possibility to provide the ngsfilter configuration as a CSV file

Former-commit-id: f0fd2cb1a7b149ae2a330edc5087b21be2c4585b

pkg/obiformats/fastseq_write_fastq.go

@@ -11,7 +11,6 @@ import (
 	log "github.com/sirupsen/logrus"
 
 	"git.metabarcoding.org/obitools/obitools4/obitools4/pkg/obiiter"
-	"git.metabarcoding.org/obitools/obitools4/obitools4/pkg/obioptions"
 	"git.metabarcoding.org/obitools/obitools4/obitools4/pkg/obiseq"
 	"git.metabarcoding.org/obitools/obitools4/obitools4/pkg/obiutils"
 )
@@ -34,7 +33,7 @@ func FormatFastq(seq *obiseq.BioSequence, formater FormatHeader) string {
 	)
 }
 
-func FormatFastqBatch(batch obiiter.BioSequenceBatch, quality_shift int,
+func FormatFastqBatch(batch obiiter.BioSequenceBatch,
 	formater FormatHeader, skipEmpty bool) []byte {
 	var bs bytes.Buffer
 	for _, seq := range batch.Slice() {
@@ -75,7 +74,6 @@ func WriteFastq(iterator obiiter.IBioSequence,
 	chunkchan := make(chan FileChunck)
 
 	header_format := opt.FormatFastSeqHeader()
-	quality := obioptions.OutputQualityShift()
 
 	newIter.Add(nwriters)
 
@@ -94,7 +92,7 @@ func WriteFastq(iterator obiiter.IBioSequence,
 		for iterator.Next() {
 			batch := iterator.Get()
 			chunk := FileChunck{
-				FormatFastqBatch(batch, quality, header_format, opt.SkipEmptySequence()),
+				FormatFastqBatch(batch, header_format, opt.SkipEmptySequence()),
 				batch.Order(),
 			}
 			chunkchan <- chunk

pkg/obiformats/ngsfilter_read.go

@@ -2,12 +2,17 @@ package obiformats
 
 import (
 	"bufio"
+	"bytes"
+	"encoding/csv"
 	"fmt"
 	"io"
 	"strings"
 
+	log "github.com/sirupsen/logrus"
+
 	"git.metabarcoding.org/obitools/obitools4/obitools4/pkg/obingslibrary"
 	"git.metabarcoding.org/obitools/obitools4/obitools4/pkg/obiseq"
+	"github.com/gabriel-vasile/mimetype"
 )
 
 func _readLines(reader io.Reader) []string {
@@ -80,7 +85,48 @@ func _parseMainNGSFilter(text string) (obingslibrary.PrimerPair, obingslibrary.T
 		true
 }
 
+func OBIMimeNGSFilterTypeGuesser(stream io.Reader) (*mimetype.MIME, io.Reader, error) {
+
+	// Create a buffer to store the read data
+	buf := make([]byte, 1024*128)
+	n, err := io.ReadFull(stream, buf)
+
+	if err != nil && err != io.ErrUnexpectedEOF {
+		return nil, nil, err
+	}
+
+	// Detect the MIME type using the mimetype library
+	mimeType := mimetype.Detect(buf[:n])
+	if mimeType == nil {
+		return nil, nil, err
+	}
+
+	// Create a new reader based on the read data
+	newReader := io.Reader(bytes.NewReader(buf[:n]))
+
+	if err == nil {
+		newReader = io.MultiReader(newReader, stream)
+	}
+
+	return mimeType, newReader, nil
+}
+
 func ReadNGSFilter(reader io.Reader) (obingslibrary.NGSLibrary, error) {
+	mimetype, newReader, err := OBIMimeNGSFilterTypeGuesser(reader)
+
+	if err != nil {
+		return nil, err
+	}
+
+	log.Infof("NGSFilter configuration mimetype: %s", mimetype.String())
+
+	if mimetype.String() == "text/csv" {
+		return ReadCSVNGSFilter(newReader)
+	}
+
+	return ReadOldNGSFilter(newReader)
+}
+func ReadOldNGSFilter(reader io.Reader) (obingslibrary.NGSLibrary, error) {
 	ngsfilter := obingslibrary.MakeNGSLibrary()
 
 	lines := _readLines(reader)
@@ -126,3 +172,104 @@ func ReadNGSFilter(reader io.Reader) (obingslibrary.NGSLibrary, error) {
 
 	return ngsfilter, nil
 }
+
+func ReadCSVNGSFilter(reader io.Reader) (obingslibrary.NGSLibrary, error) {
+	ngsfilter := obingslibrary.MakeNGSLibrary()
+	file := csv.NewReader(reader)
+
+	file.Comma = ','
+	file.Comment = '#'
+	file.TrimLeadingSpace = true
+	file.ReuseRecord = true
+
+	records, err := file.ReadAll()
+
+	if err != nil {
+		return nil, err
+	}
+
+	log.Info("Read ", len(records), " records")
+	log.Infof("First record: %s", records[0])
+
+	header := records[0]
+	data := records[1:]
+
+	// Find the index of the column named "sample"
+	experimentColIndex := -1
+	sampleColIndex := -1
+	sample_tagColIndex := -1
+	forward_primerColIndex := -1
+	reverse_primerColIndex := -1
+
+	extraColumns := make([]int, 0)
+
+	for i, colName := range header {
+		switch colName {
+		case "experiment":
+			experimentColIndex = i
+		case "sample":
+			sampleColIndex = i
+		case "sample_tag":
+			sample_tagColIndex = i
+		case "forward_primer":
+			forward_primerColIndex = i
+		case "reverse_primer":
+			reverse_primerColIndex = i
+		default:
+			extraColumns = append(extraColumns, i)
+		}
+	}
+
+	if experimentColIndex == -1 {
+		return nil, fmt.Errorf("column 'experiment' not found in the CSV file")
+	}
+
+	if sampleColIndex == -1 {
+		return nil, fmt.Errorf("column 'sample' not found in the CSV file")
+	}
+
+	if sample_tagColIndex == -1 {
+		return nil, fmt.Errorf("column 'sample_tag' not found in the CSV file")
+	}
+
+	if forward_primerColIndex == -1 {
+		return nil, fmt.Errorf("column 'forward_primer' not found in the CSV file")
+	}
+
+	if reverse_primerColIndex == -1 {
+		return nil, fmt.Errorf("column 'reverse_primer' not found in the CSV file")
+	}
+
+	for i, fields := range data {
+		if len(fields) != len(header) {
+			return nil, fmt.Errorf("row %d has %d columns, expected %d", len(data), len(fields), len(header))
+		}
+
+		forward_primer := fields[forward_primerColIndex]
+		reverse_primer := fields[reverse_primerColIndex]
+		tags := _parseMainNGSFilterTags(fields[sample_tagColIndex])
+
+		marker, _ := ngsfilter.GetMarker(forward_primer, reverse_primer)
+		pcr, ok := marker.GetPCR(tags.Forward, tags.Reverse)
+
+		if ok {
+			return ngsfilter,
+				fmt.Errorf("line %d : tag pair (%s,%s) used more than once with marker (%s,%s)",
+					i, tags.Forward, tags.Reverse, forward_primer, reverse_primer)
+		}
+
+		pcr.Experiment = fields[experimentColIndex]
+		pcr.Sample = fields[sampleColIndex]
+		pcr.Partial = false
+
+		if extraColumns != nil {
+			pcr.Annotations = make(obiseq.Annotation)
+			for _, colIndex := range extraColumns {
+				pcr.Annotations[header[colIndex]] = fields[colIndex]
+			}
+		}
+
+	}
+
+	return ngsfilter, nil
+}