From d86b934e3f1cd58a2af879cbdc1caa264cbb83c2 Mon Sep 17 00:00:00 2001
From: glichtenstein <gabriel.lichtenstein@gmail.com>
Date: Mon, 22 Jan 2024 14:33:48 -0500
Subject: [PATCH] dropped file logging to file. I need help for that.

---
 subworkflows/nf-core/bcl_demultiplex/main.nf | 54 +++++++++-----------
 1 file changed, 23 insertions(+), 31 deletions(-)

diff --git a/subworkflows/nf-core/bcl_demultiplex/main.nf b/subworkflows/nf-core/bcl_demultiplex/main.nf
index 64e8e90e..6fb108fe 100644
--- a/subworkflows/nf-core/bcl_demultiplex/main.nf
+++ b/subworkflows/nf-core/bcl_demultiplex/main.nf
@@ -7,9 +7,6 @@
 include { BCLCONVERT } from "../../../modules/nf-core/bclconvert/main"
 include { BCL2FASTQ  } from "../../../modules/nf-core/bcl2fastq/main"
 
-// invalid fastq logging
-ch_invalid_fastq = Channel.empty()
-
 workflow BCL_DEMULTIPLEX {
     take:
         ch_flowcell     // [[id:"", lane:""],samplesheet.csv, path/to/bcl/files]
@@ -64,25 +61,11 @@ workflow BCL_DEMULTIPLEX {
             ch_versions = ch_versions.mix(BCL2FASTQ.out.versions)
         }
 
-        // Split the channel into valid and invalid FASTQ entries
-        ch_fastq_meta = generate_fastq_meta(ch_fastq)
-
-        ch_invalid_fastq = ch_fastq_meta
-            .filter { it[0] == null }
-            .map { it[1] } // Extracting the file path
-            .view { println("Invalid FASTQ Detected: $it") }
-            .collectFile(name: 'invalid_fastqs.txt', newLine: true)
-            .view { file ->
-                println("Invalid FASTQs collected in file: ${file.absolutePath}")
-                println("Contents of the file:")
-                file.eachLine { println(it) }
-            }
-
-        ch_valid_fastq = ch_fastq_meta
-            .filter { it[0] != null }
+        // Generate meta for each fastq
+        ch_fastq_with_meta = generate_fastq_meta(ch_fastq)
 
     emit:
-        fastq    = ch_valid_fastq
+        fastq    = ch_fastq_with_meta
         reports  = ch_reports
         stats    = ch_stats
         interop  = ch_interop
@@ -93,6 +76,7 @@ workflow BCL_DEMULTIPLEX {
     FUNCTIONS
 */
 
+// Add meta values to fastq channel and skip invalid FASTQ files
 def generate_fastq_meta(ch_reads) {
     ch_reads.transpose().map { fc_meta, fastq ->
         // Check if the FASTQ file is empty or has invalid content
@@ -115,9 +99,7 @@ def generate_fastq_meta(ch_reads) {
             meta.readgroup = readgroup_from_fastq(fastq)
             meta.readgroup.SM = meta.samplename
         } else {
-            println("Invalid FASTQ file detected: ${fastq}")
-            ch_invalid_fastq << fastq
-            ch_invalid_fastq.view()
+            println("Skipping invalid or empty FASTQ file: ${fastq}")
             fastq = null
         }
 
@@ -133,25 +115,35 @@ def generate_fastq_meta(ch_reads) {
     }
 }
 
+// https://github.com/nf-core/sarek/blob/7ba61bde8e4f3b1932118993c766ed33b5da465e/workflows/sarek.nf#L1014-L1040
 def readgroup_from_fastq(path) {
+    // expected format:
+    // xx:yy:FLOWCELLID:LANE:... (seven fields)
+
     def line
+
     path.withInputStream {
         InputStream gzipStream = new java.util.zip.GZIPInputStream(it)
         Reader decoder = new InputStreamReader(gzipStream, 'ASCII')
         BufferedReader buffered = new BufferedReader(decoder)
         line = buffered.readLine()
-        if (line == null) {
-            println("Warning: Empty or invalid FASTQ file: ${path}")
-            //ch_invalid_fastq << path
-            return null
-        }
-        assert line.startsWith('@') : "FASTQ file does not start with '@': ${path}"
     }
+    assert line.startsWith('@')
     line = line.substring(1)
     def fields = line.split(':')
     def rg = [:]
-    rg.ID = [fields[2],fields[3]].join(".")
-    rg.PU = [fields[2], fields[3], fields[-1] =~ /[GATC+-]/ ? fields[-1] : ""].findAll().join(".")
+
+    // CASAVA 1.8+ format, from  https://support.illumina.com/help/BaseSpace_OLH_009008/Content/Source/Informatics/BS/FileFormat_FASTQ-files_swBS.htm
+    // "@<instrument>:<run number>:<flowcell ID>:<lane>:<tile>:<x-pos>:<y-pos>:<UMI> <read>:<is filtered>:<control number>:<index>"
+    sequencer_serial = fields[0]
+    run_nubmer       = fields[1]
+    fcid             = fields[2]
+    lane             = fields[3]
+    index            = fields[-1] =~ /[GATC+-]/ ? fields[-1] : ""
+
+    rg.ID = [fcid,lane].join(".")
+    rg.PU = [fcid, lane, index].findAll().join(".")
     rg.PL = "ILLUMINA"
+
     return rg
 }