From e8e584d451303efcd865c45389674b8264867e49 Mon Sep 17 00:00:00 2001
From: Chris Wright <chris.wright@nanoporetech.com>
Date: Fri, 19 Mar 2021 11:21:32 +0000
Subject: [PATCH] Set fastq option straight

---
 README.md           |  2 +-
 bin/read_lengths.py |  9 +++++++--
 main.nf             | 37 ++++++++++++++++++++++++++++++++++++-
 3 files changed, 44 insertions(+), 4 deletions(-)

diff --git a/README.md b/README.md
index 175cdb3..ed24e63 100644
--- a/README.md
+++ b/README.md
@@ -99,7 +99,7 @@ OUTPUT=output
 nextflow run epi2me-labs/wf-template \
     -w ${OUTPUT}/workspace \
     -profile standard \
-    --fastq test_data/reads.fq.gz \
+    --fastq test_data \
     --out_dir ${OUTPUT}
 ```
 
diff --git a/bin/read_lengths.py b/bin/read_lengths.py
index aec06ca..669ed47 100755
--- a/bin/read_lengths.py
+++ b/bin/read_lengths.py
@@ -23,12 +23,17 @@ def main():
     """Run entry point."""
     parser = argparse.ArgumentParser()
     parser.add_argument(
-        "directory", help="Directory containing .fastq(.gz) files")
+        "fastq", help="Directory containing .fastq(.gz) files, or single fastq")
     parser.add_argument(
         "output", help="Output file")
     args = parser.parse_args()
 
-    fastqs = glob.glob(os.path.join(args.directory, "*.fastq*"))
+    if os.path.isfile(args.fastq):
+        fastqs = [args.fastq]
+    elif os.path.isdir(args.fastq):
+        fastqs = glob.glob(os.path.join(args.directory, "*.fastq*"))
+    else:
+        raise IOError("fastq argument should be directory of file.")
     reads = itertools.chain.from_iterable(
         pysam.FastxFile(fname) for fname in fastqs)
 
diff --git a/main.nf b/main.nf
index c536fe6..fc3fec8 100644
--- a/main.nf
+++ b/main.nf
@@ -27,6 +27,40 @@ Script Options:
 }
 
 
+process concatFastq {
+    // concatenate fastq and fastq.gz in a dir
+
+    label "pysam"
+    cpus 1
+    input:
+        file "input"
+    output:
+        file "reads.fastq.gz"
+
+    shell:
+    '''
+#!/usr/bin/env python
+from glob import glob
+import gzip
+import itertools
+import os
+import pysam
+
+# we use pysam just because it will read both fastq and fastq.gz
+# and we don't have to worry about having a combination or not
+with gzip.open("reads.fastq.gz", "wt") as fh:
+    files = itertools.chain(
+        glob("input/*.fastq"), glob("input/*.fastq.gz"))
+    records = itertools.chain.from_iterable(
+        pysam.FastxFile(fn) for fn in files) 
+    for rec in records:
+        annot = " {}".format(rec.comment) if rec.comment else ""
+        qual = rec.quality if rec.quality else "+"*len(rec.sequence)
+        fh.write("@{}{}\\n{}\\n+\\n{}\\n".format(rec.name, annot, rec.sequence, qual))
+    '''
+}
+
+
 process readSeqs {
     // Just write a file with sequence lengths
     label "pysam"
@@ -75,6 +109,7 @@ workflow pipeline {
     take:
         reads
     main:
+        reads = concatFastq(reads)
         summary = readSeqs(reads)
         report = makeReport(summary)
     emit:
@@ -99,7 +134,7 @@ workflow {
 
     reads = file("$params.fastq/*.fastq*", type: 'file', maxdepth: 1)
     if (reads) {
-        reads = Channel.fromPath(params.fastq, type: 'dir', maxDepth: 1)
+        reads = Channel.fromPath(params.fastq, type: 'dir', checkIfExists: true)
         results = pipeline(reads)
         output(results)
     } else {