fix: batch submission now does not require user to write to script first

aviary/aviary.py

@@ -1078,14 +1078,14 @@ def main():
 
                                              aviary batch -f batch_file.tsv -t 32 -o batch_test
                                              
-                                             An example batch file can be found at: 
+                                             An example batch file can be found at: https://rhysnewell.github.io/aviary/examples
 
                                              ''')
 
     batch_options.add_argument(
         '-f', '--batch_file', '--batch-file',
         help='The tab or comma separated batch file containing the input samples to assemble and/or recover MAGs from. \n'
-             'An example batch file can be found at XXX. The heading line is required. \n'
+             'An example batch file can be found at https://rhysnewell.github.io/aviary/examples. The heading line is required. \n'
              'The number of reads provided to each sample is flexible as is the type of assembly being performed (if any). \n'
              'Multiple reads can be supplied by providing a comma-separated list (surrounded by double quotes \"\" if using a \n'
              'comma separated batch file) within the specific read column.',

aviary/modules/processor.py

@@ -468,6 +468,8 @@ def run_workflow(self, cores=16, profile=None, cluster_retries=None,
                 resources=f"--resources mem_mb={int(self.max_memory)*1024} {self.resources}" if not dryrun else ""
             )
 
+            logging.debug(f"Command: {cmd}")
+
             if write_to_script is not None:
                 write_to_script.append(cmd)
                 continue
@@ -495,20 +497,31 @@ def process_batch(args, prefix):
 
     logging.info(f"Reading batch file: {args.batch_file}")
 
-    header=0
+    header=None
+    separator=' '
     with open(args.batch_file, mode='r') as check_batch:
         for line in check_batch.readlines():
-            if "sample\tshort_reads_1\tshort_reads_2\tlong_reads\tlong_read_type\tassembly\tcoassemble" in line \
-                or "sample,short_reads_1,short_reads_2,long_reads,long_read_type,assembly,coassemble" in line \
-                or "sample  short_reads_1   short_reads_2   long_reads      long_read_type  assembly        coassemble" in line \
-                or "sample short_reads_1 short_reads_2 long_reads long_read_type assembly coassemble" in line:
-               header=1
-               logging.debug("Inferred header")
-            else:
-                logging.debug("No heading inferred.")
+            line = line.strip()
+            for sep in ['\t', ',', ' ']:
+                separated = line.split(sep)
+                print(separated)
+                if separated == ['sample', 'short_reads_1', 'short_reads_2', 'long_reads', 'long_read_type', 'assembly', 'coassemble']:
+                    header=0
+                    separator=sep
+                    logging.debug("Inferred header")
+                    break
+                elif len(separated) >= 7:
+                    header=None
+                    separator=sep
+                    logging.debug("Inferred no header")
+                    break
+            if header is None:
+                logging.debug("No header found")
             break
-
-    batch = pd.read_csv(args.batch_file, sep=None, engine='python', skiprows=header)
+    if header is not None:
+        batch = pd.read_csv(args.batch_file, sep=separator, engine='python', header=header)
+    else:
+        batch = pd.read_csv(args.batch_file, sep=separator, engine='python', names=['sample', 'short_reads_1', 'short_reads_2', 'long_reads', 'long_read_type', 'assembly', 'coassemble'])
     if len(batch.columns) != 7:
         logging.critical(f"Batch file contains incorrect number of columns ({len(batch.columns)}). Should contain 7.")
         logging.critical(f"Current columns: {batch.columns}")
@@ -525,11 +538,15 @@ def process_batch(args, prefix):
 
     try:
         script_file = args.write_script
-        write_to_script = []
     except AttributeError:
         script_file = None
-        write_to_script = None
+
+    write_to_script = None
+    if script_file is not None:
+        write_to_script = []
 
+    print(script_file)
+    print(write_to_script)
     runs = []
     args.interleaved = "none" # hacky solution to skip attribute error
     args.coupled = "none"

test/example_batch.tsv

@@ -0,0 +1,3 @@
+sample short_reads_1 short_reads_2 long_reads long_read_type assembly coassemble
+sample_1 test/data/wgsim.1.fq.gz test/data/wgsim.2.fq.gz NA ont NA NA
+sample_2 NA NA test/data/pbsim.fq.gz pacbio NA NA