X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/6c3e6273fa29b8004c17a7b619b3818f8e8bd84e..b6888eaa68820d4107a73fb1fea136c00866f915:/crunch_scripts/hash

diff --git a/crunch_scripts/hash b/crunch_scripts/hash
index 6d6c88f64d..3c157163d1 100755
--- a/crunch_scripts/hash
+++ b/crunch_scripts/hash
@@ -2,7 +2,6 @@
 
 import arvados
 import hashlib
-import re
 
 arvados.job_setup.one_task_per_input_file(if_sequence=0, and_end_task=True)
 
@@ -11,22 +10,25 @@ this_task = arvados.current_task()
 this_task_input = this_task['parameters']['input']
 
 if 'algorithm' in this_job['script_parameters']:
-    alg = this_job['script_parameters']
+    alg = this_job['script_parameters']['algorithm']
 else:
     alg = 'md5'
 digestor = hashlib.new(alg)
 
-input_stream = arvados.DataReader(this_task_input)
+input_file = list(arvados.CollectionReader(this_task_input).all_files())[0]
 while True:
-    buf = input_stream.read(2**20)
+    buf = input_file.read(2**20)
     if len(buf) == 0:
         break
     digestor.update(buf)
 
 hexdigest = digestor.hexdigest()
-file_name = re.sub(r'^[^/]+/', '', this_task_input)
 
-input_stream.close()
-
-this_task.set_output("%s %s\n" % (hexdigest, file_name))
+file_name = input_file.name()
+if input_file.stream_name() != '.':
+    file_name = os.join(input_file.stream_name(), file_name)
 
+out = arvados.CollectionWriter()
+out.set_current_file_name("md5sum.txt")
+out.write("%s %s\n" % (hexdigest, file_name))
+this_task.set_output(out.finish())