5787: run-command uploading uses new collection API and supports symlinks into Keep

[arvados.git] / crunch_scripts / run-command

diff --git a/crunch_scripts/run-command b/crunch_scripts/run-command
index 13ae918895d6192553107ae00273e247c6eb4c34..ae2233e4913dc9c51591d5935b17ae6585c1ff83 100755 (executable)
--- a/crunch_scripts/run-command
+++ b/crunch_scripts/run-command
@@ -57,12 +57,10 @@ else:
     jobp = json.loads(args.script_parameters)
     os.environ['JOB_UUID'] = 'zzzzz-8i9sb-1234567890abcde'
     os.environ['TASK_UUID'] = 'zzzzz-ot0gb-1234567890abcde'
-    os.environ['CRUNCH_SRC'] = '/tmp/crunche-src'
+    os.environ['CRUNCH_SRC'] = '/tmp/crunch-src'
     if 'TASK_KEEPMOUNT' not in os.environ:
         os.environ['TASK_KEEPMOUNT'] = '/keep'
 
-links = []
-
 def sub_tmpdir(v):
     return os.path.join(arvados.current_task().tmpdir, 'tmpdir')
 
@@ -415,27 +413,23 @@ signal.signal(signal.SIGINT, signal.SIG_DFL)
 signal.signal(signal.SIGTERM, signal.SIG_DFL)
 signal.signal(signal.SIGQUIT, signal.SIG_DFL)
 
-for l in links:
-    os.unlink(l)
-
 logger.info("the following output files will be saved to keep:")
 
-subprocess.call(["find", ".", "-type", "f", "-printf", "run-command: %12.12s %h/%f\\n"], stdout=sys.stderr)
+subprocess.call(["find", ".", "-printf", "run-command: %12.12s %h/%f\\n"], stdout=sys.stderr, cwd=outdir)
 
 logger.info("start writing output to keep")
 
-if "task.vwd" in taskp:
-    if "task.foreach" in jobp:
-        # This is a subtask, so don't merge with the original collection, that will happen at the end
-        outcollection = vwd.checkin(subst.do_substitution(taskp, taskp["task.vwd"]), outdir, merge=False).manifest_text()
-    else:
-        # Just a single task, so do merge with the original collection
-        outcollection = vwd.checkin(subst.do_substitution(taskp, taskp["task.vwd"]), outdir, merge=True).manifest_text()
-else:
-    outcollection = robust_put.upload(outdir, logger)
+if "task.vwd" in taskp and "task.foreach" in jobp:
+    for root, dirs, files in os.walk(outdir):
+        for f in files:
+            s = os.lstat(os.path.join(root, f))
+            if stat.S_ISLNK(s.st_mode):
+                os.unlink(os.path.join(root, f))
+
+outcollection = vwd.checkin(outdir).manifest_text()
 
-# Success if no non-zero return codes
-success = any(rcode) and not any([status != 0 for status in rcode.values()])
+# Success if we ran any subprocess, and they all exited 0.
+success = rcode and all(status == 0 for status in rcode.itervalues())
 
 api.job_tasks().update(uuid=arvados.current_task()['uuid'],
                                      body={
@@ -444,4 +438,4 @@ api.job_tasks().update(uuid=arvados.current_task()['uuid'],
                                          'progress':1.0
                                      }).execute()
 
-sys.exit(rcode)
+sys.exit(0 if success else 1)