You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Passing a schema string that contains more than the expected number of format specifiers (precisely one %d) causes an exception, but leaves the process running and stuck somewhere.
[eyh46967@pc0103 PandABlocks-client]$ pipenv run pandablocks hdf --arm 172.23.252.201 %s/%s_%d.h5
Exception in thread Thread-2:
Traceback (most recent call last):
File "/dls_sw/prod/tools/RHEL7-x86_64/Python3/3-7-2dls1/prefix/lib/python3.7/threading.py", line 917, in _bootstrap_inner
self.run()
File "/home/eyh46967/dev/PandABlocks-client/pandablocks/hdf.py", line 48, in run
data = func(data)
File "/home/eyh46967/dev/PandABlocks-client/pandablocks/hdf.py", line 96, in open_file
file_path = self.scheme % self.num
TypeError: not enough arguments for format string
The command line help for the schema parameter should probably also be changed to explicitly say "exactly 1 %d", as opposed to its current description: Uses the filename pattern specified by SCHEME, including %d for scan number starting from 1
The text was updated successfully, but these errors were encountered:
Passing a schema string that contains more than the expected number of format specifiers (precisely one %d) causes an exception, but leaves the process running and stuck somewhere.
The command line help for the schema parameter should probably also be changed to explicitly say "exactly 1 %d", as opposed to its current description:
Uses the filename pattern specified by SCHEME, including %d for scan number starting from 1
The text was updated successfully, but these errors were encountered: