| |
@@ -33,6 +33,24 @@
|
| |
# Define functions
|
| |
|
| |
# This function needs a dry-run like option
|
| |
+ def buildmeoutput(cmd, action, pkg, env, cwd=workdir):
|
| |
+ """Simple function to run a command and return 0 for success, 1 for
|
| |
+ failure. It also writes the taskID and name to a file and console.
|
| |
+ cmd is the command and arguments, action is aname for the action (for
|
| |
+ logging), pkg is the name of the packagebeing operated on, env is the
|
| |
+ environment dict, and cwd is where the script should be executed from."""
|
| |
+
|
| |
+ try:
|
| |
+ output = subprocess.check_output(cmd, env=env, cwd=cwd).decode('utf-8').split()
|
| |
+ with open(workdir+"/taskID_file", 'a') as task_file:
|
| |
+ task_file.write('%s %s\n' % (pkg, output[2]))
|
| |
+ sys.stdout.write(' Successful submission: %s taskID: %s\n' % (pkg, output[2]))
|
| |
+ except subprocess.CalledProcessError as e:
|
| |
+ sys.stderr.write('%s failed %s: %s\n' % (pkg, action, e))
|
| |
+ return 1
|
| |
+ return 0
|
| |
+
|
| |
+ # This function needs a dry-run like option
|
| |
def runme(cmd, action, pkg, env, cwd=workdir):
|
| |
"""Simple function to run a command and return 0 for success, 1 for
|
| |
failure. cmd is a list of the command and arguments, action is a
|
| |
@@ -176,5 +194,5 @@
|
| |
# build
|
| |
build = [koji_bin, 'build', '--nowait', '--background', '--fail-fast', massrebuild['target'], url]
|
| |
print('Building %s' % name)
|
| |
- runme(build, 'build', name, enviro,
|
| |
+ buildmeoutput(build, 'build', name, enviro,
|
| |
cwd=os.path.join(workdir, name))
|
| |
tdawson
commented 3 years ago
The current way of checking if a package has failed or not, misses failures due to srpm build failures.
The only way to catch those is by checking taskID's directly.
This patch prints out the package and task id to both console, and a file.
I did some time tests and found that the increase in time was undetectable.
Signed-off-by: Troy Dawson tdawson@redhat.com