Develop Reference

How to reopen a closed Gitlab Projectissue with the python gitlab module

Trying to automate reopening a bunch of issues on a kanban board in Gitlab does not seem to be working as expected. I can edit pretty much everything else: labels, title, description, due date and so on, but changing the state from 'closed' to 'opened' is not doing anything.
Funny enough the save() method, takes some time to complete and is not throwing any error (but it does when I try to update an issue without having changed anything)
import gitlab
gl = gitlab.Gitlab(URL, private_token=TOKEN, api_version = 4)
gl.auth()
Project = gl.projects.get(project_id, lazy=True)
editable_issue = Project.issues.get(issue_id)
I use this to instantiate the gitlab object
This works:
editable_issue.labels.append('some label')
editable_issue.save()
editable_issue.title = title + '\n' + 'edited.'
editable_issue.save()
But this doesn't:
editable_issue.state = 'opened'
editable_issue.save()
I have also tried changing several fields, like so:
editable_issue.state = 'opened'
editable_issue.closed_at = None
editable_issue.save()
But, albeit I receive no error, the task is not being updated.
Is there something I'm overlooking ?
edit
Oh yes, I forgot to mention that the documentation seems to be mentioning state_event as a method, while my object does not know of such a method.
Is this method only there when a projectissue is being created by the api, perhaps ?

As mentioned in the documentation, you must set the state_event attribute, then save. This attribute won't exist at first -- you must create it.
issue.state_event = 'reopen'
issue.save()
Modifying state or closed_at won't work because the issues edit API doesn't accept state or closed_at parameters.
(this is also why state_event doesn't exist on the object at first, because the issue detail API doesn't include state_event in its response -- all the attributes are set dynamically by the response JSON)

Access APScheduler cron trigger field values in python

Similar to this question I want to extract the info of a cron job trigger from APScheduler.
However, I need the "day_of_week" field and not everything. Using
for job in scheduler.get_jobs():
for f in job.trigger.fields:
print(f.name + " " + str(f))
i can see all the fields, e.g. week,hour,day_of_week , but
job.trigger.day_of_week is seemingly 'not an attribute' of the "CronTrigger" object. I'm confused as to what kind of object this job.trigger is and how its fields are packed. I tried to read the code on github, but it is even more puzzling.
How do I extract only the one field day_of_week, and how is this trigger class structured?
Diving deeper I found that
apscheduler.triggers.cron.fields.DayOfWeekField
I can find by indexing the job.trigger.fields[4], which seems really bad style, since it depends on the 'position'of the field. What I get is this DayOfWeekField, from which comically I am not able to retrieve it's value either:
a.get_value
<bound method DayOfWeekField.get_value of DayOfWeekField('day_of_week', '1,2,3,4')>
The structure of the fields is coded here, but I don't know what to do with dateval, the argument of get_value().
Eventually, after hopefully understanding the concept, I want to do
if job-day_of_week contains mon
if job-day_of_week == '*'
print ( job-day_of_week )
I am grateful for any suggestions/hints!

Looking at the code, you should be able to get the day_of_week field without hardcoding the index by using the CronTrigger class's FIELD_NAMES property, e.g.
dow_index = CronTrigger.FIELD_NAMES.index('day_of_week')
dow = job.trigger.fields[dow_index]
Getting the value of the field is a bit more complicated, but it appears that BaseField implements the str function that should give you the value of the expression that created the field as a string that you could parse to find what you want:
dow_value_as_string = str(dow)
if 'mon' in dow_value_as_string:
# do something
if dow_value_as_string = "*":
# do something else

Django: "referenced before assignment" but only for some variables

I'm writing a small app in Django and I'm keeping the state saved in a few variables I declare out of the methods in views.py. Here is the important part of this file:
from app.playerlist import fullList
auc_unsold = fullList[:]
auc_teams = []
auc_in_progress = []
auc_current_turn = -1
print(auc_in_progress)
def auc_action(request):
data = json.loads(request.GET["data"])
# ...
elif data[0] == "start":
random.shuffle(auc_teams)
print(auc_unsold)
print(auc_in_progress)
auc_in_progress = [None, 0, None]
print(auc_in_progress)
The auc_unsold and auc_teams variables work fine; the auc_in_progress variable is not seen by this method, though, giving the error in the title. If I take out the print statement and let this code assign a value to it, the exception will be thrown somewhere else in the code as soon as I use that variable again.
I have tried making another variable and this new one seems to suffer from this problem as well.
What is happening?
Edit: I found a solution: if I write global auc_in_progress just before the print statements, then everything works fine. If I try writing that as I declare the variable above it doesn't work, though, for some reason.
I am unsatisfied with this, because I don't know why this happens and because I dislike using global like that, but eh. Someone has an explanation?

You should absolutely not be doing this, either your original code or your proposed solution with global.
Anything at module level will be shared across requests, not only for the current user but for all users for that process. So everyone will see the same auction, etc.
The reason for your error is because you assign to that variable within your function, which automatically makes it a local variable: see this question for more details. But the solution recommended there, which is the same as your workaround - ie use global - is not appropriate here; you should store the data somewhere specifically associated with the user, eg the session.

Exec and variable scope

Python 3.5.1
Could you have a look at the code and the picture.
This is some Django code, but the question relates to Python.
Well, I get the error:
Exception Value:
name 'objects' is not defined
Exception Location:
/home/michael/workspace/pharchive/pharchive/general/templatetags/md_aux.py in get_aux_info, line 17
Line 17 is context = Context({"objects": objects}). In the picture it is visible.
I tried to add objects=[] a little above. But that didn't help. In this case objects was really empty. So, I commented it out.
Well, in the screenshot we can what we get if we stop at the breakpoint. We can see that objects var is really accessible in the context. And it contains something.
Anyway, I got stuck with variable scope.
I don't understand:
1. Why objects=[] didn't help me.
2. Why I get this error whereas objects is accessible.
Could you give me a kick here?
#register.simple_tag
def get_aux_info(master_document, aux_type):
md = MasterDocument.objects.get(pk=master_document)
template = get_template("general/md_tags.html")
#objects = []
code = "objects = md." + aux_type +"s.all()"
exec(code)
context = Context({"objects": objects})
return template.render(context)

A much easier approach than building and executing the code as a string would be to access the attribute using the built-in getattr function. This makes it clearer what's going on and side-steps your current scope issues entirely:
objects = getattr(md, aux_type + 's').all()
Python has extensive support for this kind of introspection; using exec (and eval) is usually a warning that you're approaching something in the wrong way.

Django method same object not saving

This is something I have been trying to solve for 3 days now and I just can't get my head around it why this is not working.
I have a method that creates a new version of an Object. It used to work, that you would pass in the sou obj. and this would be the source from which a new version is created. You can also pass in a destination, which is not really important in this example. Now I wanted to add locking to this method as we want to add multiple users. So I want to be sure that I always have the most current object from which I create a new one. So I added a line that would just get the newest object. If there is no newer object in the database it would be the same anyway.
def createRevision(request, what, sou, destination=None, ignore = [], **args):
...
if "initial" not in args.keys():
source = get_object_or_404(BaseItem, ppk=sou.ppk, project=sou.project, current=True)
print "------------"
print source == sou
print "------------"
# This outputs True
else:
source = sou
further down in the method I do something like
source.current = False
source.save()
Basically the idea is that I pass in BaseItem and if I don't specify the "initial" keyword then I get the current item from that project with the same ppk (Which is a special random pk in conduction with current). I do this just to be on the save side, that I really have the most current object. And if it is the initial version I just use that one, as there can not be another version.
So now the problem is, that everything works fine if I use sou in this method. I can save it etc .. but as soon as I use source and initial is not in the args it just doesn't save it. The print statement tells me they are the same. Everything I print after the save tells me it has been saved but it just doesn't do it.
source.current = False
source.save()
print "SAVED !!!!"
print source.pk
print source.current
rofl = get_object_or_404(BaseItem, pk=source.pk, project=sou.project)
print rofl.pk
print source.current
outputs the same pk and the same current value but somehow it is not properly saved. As soon as I look into django admin or do a select current = True.
I really don't know what to do anymore.
Why does it work without a problem if I pass in the object into the method but starts to fail when I get the exact same object in the method?
Of course I call the method with the same object:
x = get_object_or_404(BaseItem, ppk=sou.ppk, project=sou.project, current=True)
createRevision(request, "", x)

Thank you pztrick for the hint with the caches. I finally solved it. So the problem was that I was doing:
x = get_object_or_404(BaseItem, ppk=sou.ppk, project=sou.project, current=True)
createRevision(request, "", x)
# .... loads of lines of code
unlock(x)
unlock is a method I wrote that just sets a timestamp so I know no other user is editing it. So now the problem was that I was saving x in createRevision with all the correct data but of course unlock(x) still had a reference to an "old" not updated object and of course was saving it again. Hence it was overwriting my changes in createRevision.
Thank you again to everyone who helped with this.

I think you may be running afoul of model manager caching which is intended to limit database queries. However, by invoking the .all() method on the model manager you force it to hit the databse again.
So, try this: Replace your argument from the BaseItem class to the model manager's .all() QuerySet:
source = get_object_or_404(BaseItem.objects.all(), ppk=sou.ppk, project=sou.project, current=True)
# ...
rofl = get_object_or_404(BaseItem.objects.all(), pk=source.pk, project=sou.project)
get_object_or_404 supports mode classes, model managers, or QuerySets as the first parameter so this is valid.