ported to OSX. not tested; almost definitely needs more work.#1
Conversation
|
Will test over lunch. |
|
You can hold of on that, looks like I indeed broke some stuff :) fixing ... |
…error in bagit lib
|
hey Mark, go ahead and test now -- all should be fine, except it's throwing a strange error from the bagit lib: ERROR:root:expected string or buffer it appears to be creating the bag succesfully prior to throwing the error, too... any chance you changed the tag behaviour when you changed the tag behaviour last night when you changed the directory modification behaviour? |
There was a problem hiding this comment.
I didn't get as far as creating a Bag. Running it under Ubuntu results in the following error when I try::
Traceback (most recent call last): File "createbag.py", line 169, in on_folder_clicked "The Bag for folder %s has been created." % bag_dir) NameError: global name 'bag_dir' is not defined
There was a problem hiding this comment.
Changing line 169 to:
"The Bag for folder %s has been created." % folder_picker_dialog.get_filename()
Appears to have eliminated the error. I didn't get the tag error you reported. Did you add some custom tags?
There was a problem hiding this comment.
I didn't, no, which is weird. I also committed a different fix for the error a few minutes ago that's more consistent with the way you had it working previously, but either way.
|
Re. the tag error, all I did last night was make the 'create_bag_in' tag enabled (it was commented out previous to that). |
|
just pushed a fix for that -- oddly, that wasn't getting thrown on my system, probably because the tag error was coming up first -- which means that one might be local to my environment somehow. |
There was a problem hiding this comment.
This is pretty old - does it work on modern Macs?
There was a problem hiding this comment.
sure does! good old Cocoadialog.
…tory since I didn't realize that's where bags will end up without being moved elsewhere
|
most recent commit "finishes" the OSX version (it creates a bag and provides a success message), and I think all other issues should be fixed at this point! I commented out the cleaning of the /tmp directory I'd added earlier today; originally, I put that in there because I noticed while testing that if I kept trying to create the same bag, it would keep creating arbitrarily many /data directories in the same subfolder in /tmp. I think the script should probably finish by moving the /tmp bag to the desktop or the same directory as the original one so that /tmp can be cleaned -- what do you think? |
|
Awesome, will test tonight. The behavior you're describing was not intended so I'd classify it as a bug. Let me get through testing your PR and we'll take a look at that (and another bug I found today). |
There was a problem hiding this comment.
This assumes that the script is installed in a specific location. Make relative to current script with os.path.dirname(os.path.realpath(file))?
There was a problem hiding this comment.
Good call. I always forget that one -- though we're still kind of hamstrung by the need to provide straightforward automator syntax in the readme.
There was a problem hiding this comment.
Sorry, I didn't see that the path needs to be hard coded because of https://github.com/mjordan/createbag/pull/1/files#diff-04c6e90faac2675aa89e2176d2eec7d8R71. But is it essential that the application gets installed in the user's home directory? Is there a way to share one installation among many Macs?
There was a problem hiding this comment.
At that point in pretty sure we're onto thinking about app bundling (which is very easy to do on OSX, if there were a way to register automator services programmatically too)
There was a problem hiding this comment.
I think we should plan for bundling on all platforms. Can you figure out if that's feasible on OS X and if so, we can reopen this issue. For now let's go with the hard-coded installation directory and make sure it's documented.
ported to OSX. not tested; almost definitely needs more work.
|
Just merged. Awesome. I'm going to add some comments to the main script to clarify my understanding of the logic, now that there are long-span if/else blocks to accommodate Darwin.... Then, a few bug fixes, including a weird one where BagIt tags are being added using config options.... |
|
Sounds good! Looking forward to working more on this. Let me know if anything I did is weird/unclear. |
2 participants
hey Mark,
rearchitected this to make it (hopefully) work cross-platform using the Automator to add a context menu entry and Cocoadialog in place of Gtk. not yet tested, but if you want to ensure I haven't broken it on Linux/Windows, that wouldn't be a bad idea... (I'll check myself later on).