Hi, I just installed and got this error trying nbcat:

Traceback (most recent call last): File "/Users/cherny/miniconda3/envs/pandasnlp/bin/nbcat", line 8, in <module> sys.exit(cat()) File "/Users/cherny/miniconda3/envs/pandasnlp/lib/python3.7/site-packages/click/core.py", line 764, in __call__ return self.main(*args, **kwargs) File "/Users/cherny/miniconda3/envs/pandasnlp/lib/python3.7/site-packages/click/core.py", line 717, in main rv = self.invoke(ctx) File "/Users/cherny/miniconda3/envs/pandasnlp/lib/python3.7/site-packages/click/core.py", line 956, in invoke return ctx.invoke(self.callback, **ctx.params) File "/Users/cherny/miniconda3/envs/pandasnlp/lib/python3.7/site-packages/click/core.py", line 555, in invoke return callback(*args, **kwargs) File "/Users/cherny/miniconda3/envs/pandasnlp/lib/python3.7/site-packages/click/decorators.py", line 17, in new_func return f(get_current_context(), *args, **kwargs) File "/Users/cherny/miniconda3/envs/pandasnlp/lib/python3.7/site-packages/nbcommands/_cat.py", line 42, in cat click.echo("\n".join(display(merged.cells))) File "/Users/cherny/miniconda3/envs/pandasnlp/lib/python3.7/site-packages/nbcommands/terminal.py", line 15, in display cell["execution_count"] if cell["execution_count"] is not None else " " KeyError: 'execution_count'

Any ideas? Python 3.7, nbcommands-0.3.2

@vinayak-mehta I'm also wondering whether you'd like to consider distributing the package through conda-forge?

If you are interested in doing so, but don't have the bandwidth to begin, I'm happy to kickstart the process for you. Maintenance post-deployment to conda-forge is very easy, as you just have to take care of the PyPI version, and a bot will do what's needed to update on conda-forge and ping you (and me, I'm happy to look out for the conda-forge package updates too).

Please let me know!

• when a directory is given, grep all notebooks in the directory
• set type to Path to get errors from click when paths don't exist
• set required=True for "missing argument" error when no path is specified, instead of doing nothing

Looking at the docs, I was confused why nbtouch would exist when touch works exactly the same for existing notebook files, but on looking at the code I saw that it creates empty notebook files, which is useful! This behavior wasn't documented as far as I could tell, so I added a mention of it to the nbtouch docstring and summary in the docs.

@vinayak-mehta, first off, I want to say a big huge THANK YOU for this package! :tada: It'll be very helpful for me in interacting with my notebooks at the command line.

One thing I was very eager to try out was nbblack, because it has the ability to format every code cell in my notebook. The only problem that I encountered, though, was the code cell lines that have Jupyter magic inside of them. For example, if a code cell reads as:

%load_ext autoreload
%autoreload2

then nbblack will raise the following error:

black.InvalidInput: Cannot parse 10:0: %load_ext autoreload

I was wondering if it might be possible to skip code magic when extracting code from each cell? A simple search to skip lines that contain % signs at the beginning might be the way to do it, I think, but having not dug further into the code base, I would defer to your knowledge on this.

When execution_count does not exist, the commands nbcat, nbtail or nbhead throw an error. So far this has fixed it for me and now those commands work again.

Related issue: #19

Use the parse function from the ansimarkup package to render colored output instead of concatenating escape sequences from the colorama package.

En passant, also used the MarkdownLexer to highlight markdown cells.

@vinayak-mehta thanks for handling #9 so quickly! I just upgraded nbcommands, and I think I might have found one more edge case, I hope you don't mind me reporting this here.

There are some code cells where we want to show the source of the function, so we write:

func_name??

to show the full source, or

func_name?

to show just the signature + docstring (if I remember correctly).

This example, I think, also gave me an error, this time being:

black.InvalidInput: Cannot parse: 1:6: func_name??

I dug into the source, and saw that you basically added a regex replacement, is that right? Would the correct thing to do here be the following? I think the order has to be the double-?? before the single-?.

source = re.sub("^??", "#??", source, flags=re.M)
source = re.sub("^?", "#?", source, flags=re.M)

# followed by...

black_source = re.sub("^#??", "??", black_source, flags=re.M)
black_source = re.sub("^#?", "?", black_source, flags=re.M)

-B, --before-context=NUM  print NUM lines of leading context
-A, --after-context=NUM   print NUM lines of trailing context
-C, --context=NUM         print NUM lines of output context