Just for fun, here is a very lightweight test runner I have used
sometimes. First, I define a decorator that prints the name of each test
function and its docstring when the test is called:
def self_printer(f):
def new_f(*args):
print()
print ('Test:', f.__name__)
if
On Monday, February 8, 2021 at 8:48:01 AM UTC-6 Edward K. Ream wrote:
> Aha: unit testing is easy.
Indeed yes!
leoserver.py now contains the TestLeoServer class:
- setUpClass starts the bridge and begins coverage testing if coverage can
be imported.
- tearDownClass ensures that the shut_down c
It works for me :)
Most likely thing I can think of: You installed markdown with a different
Python installation than the one that launches Leo.
For example, on some Linux systems (and on my own Windows one), *python*
runs Python 2.7 while *python3* runs Python 3.8.
On Monday, February 8, 202
On Monday, February 8, 2021 at 8:48:01 AM UTC-6 Edward K. Ream wrote:
> The communication between client and server is complete.
I ran the client and server in separate Windows consoles (processes) with
the following Leonine script:
g.cls()
import time
if c.changed:
c.save()
g.execute_shell
Sorry about this newbie question on scripts within LEO.
I was generating html for an outline using the *rst3* command, but then
decided I'd rather go through the nodes and generate my own html.
I want to allow markdown in the body of the nodes, so I installed the
*Python-Markdown* library.
Whe
This Engineering Notebook post reports spectacular progress:
- Félix and I have corrected each others misunderstandings.
- The communication between client and server is complete.
- Interaction between vs-code and Leo is much simpler than I imagined.
The "language" of that interaction consists p
