| Commit message (Collapse) | Author | Age | Lines |
| ... | |
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Co-authored-by: MarkKoz <[email protected]>
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
While the previous code was safe, the synchronization was spread out
over different modules and was hard to wrap around. Additionally changes
could introduce context switches without the author being aware of them
causing potential race conditions with the refresh.
Moving the whole body into the with block solves both of these issues
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Tests for this were added additionally
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
After the move to a separate method, the docstring now documents the
behaviour so a comment is unnecessary
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
The method was also renamed from get_symbol_embed to create_symbol_embed
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The tests were not adjusted after the converter was corrected to accept
digits
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
We can still provide th user with at least the link to the docs,
for which we already have handling in the cog with a generic
"unable to parse message", using exceptions for that would mean setting
it here, immediately catching it and then providing the same or very
similar message.
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
No exceptions can be raised from the two lines of code because of the
data structures used, moving it out makes for flatter code.
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Co-authored-by: MarkKoz <[email protected]>
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Co-authored-by: MarkKoz <[email protected]>
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Co-authored-by: MarkKoz <[email protected]>
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The doc_cache coro was changed to be awaited directly instead of
creating a task to ensure the cache is populated before the item is
deleted
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The redis results are already being set in the BatchParser
for all symbols
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Co-authored-by: MarkKoz <[email protected]>
fixup! Docstring typos and style changes
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
As we extend the left side of a list now, using a deque that provides
a direct method for it is more fitting.
fixup! Use deque instead of a list
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This introduces a possibly redundant check for the doc_item being None
but results in flatter code with less duplication
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Using extend caused old items, including the ones that were requested
by users and pushed to the front, to be pushed back by all of the items
on the new page, possibly significantly delaying their parsing
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The last index will always be the largest one so there's no need
for max to search for it
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
As we check for non dt symbol names to be sure in case something pops up
, we can skip the initial group check as all the symbols from those
groups should point to non dt tags.
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
the placeholder contains a space and is actually 4 chars because of that
with a comma that adds up to 5 characters in the signature instead
of 4
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
The comments explain things that should be clear, or basic concepts
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Previously in some context inventory referred both to a single
"inventory" that we got from a remote objects.inv and to the internal
cog inventories. Always referring to the cog's inventories as plural
The update_single docstring was also changed from rebuild to build,
as the method doesn't handle anything with a preexisting inventory
with the same symbols being in the cog's inventories
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The main simplification was getting rid of keeping track of string depth
which was unnecessary, as we can just always skip them as was being
done for strings inside of brackets.
The branching was also simplified to make sure less unnecessary checks
were being done with a bit less confusing elifs.
|
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
run_in_executor can provide args to the func it's passed in,
making the use of partial unnecessary. This will also make it more
convenient to move to asyncio.to_thread when the codebase is switched
to python 3.9
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
The tests import the modules the other way around causing a circular
import
|