Permalink
Comparing changes
Choose two branches to see what’s changed or to start a new pull request.
If you need to, you can also .
Open a pull request
Create a new pull request by comparing changes across two branches. If you need to, you can also .
8
contributors
Commits on Nov 02, 2018
This is needed as we can't have a constant intarray ....
Commits on Nov 03, 2018
In my tests, repeating a benchmark many times gives roughly a 1% to 4% spread in the resulting values, so it makes some (limited) sense to keep three significant figures.
Commits on Nov 04, 2018
This fixes overflows when code to be benchmarked took very short amounts of time, and this should also avoid getting the number of operations to be infinite.
If an expandable token is passed as a pole, then the expanded version would be printed rather than the exact token list that was searched.
Commits on Nov 19, 2018
Commits on Nov 20, 2018
Commits on Nov 26, 2018
Commits on Nov 29, 2018
Commits on Dec 02, 2018
Also \int_rand:n, which is a trivial extension of int_rand:nn.
Commits on Dec 04, 2018
Enhanced the documentation for tl_range:nnn according to #495 (comment)
Commits on Dec 10, 2018
To follow \Ucharcat changes.
Commits on Dec 11, 2018
Commits on Dec 12, 2018
Commits on Dec 13, 2018
Commits on Dec 17, 2018
Updated documentation for \str_range:nnn to match \tl_range:nnn
Commits on Dec 20, 2018
I had thought integers are always normal fp numbers, but they can of course be zero too
Commits on Dec 23, 2018
The default value for log-declaration has been changed to false, but the documentation was not updated
Commits on Dec 24, 2018
Commits on Dec 29, 2018
Presumably, over time the ChangeLog should have everything of this type in it: we might consider retiring `l3obsolete` eventually.
Commits on Dec 30, 2018
Commits on Jan 01, 2019
... also copyright for l3kernel files.
Commits on Jan 06, 2019
Commits on Jan 07, 2019
Commits on Jan 11, 2019
This makes a real difference to load time with Unicode engines: they map over all of the Unicode data.
FMi's suggestion.
Commits on Jan 12, 2019
Commits on Jan 13, 2019
There was an unclosed \ifx here: the performance gain doesn't justify the required complexity.
The data file for case folding instructs to fold some code points to a sequence of 1,2,3 code points (given by hex numbers). In the case of 3 hex numbers the parsing code ended up with a space at the end of the third number
Commits on Jan 14, 2019
This way, things like uninstall 'just work'.
... until WSPR sorts it :)
Need to fix Lua ...
Commits on Jan 15, 2019
Commits on Jan 16, 2019
Somewhat less complex than \tl_set_from_file:Nnn
To be addressed properly later.
Rather than complicated business about x-expansion it is enough to perform the assignment before the end of file, then let the file end.
Commits on Jan 18, 2019
This reverts commit 12749b2.
Commits on Jan 20, 2019
This was never documented, so no deprecation required.
Commits on Jan 21, 2019
Tidy up the test files and documentation and make sure that behaviour and docs match.
Commits on Jan 22, 2019
We need to tighten this up depending on final approach to these functions. At present, they certainly are not global.
Commits on Jan 23, 2019
These are always set at the current group level.
The work on global box functions shows this up. These are experimental and unused, so a direct change.
Commits on Jan 24, 2019
Commits on Jan 25, 2019
There is an issue causing the CTeX tests to fail ...
Commits on Jan 26, 2019
Commits on Jan 28, 2019
There are no files selected for viewing