Max Haughton

Fix Issue 23284 - better floating point error messages.

2

There was also a bug here in that real literals did not get a suffix attached.

newCTFE rebase-dump

2

This adds a preview flag but it shouldn't be merged just yet but you can play with it yourself. I'll add a markdown file that explains the design in more...

Fix Issue 15749 - Implement with expressions

19

I've definitely missed some things but this is enough to be reviewed (i.e. non-draft). I haven't decided what to do with the `union` usage yet. Hypothetically one could use the...

Make -vtemplates measure the time spent instantiating

6

This times the total time for a given instantiation, but the individual measurements can be done in a followup

Fix Issues 23130, 19528 : Inline asm was too liberal about operand si…

10

…zes. A 8-byte type was only considered 8 bytes in 64 bit code, for some reason. This led to inconsistent compilation of extended asm. I tried to add an error...

Refactor address-taking code and use it to give a better error message

1

Sketch towards structuring the escape code. Strange functions lying in ponds manipulating integers is no basis for a system of compilation. much to clean up but enough to be looked...

Merge and use time trace sources into dmd

16

For anyone who doesn't know this is the time-trace facility from LDC. It's effectively a one-stop-shop for profiling compile times, simultaneously providing a simple high level view (with an appropriate...

Fix Issue 22987 - Make traits getLocation have an option to return an…

12

… absolute path The feature is required for an application of metaprogramming within Symmetry. A flag is added rather than changing the default in case it breaks something.

ImportC - First go at implementing mixinC (In this iteration it has t…

13

…he syntax mixin[C], bikeshed away) - Currently just for expressions. We don't capture macros yet so this isn't particularly useful just yet (we do have a small macro engine for...

assert(0) as ternary RHS => crash

2

```d int front(int param) { return param ? 1 : assert(0); } ``` Also causes an assertion failure in dmd (https://issues.dlang.org/show_bug.cgi?id=22587). Is it even valid code?