Implement for fast DFA engine, value tracking and uninitialized memory preventing of reading.

[rikkimax]

This PR is to see if there are any failures in the CI beyond what I already know.

EDIT: I've had to change if statements location for assignments, so that it reflects the condition, not the if statement.

[dlang-bot]

Thanks for your pull request and interest in making D better, @rikkimax! We are looking forward to reviewing it, and you should be hearing from a maintainer soon.
Please verify that your PR follows this checklist:

• My PR is fully covered with tests (you can see the coverage diff by visiting the details link of the codecov check)
• My PR is as minimal as possible (smaller, focused PRs are easier to review than big ones)
• I have provided a detailed rationale explaining my changes
• New or modified functions have Ddoc comments (with Params: and Returns:)

Please see CONTRIBUTING.md for more information.

---

If you have addressed all reviews or aren't sure how to proceed, don't hesitate to ping us with a simple comment.

Bugzilla references

Your PR doesn't reference any Bugzilla issue.

If your PR contains non-trivial changes, please reference a Bugzilla issue or create a manual changelog.

Testing this PR locally

If you don't have a local development environment setup, you can use Digger to test this PR:

dub run digger -- build "master + dmd#22400"

[ibuclaw]

GDC disagrees. :-)

https://compiler-explorer.com/z/1qhW7T9dT

[rikkimax]

I don't do truthiness interprocedurally, it's intentional.

[ibuclaw]

I'm pretty sure it's nothing to do with detecting truthiness, and more to do with representing variables as PHI nodes for each diverging path taken to reach the current point in code.

[rikkimax]

I'm pretty sure it's nothing to do with detecting truthiness, and more to do with representing variables as PHI nodes for each diverging path taken to reach the current point in code.

https://compiler-explorer.com/z/PqaYojh85

It is modelling truthiness.

[rikkimax]

I'm looking into the following error:

std/utf.d(2277): Error: Argument is expected to be non-null but was null
        static assert(isSafe!({ S str; size_t i = 0; decode(str, i);      }));
                                                           ^
std/utf.d(1171):        For parameter `str` in argument 0
auto ref scope S str, ref size_t index) @trusted pure
                 ^
std/utf.d(2277):        while evaluating: `static assert(isSafe!(() pure @safe
{
char[] str = null;
ulong i = 0LU;
decode(str, i);
}
))`
        static assert(isSafe!({ S str; size_t i = 0; decode(str, i);      }));
        ^

A quick look over at decode, shows that it does indeed cause a bounds check error if that function were ever be called.

This is another example of where @__ctfe kind of flag would be the solution. Unlike the last time I cannot solve for this.

There are a few more examples of this, relating to decode(front/back), and they appear to all be doing the same thing.
Giving the string a non-empty value, would solve this.

decode is correctly inferring as non-null for its parameter now.

[rikkimax]

I don't know why core.exception.AssertError@std/int128.d(480): unittest failure is happening.

Given that it is a runtime error, not a CT one, I'm inclined that it isn't related.

EDIT: likely is something to do with me, other PR's are green for phobos.

Confirmed its me, toString error'd, and formattedWrite reverted to built in behavior.

testdfa\std\int128.d(445): Error: Expression reads from an uninitialized variable, it must be written to at least once before reading
        const minw = (buf.length - bufStart) + int(signChar != 0);
                                   ^
testdfa\std\int128.d(389):        For variable `bufStart`
        size_t bufStart = void;
               ^
``

[rikkimax]

boilerplate is a known error, as this is industry code I'm waiting on feep to deal with it.

That covers serialized.

Ocean is known good error.

Okay I'm calling buildkite doing what it is supposed to.

[rikkimax]

Lists of tests in testsuite where error is correct:

• runnable/xtest46_gc.d
• runnable/xtest46.d
• runnable/exe1.c
• runnable/exe2.c
• runnable/exe3.c
• runnable/nullderefcheck.d
• runnable/test12874.d
• runnable/testassert.d
• compilable/arraytopointer.c
• compilable/fix20425.c
• compilable/interpret3.d
• compilable/test22755.c
• compilable/test22842.c
• compilable/test22875.i
• compilable/test23044.c
• compilable/test23875.c
• compilable/testcomplex.c
• compilable/testcstuff1.c
• compilable/testcstuff2.c
• fail_compilation/failcstuff2.c

Only one known bug: runnable/bitfieldsposix64.c

I also want to fix the failures in phobos for decode somehow.

[rikkimax]

Functionally, it is now complete.

I need to write up a changelog and do an audit, but otherwise its done.

[rikkimax]

Okay, I'm now happy, but I'll leave it a few days to see if I get annoyed with anything else.

[thewilsonator]

Do ping, when this is good to go.

[rikkimax]

@thewilsonator

[thewilsonator]

Apart from one suspect change looks OK to me.