std.complex: Add complex exp(), log(), log10(), and pow()

[ibuclaw]

Pushing out initial implementation. Unittest incoming...

[dlang-bot]

Thanks for your pull request, @ibuclaw!

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 + phobos#7456"

[ibuclaw]

Looks like tests are failing because std.complex doesn't have any template constraints for all it's functions. Apparently indirect constraints aren't enough (Complex!T foo(T)() { ... } doesn't inherit constraints from it's return type struct Complex(T) if (isFloatingPoint!T))

[ibuclaw]

Actually, I can't reproduce with what I thought would be a reduced test, so must be a bug in partial then (I get the same error when replacing pow with cos).

Euler is appeased.

[ibuclaw]

Still WIP are:

• log() special cases.
• Put both exp() and log() special cases in Ddoc table.
• Changelog entry?

Best for another PR...

• tan is missing, as are also:
• asin, acos, atan
• sinh, cosh, tanh
• norm (squared magnitude)
• proj (projection onto the Riemann sphere)

[ibuclaw]

@thewilsonator - green after changing a few more tests from == to approxEqual.

These are a little tedious to write out, would it make sense to add an approxEqual helper for Complex!T?

[PetarKirov]

would it make sense to add an approxEqual helper for Complex!T?

Or better yet, to isClose (#7384).

[ibuclaw]

Excellent, I better update all my tests then. :-)

[ibuclaw]

Excellent, I better update all my tests then. :-)

That can wait for a follow-up PR anyhow, as most tests are isClose, but some only pass with approxEqual.

[ibuclaw]

@thewilsonator - anything left here?