Fixed operator precedence

[TheGrandmother]

Operator precedence is now normal.
Now equality checks bind stronger than logical operators.

Before one had to write 1 == 2 and 5 <= 2 as (1 == 2) and ( 5 <= 2 )

Relates to issue #385

[supercooldave]

I'd like to see some more extensive testing, which will help define/understand what normal is. A single file test with lots of possibilities should do the trick.

Your tests could be simply showing that 1 == 2 and 5 <= 2 and (1 == 2) and ( 5 <= 2 ) are the same, for many such expressions.

[TheGrandmother]

@supercooldave
Oki I'll fix. Just didn't think this was a big enough thing to test. But I'll do it.

[TheGrandmother]

Added tests.
But wait with merging until #380 is merged.

The tests uses the new printing to avoid having to change them once the new printing is merged.

[albertnetymk]

#380 is merged, better to do a rebase to make CI happy.

[TheGrandmother]

Fixed!

[albertnetymk]

operator_precedence.enc breaks some coding convention, but it's very minor. I would merge it in one hour if no objections.

[supercooldave]

@albertnetymk This is not a very useful comment.

[albertnetymk]

@supercooldave Because I didn't elaborate on what coding conventions are broken or what?

[supercooldave]

@albertnetymk Precisely. Your comment was akin to "Something is wrong – but I'm not going to tell you what it is."

[TheGrandmother]

@albertnetymk
Is the coding standard defined somewhere?
I was made aware that we had one during the hackathon :p

[albertnetymk]

There's no strict coding standard, but we have a convention not to exceeding 80 columns and not having trailing spaces. They are very minor. I would merge it as it is if you don't feel like fixing it.

[TheGrandmother]

Fixed!

[EliasC]

This looks reasonable! Merging before I go home today.