Fixed debugger crash when trying to represent angles in tables in E2

[Alexandre425]

If an angle is put into a table and the E2 chip is debugged, the debugger crashes with the following error:

[wire] entities/gmod_wire_expression2/core/table.lua:93: bad argument #2 to '?' (number expected, got nil)
  1. unknown - [C]:-1
   2. temp - entities/gmod_wire_expression2/core/table.lua:84
    3. unknown - entities/gmod_wire_expression2/core/table.lua:117
     4. updateForPlayer - addons/wire/lua/weapons/gmod_tool/stools/wire_debugger.lua:299
      5. unknown - addons/wire/lua/weapons/gmod_tool/stools/wire_debugger.lua:329

Timer Failed! [Wire_DebuggerThink][@addons/wire/lua/weapons/gmod_tool/stools/wire_debugger.lua (line 333)]

This can be easily confirmed with the following E2:

@name test
@outputs T:table
if (first() | dupefinished())
{
    T = table()
    T["a", angle] = ang()
}

In lua/wire/server/debuggerlib.lua, the angle's values are accessed by .p .y .r. They're not assigned as such, as in lua/entities/gmod_wire_expression2/core/angle.lua the function ang places the pitch yaw and roll in the table's numerical indices (1, 2 and 3).

Angles and vectors are formatted in the same way, so some redundant angle checks were removed, and formatPort.VECTOR is used for both angles and vectors when formatting the values in tables and arrays

[Alexandre425]

Travis Cl raises the following error:
lua/wire/server/debuggerlib.lua:78:67: expected ')' near '!'
in the following line:
elseif(type(Element) == "PhysObj" and e2_tostring_bone(Element) != "(null)") then --Bone
Naturally != isn't valid in vanilla lua, but it is in Glua

[thegrb93]

Ya, I think it's part of the code convention to not use glua syntax

[Alexandre425]

Ya, I think it's part of the code convention to not use glua syntax

Roger that, the error was in old code I didn't touch however

[bigdogmat]

That's weird that it errors because the isangle check should've only went through for actual angle objects. Can you run isangle({1,2,3}) and see if it prints true or false.

Edit: Actually where in the stack trace does it point to the code you changed?

[Alexandre425]

That's weird that it errors because the isangle check should've only went through for actual angle objects. Can you run isangle({1,2,3}) and see if it prints true or false.

It prints false.

Edit: Actually where in the stack trace does it point to the code you changed?

table.lua 84 is this line:
ret = ret .. formatPort[longtype]( v, orientvertical )
This calls the function I changed.

[bigdogmat]

Ah I see it now, my branches line numbers didn't match your trace.

[thegrb93]

The debugger tool doesn't have any errors when I look at angles in tables.

[thegrb93]

Tested with

@outputs A:array

A = array(ang(24,24,24))

[thegrb93]

Nvm got it to happen with

@outputs A:table

A = table()
A["Test",angle] = ang(24,24,24)

[thegrb93]

Can you remove the other changes except for line 16?

[thegrb93]

^

[Alexandre425]

Can you remove the other changes except for line 16?

Changed as requested

[thegrb93]

Ok. I'll test it in a few hours.