Doubled modifiers from ship not showing in Affected By

[blitzmann]

https://forums.eveonline.com/default.aspx?g=posts&m=4920037#post4920037

[blitzmann]

Checked on SISI with my alt - shows the same stats between pyfa and SISI.

This could be a bug with multiple modifications of the same attribute not showing up (possible related to work done with #101?). Although, I thought I had specifically checked for this possible regression when working on it.

EDIT: checked with pyfa 1.1.22 which was released before changes in #101 were made, still shows at 1 RoF modification of -25%

[blitzmann]

So, this extends past the Claymore as I thought and seems to affect any ship which modifies any attribute twice. The harpy is another example where the optimal range of Small Hybrids is modified by the Assault Frigate and Caldari Frigate skills.

Additionally, if the two skills are at a different level, both modifiers show up. This leads me to believe it's an issue with adding the same modifier information to a set() or dict().

EDIT: it does seem to be a set() issue. from eos.modifiedAttributeDict:

affs.add((modifier, operation, bonus, used)) 

affs is a set of affectors. When dealing with ship bonuses that affect the same attribute with the same modifier, it tries to add it twice:

(<eos.saveddata.ship.Ship object at 0x0825F710>, '*', 0.75, True)
(<eos.saveddata.ship.Ship object at 0x0825F710>, '*', 0.75, True)

Being a set, it disregards the second one because it is not unique.

A fix could be simply switching to a list - I'd have to find all instances of such tho, and make sure there is no regression (EOS may be using set-specific methods for calculation, in which case this issue will probably not get fixed soon)

[blitzmann]

Actually, from what I remember and with playing around with it today, the affector code doesn't actually affect the calculations at all. It simply keeps up with what is affecting what, for the sole use in the Affected By tab. I have a fix pending after I test it a bit more thoroughly