Diff page still compares things that shouldn't be compared

[atomicptr]

Just seen this:

that shouldn't happen, have to figure out how to change that ho :>. Obviously the node didn't get renamed and all drops changed but instead it compared two completely different nodes with each other.

[TobiTenno]

It's probably because they're never really sorted in a particular order....

[atomicptr]

HM, I was more thinking in the lines of giving everything an _id, but sorting might actually be a good idea 🤔

[Senexis]

The referenced PR adds an "_id" field and fixes the issue, and it would fix the issue if merged, though instead of using the implementation I suggested in the PR might not be perfect as the conversation in the PR shows. 😄

A live preview of this: https://senexis.github.io/warframe-drop-data/diff.html, though due to the newly added "_id" field it all shows as new items.

[TobiTenno]

right, @atomicptr doing both.... give an id, sort by ID

[Senexis]

There's no need to sort the array as long as it has IDs. Sorting it doesn't add anything because all the transients already have a unique ID at that point and is good enough.

[TobiTenno]

right, but if the diff is comparing plain text and not id to id, then that would still be needed...

[Senexis]

It's not. It seemingly comparing to text (it doesn't; it's the order of the array) is only being done in the case when the ID is missing, so adding an ID to the transients is really good enough.

Edit: Turns out the entire thing is being sorted already anyway.

    function sortModName(a, b) {
        if (a.modName)   { return a.modName.localeCompare(b.modName) }
        else if (a._id)  { return a._id.localeCompare(b._id) }
    }

    function sortEnemyName(a, b) {
        if (a.enemyName) { return a.enemyName.localeCompare(b.enemyName) }
        else if (a._id)  { return a._id.localeCompare(b._id) }
    }

    function sortItemName(a, b) {
        if (a.itemName)  { return a.itemName.localeCompare(b.itemName) }
        else if (a._id)  { return a._id.localeCompare(b._id) }
    }

[atomicptr]

@Senexis your diff page actually shows the one thing that bothers me about this, instead of seeing hey shit changed for this thing you just see "THIS GOT REMOVED", "THIS GOT ADDED". :D

[Senexis]

Yep, which is why the PR exists, to attempt to fix it. :-)

The only reason this is happening is because the transients never had an "_id" field.

[atomicptr]

Fixed with #17