ipairs bugged?

[zgrssd]

As far as I understand the only difference between pairs and ipairs should be that ipairs goes by alphabetical/numerical order. Both should still go over all the data.

But when given a table like this:
[1] = "CHAT_CATEGORY_SAY"
[2] = "CHAT_CATEGORY_YELL"
[3] = "CHAT_CATEGORY_WHISPER_INCOMING"
[4] = "CHAT_CATEGORY_WHISPER_OUTGOING"
[6] = "CHAT_CATEGORY_ZONE"
... (the table is in total 41 entries long)

ipairs will stop dead after entry 4.
As opposed to pairs, wich goes properly over the full table.

As far as I understand the syntax those are global functions.
Any chance Zenimax bugged ipairs up?
Any chance some addon is overwriting ipairs with a faulty version?

[Khaibit]

Originally Posted by zgrssd As far as I understand the only difference between pairs and ipairs should be that ipairs goes by alphabetical/numerical order. Both should still go over all the data. But when given a table like this: [1] = "CHAT_CATEGORY_SAY" [2] = "CHAT_CATEGORY_YELL" [3] = "CHAT_CATEGORY_WHISPER_INCOMING" [4] = "CHAT_CATEGORY_WHISPER_OUTGOING" [6] = "CHAT_CATEGORY_ZONE" ... (the table is in total 41 entries long) ipairs will stop dead after entry 4. As opposed to pairs, wich goes properly over the full table. As far as I understand the syntax those are global functions. Any chance Zenimax bugged ipairs up? Any chance some addon is overwriting ipairs with a faulty version?

From what I understand, ipairs only works with "well-formed" tables (i.e. sequential indices without any gaps, 1-indexed). Your table is missing index 5, which is why ipairs stops there.

[zgrssd]

Originally Posted by Khaibit From what I understand, ipairs only works with "well-formed" tables (i.e. sequential indices without any gaps, 1-indexed). Your table is missing index 5, which is why ipairs stops there.

Now that you mentioned it I looked up the definition. It appears like "ipairs" for really only is a short version of "for i=1, #table, 1 do":
http://www.luafaq.org/#T1.10

Looks like the source that originally pointed me to ipairs was not precise enough in representing what it cannot do.

[merlight]

Originally Posted by zgrssd Now that you mentioned it I looked up the definition. It appears like "ipairs" for really only is a short version of "for i=1, #table, 1 do": http://www.luafaq.org/#T1.10

With the minor difference that ipairs doesn't even need to get #table. The generator it returns simply stops at the first nil element.

[Xrystal]

I always translated ipairs to mean indexedpairs so only ever use it for tables that I know were indexed 1 to #tab. For the most part I use pairs as most of my tables are indexed by name rather than a running counter.

If I use table.insert and I don't delete elements on the fly I use ipairs. Otherwise I would use pairs.

[Deome]

Originally Posted by merlight With the minor difference that ipairs doesn't even need to get #table. The generator it returns simply stops at the first nil element.

Yep. My addon relies HEAVILY on tables, and tables within tables ad nauseum, and I think there's only a single ipairs() call in 800 or so lines of code. Could be wrong, maybe two. If your table contains, or may contain, or may contain sometime in the future, anything other than a pure sequential index that starts with 1, just use pairs() to cycle through the whole thing. Besides, you can do so much more with pairs() than ipairs().

Also, my new favorite ZO function is NonContiguousCount(table). I haven't seen it mentioned much, but it's basically #table for non-sequential or alphabetical tables, which is a gods-send for me.

[katkat42]

Originally Posted by Deome Also, my new favorite ZO function is NonContiguousCount(table). I haven't seen it mentioned much, but it's basically #table for non-sequential or alphabetical tables, which is a gods-send for me.

Really? And I've been using one I had rolled on my own!

[Xrystal]

Originally Posted by katkat42 Really? And I've been using one I had rolled on my own!

Woot, no more for loops rofl. Although that function probably does all that stuff itself anyway.

[Sephiroth018]

Originally Posted by Xrystal Woot, no more for loops rofl. Although that function probably does all that stuff itself anyway.

One less self made helper function for me too