AutoHotkey Community

[infogulch]

[Edit]
This is version Beta3, and I've made lots of changes, some of which may not be reflected in these (already sparse) docs. This has been released early because of the proddings of bmcclure. In light of that, some code is still messy (i.e. I forbid anyone from looking at SA_Del() ) and not completely tested, so please gimme a couple days before you start code-bashing my lib.
[/Edit]

SimpleArray is a single variable list/array editing function library focused on speed and simplicity. It does not loop, parse or split the array when working with it, and all the functions can finish while executing less than 20 total lines of code.

I must credit olegbl for his excellent AHKArray library, which is similar in concept, but unfortunately not compatible. Since SA is 'simple' I won't be adding some advanced features such as those found in AHKA, so if you are looking for advanced features, I must refer you to olegbl's thread. However, I am open to suggestions and if you have a feature / improvement idea that fits in the bounds of the title, it would be very welcome.

Note: it is still to be tested further, so please be careful with your data! If you find any bugs I would very much appreciate it reported!

[About]

• Lib: These functions are library compatible, so you can save them in one of your lib folders as SA.ahk and they will be accessible without using #Include.
  
• Number Index: (param name Idx may only be a number)
  For values above 0, a position relative to the beginning of the array with an index of 1. So the first element is '1', the second '2' etc.
  For values below 0, a position relative to the end of the array with an index of -1. So the last element is '-1', second to last '-2', etc.
  Zero is not used, do not pass 0 as an index. (Note that this is different from AHKA)
  If you want to add a value at the end of the array with SA_Set() use '+' as the first character and the number following will be counted as an offset from the end of the array. This allows you to add it arbitrarily farther than the end. (Note that the default index param is '+1' so you can add it to the end without specifying a number yourself)
  
• Name Index: (param name nIdx May be either a number or a name.)
  A valid name is any non-blank, non-number string.
  If an element has a name, you can reference by it's name instead of element index.
  
• Functions:
  SA_Set( Array, Val, [nIdx="+1", Insert=false, giveName=""] ) ~ returns the new array with Val set to nIdx. If you pass a name for nIdx, the name will be preserved unless you pass something else to giveName. If you pass a number for nIdx, the element's name (if any) will be discarded. (this behaviour may be modified)
  SA_Get( Array, nIdx1=1, nIdx2="", nIdx3="", nIdx4="", nIdx5="" ) ~ returns the array element at nIdx1. Also retrieves nested arrays' elements. If you want to retrieve values more than 5 elements deep, add more params numbered sequentially.
  SA_Del( Array, nIdxFm, nIdxTo="", Count=0 ) ~ returns the new array with elements from nIdxFm to nIdxTo removed. OR you can pass nIdxFm and Count, in which case it deletes Count number of elements starting at nIdxFm
  SA_Trim( Array, nIdxFm, nIdxTo=-1 ) ~ returns the new array with everything outside of nIdxFm to nIdxTo removed.
  SA_Len( Array ) ~ returns the length of the array
  SA_IdxName( Array, Name ) ~ returns the index of Name in Array as a number
  SA_NameIdx( Array, Idx ) ~ returns the name (if any) of Idx as a string
  SA_ValidName( x ) ~ used internally, returns true if the value passed is acceptable as a name
  [i]SA_Enc( Array )[/i] ~ used internally, returns a value ready to be used as an element
  [i]SA_Dec( Array )[/i] ~ used internally, returns a value ready to be used outside the array
  SA_EncRE( x ) ~ used internally, escapes a string for use in RegEx

[Code] Download

[Laszlo]

Nice! You could get some more ideas in the 3.5 years old List manipulation functions thread. Lists are one dimensional arrays, represented as strings, with their elements separated by specific delimiters. Since the original library was written AHK evolved quite a bit. E.g. Regular Expressions could greatly reduce the code size and improve the speed.

[olegbl]

Cool! I like it! Good job on making it fast!
(I know that takes a ton of work =P)
Are you using a delimetered (|) string to store it (with encoding)?
If you want, I can work on some conversion function between our libs =)

[infogulch]

Thank you Lazlo and olegbl! I appreciate your support.

Ah, the List Manipulation Functions, I will definitely have to take a close look at those, thanks Lazlo. Actually, I was trying to avoid RegEx for speed purposes, since the most demanding commands in a function are StringReplace*2, StringGetPos*2 and SubStr()*2. But if you can convince me that RegEx would help speed, I'd be more than willing to attempt it.

[olegbl]

Yup, I was thinking of waiting for a settled format, I remember it took me months before I settled on something XD (Though I'm also going to college in about a month heh)
For your encoding/decoding format, do the %e, %d characters get recoded when included?

[defrostedchemist]

[infogulch]

[bmcclure]

Great functions! I'm going to give them a try for some of my simpler arrays in SteamLab and see how it works.

I would like to see how this compares to AHKArray (which I've been using in SteamLab) for larger, multi-dimensional arrays. Do we know how much faster the encoding is here than AHKArray's hexing?

Has this been tested on large/deep arrays to still have a good speed advantage?
_________________
Ben

My Trac projects
My Wiki
[Broken] - My music

[olegbl]

[bmcclure]

Thanks; That's what I figured.

I really love the advanced functions of AHKA, however, and have never had a big problem with my apps running slow.

A lot of extra speed can never hurt though.

The only thing that's really missing from SimpleArray that I would love to see is a simple SA_Find function that locates the index of the first instance of a value in an array.

Perhaps something like SA_Find(Array, Value, Pos) where Pos could be the positive or negative index number to start searching forwards or backwards from?

I know that it would probably be the slowest function included with SA, but in my case (and I'm assuming many others), I am always trying to figure out the index of a particular value in a dynamic array and this would be a great help.
_________________
Ben

My Trac projects
My Wiki
[Broken] - My music

[infogulch]

Hey guys!

That's a good idea, bmcclure. So I'm implementing element names, and ways to get their indexes. I'll probably overload most of the functions, allowing you to set and get nested arrays with one function, referencing either by name or index. It will still be fast, and I can keep my "no looping" staus.

Thanks for answering olegbl. Yeah, that's about as well as I could explain it.

Actually, as I'm going through and rewriting I'm actually shortening these functions quite a bit. E.g. sa_set, it was ~16 lines, and I shortened it to 11. And get went from 10 to 6. Hopefully shorter = faster in this instance. So maybe the name addition won't add much more overhead .

I wanted to post something here just to let you guys know I'm not abandoning this.


_________________
A great Beginner's Tutorial

[bmcclure]

Sounds cool! By names do you mean it will support named indexes?

If so, and with the lookup functions you're working on, I'm very excited! It could be exactly what I'm looking for for at least some of the SteamLab arrays, and if there is still a performance benefit over AHKA that would be an added bonus of maybe speeding up the application a bit.

Looking forward to the next code update
_________________
Ben

My Trac projects
My Wiki
[Broken] - My music

[CannedCheese]

This looks cool and I will definitely be playing with it. My only suggestion would be to add a second parameter so that one can optionally pass a delimiter (obv. I will add this myself, and I know you want to keep things simple and fast, but I'm not sure how much a delimiter change will slow things down). I often have to import user entered data from a text file, and some of the entered data can contain the pipe "|" character. I'd like the option to switch to a character that they can't enter, such as A_Tab or the standard comma, or something. Again, this is something that I should be able to implement for myself, but as I'm not sure it would really slow down the code any, I thought I'd suggest it.

[bmcclure]

Your data can contain the pipe character; that's the best part about it. The data you pass into the functions is encoded, so pipes will be encoded and decoded correctly at each level.

That is my understanding, at least; correct me if I'm wrong.
_________________
Ben

My Trac projects
My Wiki
[Broken] - My music

[olegbl]

Actually, from what infogulch said, I understand it as the delimiter being encoded rather than the pipe character. Which is much much more efficient.