Hash UDF for Standard Library

[aGorilla]

Based on DaleHohm's comments in this thread, and GaryFrost's dictionary object thread

Out of personal preference, I did a bit of renaming, and changed the indent to 2 spaces. If that's sacrilege, let me know.

I've added error handling (possibly too much), and the function headers needed for documentation.

I'm a bit confused by Gary's code referring to a COM error handler, and I'm not quite sure when a COM error would occur, and whether @error would be set or not, so some of the error handling is likely to be off.

Let me know if you see anything wrong, or have suggestions for improvement. Please check both, the headers, and the code.

Thanks for your help.

I'd post as an attachment, but it's not liking Mozilla. I broke down and ran IE.

Hash.au3Edit: updated code, Added attachment.

Edited May 17, 2008 by aGorilla

[sbgwb0769]

what?

[aGorilla]

While I appreciate the bump, that doesn't really add much to the conversation.

Do you actually have a question?

[JRSmile]

nice udf, adds many abillities i missed in the previous ones, Thumbs UP :-)

[aGorilla]

nice udf, adds many abillities i missed in the previous ones, Thumbs UP :-)

Thanks, but I can't take credit for that. The feature set was pretty much laid out by DaleHohm and GaryFrost, I just added a bit of error checking (which I'm hoping is correct), and function documentation.

I'm hoping to work on examples next week, and that should help me find out if the error checking is ok.

In the mean time, if you use it, I'd love to hear how it goes.

[DaleHohm]

I haven't had time to stress test it, but by observation it looks good. There are some issues with the error handling however.

First on _HashAdd you need to check IsObj after the ObjCreate instead of relying on @error.

In most of the other routines checking @error as you are won't help you because without a COM error handler instantiated, the COM errors will be fatal and you'll bomb out of the script before @error can be checked. This is a difficult issue... If you instantiate your own COM error handler you can conflict with IE.au3 and Word.au3 and possibly Excel.au3 error handlers. Best solution would probably be to try to anticipate and avoid as many errors as possible and then document a suggestion that the use instantiate their own COM error handler such that the errors are no longer fatal and your @error checks would work.

Speaking for me, I need no attribution in this code and I'm guessing that Gary doesn't either (although he can speak for himself). This is why I offered it up for someone to complete. I'd prefer you simply took ownership of it.

Dale

[JRSmile]

In the mean time, if you use it, I'd love to hear how it goes.

I'm currently developing a µTorrent UDF and for now it works like a charme.

It is utilized to make a "config_variable_name" <> "config_variable_value" relation table.

I will use it for nearly every funtion that will be in that UDF.

Best regards,

J.

[aGorilla]

Thanks for the feedback.

@DaleHohm:

I'm thinking you meant _HashNew(), not _HashAdd(), that said... should I check both, IsObj, and then @error? Is it possible that the object would be created, but some other error occured? I'm guessing no, but I'm not sure.

I noticed in Gary's example that he was doing the COM error handler in the example code, so I was thinking it would be best to leave that to the user. I'll point that out in the _HashNew() example.

I'll also try to dig into the MS docs for Scripting.Dictionary to see if I can figure out when I should use @error.

@JRSmile:

Thanks for testing it. It should be fairly stable, and I'll try to improve on the error handling this weekend. The biggest changes would likely be in the error return values, but I'll try to keep changes to a minimum.

[aGorilla]

Updated code in 1st post.

Added (optional) 'compare mode' to _HashNew()

Removed @error checks from all but _HashAdd() and _HashDelete() (following GaryFrost's lead).

Added links to MSDN docs, and some remarks from those pages.

Reduced use of _HashFind() to match the above mentioned docs (although I don't agree with their approach).

Took ownership.

Set up shell files for examples, but haven't started on them yet. May work on them tommorow.

Considering adding a display function (inspired by Array library), would come in handy for examples.

[DaleHohm]

Does it break the comment lines in the code box for you like it does for me?

; ====================================================================================================
===========================
Func _HashSet($oHash, $vKey, $vValue)

I have to spend a few minutes cleaning it up before I can test it. Perhaps you can correct the basenote post.

Dale

[aGorilla]

Does it break the comment lines in the code box for you like it does for me?

; ====================================================================================================

===========================
Func _HashSet($oHash, $vKey, $vValue)

I have to spend a few minutes cleaning it up before I can test it. Perhaps you can correct the basenote post.

Dale

Yep, was annoying me too. Added attachment.

[aGorilla]

Best solution would probably be to try to anticipate and avoid as many errors as possible and then document a suggestion that the use instantiate their own COM error handler such that the errors are no longer fatal and your @error checks would work.

Removed @error checks from all but _HashAdd() and _HashDelete() (following GaryFrost's lead).

It seems I may have jumped the gun on that one. Some day I'll learn to not 'skim' while reading replies.

[DaleHohm]

Another check you can to is to insure that the object actually IS a Dictionary object:

If Not (ObjName($oHash) = "IDictionary") Then Return SetError(2, 0, False)

Dale

[jaui]

Updated code in 1st post.

...

you have also to update _HashClear, because .RemoveAll didn't require an argument.

; #FUNCTION# ====================================================================================================================
  ; Name...........: _HashClear
  ; Description ...: Remove all key/value pairs.
  ; Syntax.........: _HashClear($oHash)
  ; Parameters ....: $oHash - Hash object, created with _HashNew
  ; Return values .: Success - Returns True
  ;                  Failure - Returns False, and sets @error:
  ;                  |1 - Invalid $oHash - Not an object
  ; Author ........: Bill Guindon (aGorilla)
  ; Modified.......:
  ; Remarks .......: http://msdn.microsoft.com/en-us/library/45731e2w(VS.85).aspx
  ; Related .......:
  ; Link ..........;
  ; Example .......;
  ; ===============================================================================================================================
  Func _HashClear($oHash)
    If Not IsObj($oHash) Then Return SetError(1, 0, False)
    $oHash.removeAll()
    Return True
  EndFunc

[nobbe]

there is also a hash table class in the .net

http://msdn.microsoft.com/en-us/library/sy....hashtable.aspx

; hash tables

Global $hash = ObjCreate("System.Collections.Hashtable")

; add

$hash.add ($key, $value)

;get value

$value = $hash ($key) ;

[big_daddy]

Very useful library!

I found a bug in _HashClear(), the .removeAll method should be called without parameters.

; #FUNCTION# ====================================================================================================================
; Name...........: _HashClear
; Description ...: Remove all key/value pairs.
; Syntax.........: _HashClear($oHash)
; Parameters ....: $oHash - Hash object, created with _HashNew.
; Return values .: Success - Returns True
;                  Failure - Returns False, and sets @error:
;                  |1 - Invalid $oHash - Not an object
; Author ........: Bill Guindon (aGorilla)
; Modified.......: Bob Anthony (big_daddy)
; Remarks .......: http://msdn.microsoft.com/en-us/library/45731e2w(VS.85).aspx
; Related .......:
; Link ..........;
; Example .......;
; ===============================================================================================================================
Func _HashClear($oHash)
  If Not IsObj($oHash) Then Return SetError(1, 0, False)
  $oHash.removeAll()
  Return True
EndFunc

[Skizmata]

Based on DaleHohm's comments in this thread, and GaryFrost's dictionary object thread

You guys have made my day. In my work I use perl daily and one of the few big points that I have allways wished AutoIt had was a perl style Hash. This is going to be used in everything I use from here on out I think. I cant thank you guys enough.

Thank you! aGorilla DaleHohm and GaryFrost

*Edit*

Is there a reason I should use _HashAdd over _HashSet? In this example it seems HashSet has no problem adding a hash element.

$hash = _HashNew()

_HashSet($hash, "Key", "Value")
MsgBox(0, "", _HashGet($hash, "Key"))

_HashSet($hash, "Key", "NewValue")
MsgBox(0, "", _HashGet($hash, "Key"))


_HashAdd($hash, "Key2", "Value2")
MsgBox(0, "", _HashGet($hash, "Key2"))

_HashAdd($hash, "Key2", "NewValue2")
MsgBox(0, "", _HashGet($hash, "Key2"))

Edited May 10, 2009 by Skizmata

[hunt]

Hello! How to set event for _HashAdd or _HashSet?

[Zedna]

@aGorilla

Very nice!

Hash.au3 in first post seems to be ready for standard UDF.

Why didn't you send it as proposition for standard UDF (on the Trac) already?

Trac: http://www.autoitscript.com/trac/autoit

[monoceres]

@Zedna, as you already know, there are already a hash/crypt UDF on the way into the standard UDF library. I don't know what advantages this have, mine is pretty much as complete as it gets.