Some conclusions from experiments on various codes types.
Two main ‘_- observations
‘_- Stopping the function.
I have done a number of experiments, as briefly explained at the end of the last post.
Some important observations:
It seems that my “hook procedure” function will potentially be triggered indefinitely. Possibly this is reasonable. It is further reasonable that the simple API Function call already discussed is needed to stop the function being triggered once my function has done its main purpose:
Here an Ali -As-‘d version of this code line:
UnHookWindowsHookCodEx hHookTrapCrapNumber
The corresponding Declare line for that would be
Private Declare Function UnHookWindowsHookCodEx Lib "user32" Alias "UnhookWindowsHookEx" (ByVal hHookTrapCrapNumber As Long) As Long
‘_- Doing the Pop up window sizing and positioning ( API User32 dll program SetWindowPos)
Originally the issue for which my Hook procedure Function was required was the _ Part _b)(ii) the positional arguments _ .
The API Function which appears to have been used in Visual Basic and VBA codes in similar issues to this is the
SetWindowPos ( https://msdn.microsoft.com/en-us/lib...(v=vs.85).aspx ). This looks potentially very useful as it can change_..
_..all three positional co ordinates, horizontal (x), vertical (y), and the “z axis” “Z order”. ( The last z coordinate can be thought of to a first approximation as the axis looking onto the screen, the “top z” then being the window “seen” above all others ).
_.. also, as a bonus, it can also be used to change the seen Window width and height.
The syntax in a code line use of this function is pseudo like
_ SetWindowPosition _ WindowIdentifyinghandle, zorder , x , y , width , height , zFurtherInfo
The two z__ options are not too intuitive, and even some computer Professionals make an inspired guess for them based on a partial understanding of the documentation. The rest of the argument parameters are as “written on the can” _ x , y , width , height ( “says the tin”.. https://www.excelforum.com/excel-pro...ml#post4831198 )
From the experiments with use of this I was concerned about a seemingly unnecessary recursion process that seemed to set in. I also observed this in published Functions which were intended to do something similar.
My final Hook procedure Function differs therefore slightly in any I have seen so far in that it a globinal variable, GlobinalCntChopsLog, is used to keep track of the times that my Hook procedure Function is run. The constant jerking back and forth of the function, repeating, presumably by the “Bookmarks” reacting seemingly indefinitely is actually needed. As noted in the previous post we will do some selection to hopefully pick out the correct “event”, that is to say the appropriate event in the chain of events in which the “popping up” of our message box is included. But I am not sure if the extra recursion runs of copies of my function are necessary or desired. - The required action appears to be done on the first use of the SetWindowPosition . However a usage seems itself to trigger the function. I am assuming this is a recursion process so that another copy of my function is made and that is run. It does not, however, go into stack overflow. Consistently at the 29th recursion, it would appear that the recursion stops and each copy function is ended one after the other. ( A guess is that the stack might be limited somehow to 30 ). For my codes this resulted in a total of 6+29=35 runs: My function jerked itself off 5 times until a condition I set to hopefully find the event I was wanting to be caught was satisfied. So on the 6th run the SetWindowPosition code line was done. Experimenting on that showed that at this point all was done and OK as regards the positioning. A further 29 copy runs of the function appeared to be done, suggesting that the SetWindowPosition code line was responsible somehow for this. ( Example of this test code with output Log is shown here: http://www.excelfox.com/forum/showth...0478#post10478 )
Logic of my final Hook procedure Function
At the outset of the function, my gobinial variable is increased by 1. This intended to keep count of the copy number of the Function being run. It will therefore be reduced by 1 at the end of the function. Therefore, should a recursion run take place, ( that is to say a copy of the function is made and run ) , then the gobinial variable will hold the value of 2. A test for that will be made. This is tested for and If a value of 2 is found Then the “hook is released” and the function ended.
When my function is no in a recursion run, that is to say GlobinalCntChopsLog is not equal to 2 then the code proceeds as follows:
The value of the first argument lMsg at the incoming signature line _..
_ Function HoldYaBackCalledYaBackClapTrapRuc(ByVal lMsg As Long, ByVal wParam As Long, ByVal lParam As Long ).
_.. is investigated
This is done for the following reason:
As explained in the lost few posts, one result of the “hook” being “hanged” is that when one of our “bookmarks” is “triggered”, then some coding ( possibly the Microsoft CBTProc callback function ( https://msdn.microsoft.com/en-us/lib...(v=vs.85).aspx ) ) is responsible for passing information into the three arguments of my Hook procedure Function. ( Possibly this coding itself is responsible for all the other “placing of bookmarks” etc.. previously discussed ).
The first argument defines the type of “event” that has done the triggering. ( Remember we mentioned previously that it is an unfortunate co incidence that the number of 5 is what we are looking for here.. This was also the number which we needed to stipulate in the setting of the hook code line, hHookTrapCrapNumber = SetWindowsHooksExample(5, AddressOf , …….. ). That had a different meaning so should not be confused with the current 5 under discussion. )
The value of 5 specifies that the system is about to activate a window.
So If lMsg = 5 Then the code line_..
_ SetWindowPosition(wParam, 0, 10, 50, 400, 150, 40)
_.. is carried out.
_ wParam ________ is the passed windows identifying number for my Message box
10, 50, 400, 150 ___ are the horizontal and vertical size and positions
_ , 0 , _ , _ , _ , 40 __ The two numbers 0 and 40 are chosen after a bit of intuitive guessing based on the previous given Microsoft references. The end effect is to have the window seen as dominantly as wanted. They are likely to be based to some extent on experimenting in a particular requirement.
( As a function, the SetWindowPos is designed to return a value. In this usage I have not experienced problems using it as a Sub routine Call _..
_ Call SetWindowPosition( , , , , , , )
_.. but to be on the safe side I have used it as a Function returning its return in a Boolean variable, Booloks )
_.__________________
All the ground has been covered and explained as far as I am able to give finally a Pop Up User pseudo InputBox with range selection alternative with API User 32 dll Programs.
The next post will give a Summarised working full codes solution.
Reply With Quote
Bookmarks