I develop language-extension with SignatureHelpProvider.
Some functions have several signatures, so I push several SignatureInfo to SignatureHelp.signatures array.
The problem is - how to get current active signature selected by user?
Current realization switches the activeSingature to 0 every time User types a character.
Also in vscode >=1.6 you must set an activeSingature to some value - otherwise the hint mini-window won't pop up.
Any ideas how to get this work properly?
It's up to you to decide which signature you want to set to active for your user.
It might for example be the one with fewer parameters that satisfies your activeParameter. It's your call as an extension developer.
When he types a character, you are probably invoking the signatureHelpProvider (with your trigger characters) and that's reseting activeSignature back to 0.
Related
recently I stumbled upon one of the videos of Benjamin Schumann titled: What are dynamic and action parameters and when should you use them in your AnyLogic model.
I tried to further adjust the functions of dynamic and action based parameters for a problem of mine. Just to give a heads up, I am fairly new to Anylogic (only worked through that one book, and some minor projects and tutorials) and been decent in Java (been a few years since I've been actively working in Java but currentlystarting to get back in [still rusty]).
Regarding my actual problem, in the video Mr. Schumann has an agent with three parameters. One static, one dynamic and one action. In addition to that he has a variable (double) all set in his agent. On his main is a button to increment the value of the variable with the help of the parameters and to trace the lines in the console (= giving out a string if a certain threshold of the variable is passed).
I created a similar setting, however I happen to run into a lot of variable errors during time to time while compiling.
Here some example code snippets:
dynamic parameter p_Station of the type String
v_myFahrt < 222 ? "Wiesbaden Hbf" :
v_myFahrt < 442 ? "Wiesbaden-Biebrich Bahnhof Wiesbaden Ost" :
v_myFahrt < 663 ? "Wiesbaden-Mainz-Kastel Bahnhof" :
"Hochheim (Main) Bahnhof"
therefore my variable is called v_myFahrt, a double with the initial value of 0
action parameter p_durchFahrt with the default action:
v_myFahrt = v_myFahrt + 220;
and my Button on the main:
myAgent.p_durchFahrt();
traceln(myAgent.p_Station());
So basically it is a somewhat similar code as in the reference. I tried to than add another instance of the agent with a different set of "code" for the dynamic parameter (different Strings and values) as well as a different "code" for the action parameter (e.g. + 208 instead of + 220). To then wanting to trace the lines in the console with the button again.
I tried to add
myAgent1.p_durchFahrt(); traceln(myAgent1.p_Station());
to it.
But before I coul even run it, I keep getting the error "v_myFahrt cannot be resolved to a variable" for myAgent1. Inspecting the error it keeps referring to myAgent1 with the newly added code for p_Station and I can't seem to find a way around it.
What am I doing majorly wrong here?
it looks like you have created v_myFahrt in main, right? (that would explain your symptoms).
If yes, you should create it in MyAgent instead.
I'd like to use EA to generate Requirement elements programatically. I need to use the same sequence numbering (REQ00000xy), as with the GUI when pressing "Auto" button in "Add Element ..." dialog in order to keep´consistent numbering for Requirement elements created either from GUI or from API.
Selecting the last used sequence number from already existing Requierement elements won't help, as it don't move the sequence number up and next Requirement created from GUI .
Is there a way to get (and properly use) the sequence number via EA API or EA SQL?
The table you're looking for is t_trxtypes. This contains something like (EA's output)
Description;NumericWeight;Notes;TRX;TRX_ID;Style;
Autocount;1,00;prefix=bla;suffix=x;active=1;active_a=0;counter=126;;Class;1; ;
You're interested in the column Notes which holds as CSV list like
prefix=bla;suffix=x;active=1;active_a=0;counter=126;
This is a test setting for a class which currently has the number 126. So the next created class would be named bla126x and the entry would change to
prefix=bla;suffix=x;active=1;active_a=0;counter=127;
Just keep the columne t_trxtypes.notes in synch with your creations.
Note EA does not (seem to) allow direct DB access. However, it has a proven back door:
Repository.Execute("UPDATE t_trxtypes SET Notes='prefix=bla;suffix=x;active=1;active_a=0;counter=127;' WHERE TRX_ID=<your id>")
will do the update (replace <your id> with the appropriate key). Though Execute is undocumented it works ever since and for sure Sparx will not limit it as nowadays everyone relies on it.
As a side note: This counter is not safe. There are lots of ways (the easiest is a simple rename) to break it. You'd need some script/add-in to have regular checks your numbering is still consistent. If you rely on requirement numbering you better use an external system like, I dare to say, DOORS.
Finally, RTFM....
For elements, where sequence is defined, if you use empty name in set =AddNew() function, EA generates the sequence upon .Update(). Not earlier. So if you plan to use the generated sequence and add some description, you need to create the element with empty name first, then Update() it and after that append your description to the content of the Name field.
As easy as this.
The files/list _folders method in api docs has the following for the description of the cursor parameter:
cursor - Pass the cursor into list_folder/continue to see what's changed in the folder since your previous query.
has_more - If true, then there are more entries available. Pass the cursor to list_folder/continue to retrieve the rest.
https://www.dropbox.com/developers/documentation/http/documentation#files-list_folder
Which is it then? If I make another request to list_folders_continue passing in the token, am I requesting another page of results or changes to the folder? If I am requesting another page of results, what is the limit that was applied? And can I control that?
The docs also have the list_folder method under sharing, which appears to have more precise definition of the cursor and the support for the limit parameter
https://www.dropbox.com/developers/documentation/http/documentation#sharing-list_folders
Will the method under files have that too?
Which is it then? If I make another request to list_folders_continue passing in the token, am I requesting another page of results or changes to the folder?
It's both. The /files/list_folder/continue endpoint is used both for returning additional pages of items, as well as getting further updates in the future.
Each page returned by /files/list_folder[/continue] is limited in size, but there is no guaranteed limit, nor can you control that limit. Your app should just always check has_more and call back to /files/list_folder/continue if it is true.
Also, the endpoints /files/list_folder and /sharing/list_folders are different endpoints with different functionality.
The /files/list_folder endpoint is documented as:
Returns the contents of a folder.
The /sharing/list_folders endpoint is documented as:
Return the list of all shared folders the current user has access to.
The two endpoints use different cursors, for use with /files/list_folder/continue and /sharing/list_folders/continue, respectively.
Which is it then? If I make another request to list_folders_continue
passing in the token, am I requesting another page of results or
changes to the folder? If I am requesting another page of results,
what is the limit that was applied? And can I control that?
My own trial and error experimentation has determined that the "It's both" answer posted by Greg is not correct. Specifically, the correct answer is "it's either". See below for more details.
When you call files/list_folder and it returns the entire list of contents (with a has_more value of false), then the cursor returned is suitable for monitoring subsequent changes to the path specified by passing that cursor to files/list_folder/continue.
If files/list_folder returns the first of multiple pages of contents (with a has_more value of true), then the cursor returned is suitable for getting the next page of contents by passing it to files/list_folder/continue. This cursor is not suitable for monitoring for changes (try it yourself, you can call continue as many times as you want with that cursor and you will always get the second page of contents).
It's not entirely clear from the docs, but when getting pages of contents from files/list_folder/continue, each call will return a new cursor, and that new cursor must be passed to a subsequent call to files/list_folder/continue in order to get the next page. You must chain the cursors together in this way until you get the last page (where has_more is false).
The cursor returned in that last call to files/list_folder/continue can then be used in subsequent calls to files/list_folder/continue to monitor for changes to the original path specified in the very first call to files/list_folder.
So the cursor is either suitable for monitoring for changes (when it is returned with a has_more value of false) OR it is suitable for getting a subsequent page of contents (when it is returned with a has_more value of true), regardless of whether the call that produced the cursor was files/list_folder or files/list_folder/continue, but it is never both.
I have a form that loads a single record. The user does what they need to do on the form...in this case, they enter a date, and a button becomes available to click to advance the record to the next step in the process.
I have a public function that is logging the activity to tblActivity, and sets the record's new Status and Location. This Function takes 3 variables, and was working fine until today.
'I'm calling the function with this line from the button's Click event
LogActivity 15, Screen.ActiveForm, Me.Recordset
Public Function LogActivity(ByVal lSID As Long, Optional fForm As Form, Optional ByRef fRS As Recordset)
With fRS
Do Until .EOF
Debug.Print .Fields(5)
.MoveNext
Loop
End With
...
End Function
This should be printing the form's Status value, but fRS is passed in with no values. The form's recordset has values prior to being passed as the form has data. Some how it is getting lost in the pass. This was working fine, I have multiple buttons across 5 different forms that all call this same Function. Suddenly today, none of them can pass the recordset. I can think of nothing that was changed that would effect this. Most of the changes recently involved locking down fields and the appearance of buttons at the right time...nothing related to the recordset.
Naturally, this DB is supposed to go live on Monday.
Found the problem.
I had a backup from yesterday that was working fine.
One by one, I went through the changes I logged from yesterday and found that by changing some fields to .enabled = False and .locked = True is what was doing it. Apparently that was enough to clear all the values when passing.
Left the fields enabled, just locked them and it passes all values correctly.
Even though this was a failure on my part, I'll leave this up in case some one else makes the same mistake I made.
**** Update ****
I also found out that if I did a
fRS.movelast
fRS.movefirst
before anything else, it found the data. Not sure why it started happening, but these two things seem to have fixed it completely.
I need to automatically generate a 12 character value for my Business Key. Without any user interaction.
8 character -> Today Date (yyyymmdd or ddmmyyyy).
+
4 character -> Sequential Number (0001,0002,0003).
The Sequential Number must reset on each new day.
Is it possible to do this in Microsoft Access 2010+ without any coding involved?
Since you are using Access 2010+ the best way to accomplish your goal would be to use a Before Change data macro like this
To create the Before Change macro, click the "Before Change" button on the "Table" tab of the ribbon when the table is open in Datasheet View:
For more details on Data Macros see
Create a data macro
Good question, thanks for the challenge!
After some search, it seems it's possible to do that.
You can prefix the AutoNuber value by processing like the explanation available here: http://www.databasedev.co.uk/add_prefix.html
You can try to specify in the format of the field a format(now(),"ddmmyyyy").
Check this page for more informations, another user seems to have the same problem and got a solution: http://bytes.com/topic/access/answers/695188-custom-made-autonumber-show-todays-date
Hope it's helping you!