ALV graph dependency on Data Types.

Hello Friends,
I have to display some fields in the graph option of ALV.For that I need to
declare its data type as Type INT4 or INT2.But in that case I cannot declare it with Decimal places.I can declare it as Type F but in that case I not getting desired display(like 6.130000000000E).
I can use Type P but I am not getting this option when I am doing F4 in data dictionary.So what should I use ,please help me if you have some solution?
Regards
Ashish.

Hi Sam,
You might be searching for the below site.
Report with a Graph.. An Approach!
Hope this helps you.
Cheers!1
VEnk@

Similar Messages

Linking x-axis of two graphs (with seperate data types)

Hi all,
I am plotting a sine wave and a resulting waveform generated after driving a component with that sine wave.
I have 2 data types:
i) One is a waveform - of the sine wave
ii) The other is a cluster - resulting from two arrays (DBL)
Ideally I would like to display both of these on a single plot. But I can't seem to do that. If you know how to do this, then please let me know.
So, the next solution was to display them on a waveform chart and XY-plot, respectively.
Is there some way of linking the x axis, such that when I zoom into one plot, I simultaneously zoom into the other?
I hope my question is clear.
Thanks

If you are talking about the amplitude then it is the Y axis scale. Have a look here on how to add a second scale (http://forums.ni.com/ni/board/message?board.id=170&message.id=314415&query.id=105962#M314415).
Then you can have a separate scale for each plot. All plots on one graph with separate scales... you've got your cake and you can eat it too!
Message Edited by Troy K on 07-25-2008 03:40 PM
Troy
CLDEach snowflake in an avalanche pleads not guilty. - Stanislaw J. Lec
I haven't failed, I've found 10,000 ways that don't work - Thomas Edison
Beware of the man who won't be bothered with details. - William Feather
The greatest of faults is to be conscious of none. - Thomas Carlyle

What is the data type for the Data Dashboard XY Graph?

Hey all,
I am playing with the new Data Dashboard (!!) on my iPad 2 and am having trouble setting up the shared variables that host data to the XY Graph.
On my PC, in my Project (inside of a variable library) I have set up shared variables for all of the other controls/indicators available in the app; They all work great. However, I have not been able to identify the data type that is for the XY Graph in Data Dashboard. Of course, inside LabVIEW, I use a cluster of 2 arrays; this is not an available data type that I can select while creating a new variable on the hosting side. I have even gone so far as to create a custom control from my XYGraph in LabVIEW and then creating the data type "From Custom Control..." with no success.
What gives? Am I missing something obvious? Thanks for the help!!
Tom
Solved!
Go to Solution.

Great question, Aerogoob.
The XY graph can be bound to a 1d array of "points", where each point is a cluster of two numerics (X and Y). To create a shared variable of this type, you can set the data type to "From Custom Control..." in the shared variable properties dialog. Of course, first you'll have to build the custom control of the correct type: array of cluster of two numerics.
If any of that doesn't make sense, please post back and we can walk you through it in more detail.
Also, just for completeness, the chart indicator can be bound to a scalar numeric or to an array of numerics. The graph indicator can only be bound to an array of numerics.

The data type DEC while downloading from ALV to EXCEL

The data type DEC while downloading from ALV to EXCEL is giving some invalid data .
Ex : In ALV field is the time difference in hours:minutes
if the value is :23 in Excel it will be 22:59 .
Please Suggest immediately

Hi ....
I have the similar problem. My coumns are not in any sequence when I download it to excel sheet.
How do I fix it.
Neha

How do I Change the plot waveform colour on a Dynamic Data type Graph?

I have multiple different plots on different graphs, and I want to set the colours of the different waveforms.
In order to display the waveforms on different graphs, I switched to Dynamic data type and then split the wire
according to how many channels I have.
I can set the graph background colour with a property node, and create cursors of different colours no problem
using the array of references that I have created, so there is nothing wrong with my array of references or the property node.
As soon as I try and wire a colour box into the Plot>Plot Colour control, I get error 1055 out of the property node.
I was wondering if it was an error between the keyboard and the chair, but now I am thinking it is someting more sinister.
The Error box pops up and can't be dismissed remaining the focus until I run TASKKILL on LabVIEW and terminate it.
Any ideas on how to achieve my goal would be appreciated.
(LabVIEW 8.6)
Solved!
Go to Solution.

Hi Sheela,
It's 1 plot per graph, up to 4 graphs displayed on different tabs.
Here are some screenshots
Attachments:
graph_display.JPG ‏135 KB
set_Plot_colour.JPG ‏79 KB

XControl Data wiring problem - Data type is Waveform Graph (Strict)

Hi,
I created an Xcontrol for a waveform graph. The datatype (under Data 1.ctl) is a WaveformGraph reference type.
When I wire a Waveform graph reference to the Xcontrol (as an indicator), there is a wire conflict.
I checked the strict definition and the reference types are theoretically congruent but I cannot seem to wire it together. Any solutions?
Please see the attached code.
Thanks!
Attachments:
GraphDisplay XControl.7z ‏1717 KB
Tester.vi ‏15 KB

It looks like the XControl data is a waveform reference (strict). I can't open the XControl though, because I don't have a 7z compiler, but the problem is probably that the refnum will only take in a reference to an exact waveform graph because it's strict.
You can create a non-strict reference and use that for your data by:
Open your data control FP
Press Ctrl-Space
Type "Control Refnum"
Press Enter and place that on front panel
Right click it and brows to the correct data type:
James
LabVIEW Professional 2014

The data type of the variant is not compatible with the data type wired to the type input.

Hello all..
Iam curently struggling with the following:
On my main(top)_VI, I have a mix-signal graph which will output several or one signals depending on the user. I also have an inner subvi which has the analisys i want to give the the data before outputing it to the main(top) VI, which has the mix-signal graph control.
Now, I created a reference from the mix-signal graph and Iam imputing this reference into the subVI (which inside has a property node type_control_mix-signal graph in which Iam connecting the refnum from the top VI to view the output data). The data consists of an array of clusters of 2 arrays of doubles each.
-- let say I create the property node type control mix-signal graph and i choose value property; it lets me put the array of clusters of 2 arrays of doubles each without problems; Now the second I connect the refnum pointing to the mix-signal graph on the Main_VI it brakes the cables and only accepts one cluster of two arrays os doubles instead.
--if on the other hand I connect the refnum to the property node type_control_mix-signal graph before connecting the imput it doesnt complain until I run the VI where it gives the error 91.
I have also notices a (strict) parenthisis at the end of the property node sometimes. [what does this (strict) means?]
FInally if instead of creating the refnum from the Main_VI, I create either a xy graph or a mix-signal praph it lets me connect the an array of clusters of 2 arrays of doubles each without any problem. The refnum connection seems to be the problematic factor, but by the same tolken I dont know any other means of sending that data from the inner to the outter VI.
I have the feeling Iam missing on one or more Labview fundamentals regarding refnums.
Any help will be apreciated

You almost found it on your own. The refnum needs to be 'strict', that means that the data type of the graph is set (2D dbl), otherwise it will use the default data type. To do this you create the refnum directly from the graph after you have wired the correct data type to the terminal. In this case the refnum wire will break if you change the data type of the graph.
Independend of this issue, I suggest you not to use the value property on the graph but directly wire the data (e.g. the 2D double) to the SubVI. The property value is by far the slowest way to get data, and 2D arrays are really some amount of data.
Felix
www.aescusoft.de
My latest community nugget on producer/consumer design
My current blog: A journey through uml

AnyGantt: How to set different bar color depending on data value.

Need to have different colors for bars in a gantt resource chart.
The color depends on certain data value not included in the visible part of the chart, so it must be provided 'extra' for the chart in some way without beeing visible.
Have studied some examples but cannot figure out the mechanism for this, and the procedure to implement it.
Any ideas?
Apex 4.2.2
Regards
Per.

Hi Per,
You mention "+Among the samples I can see that it has been implemented, but not how.+" - in the Information region below the chart that Tony referred to (https://apex.oracle.com/pls/apex/f?p=36648:50) that I've stated the following:
"The elements of the Timeline region have been customized via the Chart Attributes page, using the following settings: *Custom Colors*: #000000,#00FF00,#0033FF"
So that's how I applied custom colours to that particular chart. However, that declarative option won't meet your requirements, where the colours you wish to apply are dependent on data not included in the chart series query. You mention "+Now, if I add another column to the statement I get a yellow error message telling that the statement is not valid.+". It is not possible to simply add an additional column to the chart series query. Each supported chart type expects the associated chart series query to use specific syntax - see About Creating SQL Queries for Charts in Chapter 10 of the APEX User's Guide, http://docs.oracle.com/cd/E37097_01/doc/doc.42/e35125/bldapp_chrt.htm#BCEIJJID.
So to answer your question "+How to add the new value if not providing another column to the graph?+", if you wish to customise your chart to that extent, then you will need to use custom XML. Depending on what exactly you're trying to achieve, this may require the generation of custom data for your chart. If you edit your chart, and take a look at the Custom XML region, you'll notice a substitution string, #DATA#, towards the end of the XML. When a chart is rendered, we use the chart series query to generate the necessary XML to represent the data on the chart, replacing the string #DATA# with the actual XML. The format of the data XML generated for gantt charts uses a bar style called "defaultStyle", and this handles the default appearance of the bars in the timeline region e.g.
<period resource_id="1" start="2009.03.21 00.00.00" end="2009.09.21 00.00.00" style="defaultStyle"/>You'll see that the 'style="defaultStyle" ' attribute of the data corresponds to a segment of XML outlining what that style is e.g.
<period_style name="defaultStyle">Now take a look at the Bar Style in Resource Project example in the AnyChart online documentation http://www.anychart.com/products/anygantt/docs/users-guide/bar-style.html#bar-style-application-resource-project. In that particular example, a number of different bar styles are used i.e.
<periods>
<period resource_id="server_1" start="2008.07.07" end="2008.07.8" style="Working" />
<period resource_id="server_1" start="2008.07.8" end="2008.07.12" style="Maintance" />If you need to add a different bar style to particular rows of data on your chart, then you'll need to add new bar style XML for each colour you wish to use in the Custom XML on the Chart Attributes page, and then you'll also need to generate the data XML yourself, applying the necessary bar style to the relevant row of data. I've got an example of how custom data can be generated for a Resource Gantt here e.g. https://apex.oracle.com/pls/apex/f?p=36648:73 but keep in mind that particular example doesn't demo adding different styles to the data...but it can certainly give you an idea of what's entailed in the generation of custom data for your chart.
I hope this helps.
Regards,
Hilary

Any idea what this errorr means? the data type of the reference does not match the data type of the variable

I am using Veristand 2014, Scan Engine and EtherCat Custom Device. I have not had this error before, but I was trying to deploy my System Definition File (run) to the Target (cRio 9024 with 6 modules) and it failed. It wouldn't even try to communicate with the target. I get the 'connection refused' error.
I created a new Veristand project
I added the Scan Engine and EtherCat custom device.
I changed the IP address and auto-detected my modules
i noticed tat Veristand didn't find one of my modules that was there earlier. (this week)
So, i went to NiMax to make sure software was installed and even reinstalled Scan Engine and Veristand just to make sure.
Now, it finds the module, but when i go to deploy it getsto the last step of deploying the code to the target, and then it fails.
Any thoughts?
Start Date: 4/10/2015 11:48 AM
• Loading System Definition file: C:\Users\Public\Documents\National Instruments\NI VeriStand 2014\Projects\testChassis\testChassis.nivssdf
• Initializing TCP subsystem...
• Starting TCP Loops...
• Connection established with target Controller.
• Preparing to synchronize with targets...
• Querying the active System Definition file from the targets...
• Stopping TCP loops.
Waiting for TCP loops to shut down...
• TCP loops shut down successfully.
• Unloading System Definition file...
• Connection with target Controller has been lost.
• Start Date: 4/10/2015 11:48 AM
• Loading System Definition file: C:\Users\Public\Documents\National Instruments\NI VeriStand 2014\Projects\testChassis\testChassis.nivssdf
• Preparing to deploy the System Definition to the targets...
• Compiling the System Definition file...
• Initializing TCP subsystem...
• Starting TCP Loops...
• Connection established with target Controller.
• Sending reset command to all targets...
• Preparing to deploy files to the targets...
• Starting download for target Controller...
• Opening FTP session to IP 10.12.0.48...
• Processing Action on Deploy VIs...
• Setting target scan rate to 10000 (uSec)... Done.
• Gathering target dependency files...
• Downloading testChassis.nivssdf [92 kB] (file 1 of 4)
• Downloading testChassis_Controller.nivsdat [204 kB] (file 2 of 4)
• Downloading CalibrationData.nivscal [0 kB] (file 3 of 4)
• Downloading testChassis_Controller.nivsparam [0 kB] (file 4 of 4)
• Closing FTP session...
• Files successfully deployed to the targets.
• Starting deployment group 1...
The VeriStand Gateway encountered an error while deploying the System Definition file.
Details:
Error -66212 occurred at Project Window.lvlibroject Window.vi >> Project Window.lvlib:Command Loop.vi >> NI_VS Workspace ExecutionAPI.lvlib:NI VeriStand - Connect to System.vi
Possible reason(s):
LabVIEW: The data type of the reference does not match the data type of the variable.
=========================
NI VeriStand: NI VeriStand Engine.lvlib:VeriStand Engine Wrapper (RT).vi >> NI VeriStand Engine.lvlib:VeriStand Engine.vi >> NI VeriStand Engine.lvlib:VeriStand Engine State Machine.vi >> NI VeriStand Engine.lvlib:Initialize Inline Custom Devices.vi >> Custom Devices Storage.lvlib:Initialize Device (HW Interface).vi
• Sending reset command to all targets...
• Stopping TCP loops.
Waiting for TCP loops to shut down...
• TCP loops shut down successfully.
• Unloading System Definition file...
• Connection with target Controller has been lost.

Can you deploy if you only have the two 9401 modules in the chassis (no other modules) and in the sysdef? I meant to ask if you could attach your system definition file to the forum post so we can see it as well (sorry for the confusion).
Are you using any of the specialty configurations for the 9401 modules? (ex: counter, PWM, quadrature, etc)
You will probably want to post this on the support page for the Scan Engine/EtherCAT Custom Device: https://decibel.ni.com/content/thread/8671
Custom devices aren't officially supported by NI, so technical questions and issues are handled on the above page.
Kevin W.
Applications Engineer
National Instruments

Join columns of different data types

Can i join two columns of different data types in a single statement. Is there any cast or convert statement
SELECT B~PARTNER AS RESP_GROUP
FROM BUT000 AS B
INNER JOIN CRMV_LINKPARTNER AS C ON BPARTNER_GUID = CPARTNER_NO.
Here PARTNER_GUID is Raw 16 and PARTNER_NO is Char 32, the query is not returning any rows.
Thanks
Akila.R

Hi!
Basically we can say, you can't join them, but you might try this way:
SELECT B~PARTNER AS RESP_GROUP
FROM BUT000 AS B
INNER JOIN CRMV_LINKPARTNER AS C ON BPARTNER_GUID = CPARTNER_NO(16). "add
(16)
Or an other version
C~PARTNER_NO+16(16) "depends on, where is the real number stored
Regards
Tamá

Performance and data types: which to use?

Hi All,
I am wondering what data type to use and the effect of them on memory/speed.
1. What is the difference (if any) of using sgl, dbl, int etc. Looking at the LabVIEW help there seems to be a range of 8-256 bits of storage according to the data type. Is it basically choose the one with the smallest storage that can fit the data?
2. I currently have a cluster flowing through subVI's. The cluster contains the start time (or freq), the delta t (or f) and the array of data (about 500-5000 elements). I tried to use the waveform datatype but it couldn't handle a delta t of 2 nanoseconds (500 MHz signal). Am i ok using the cluster, or should i seperate the components and pass them along? What data type should i use for each of the components?
Thanks

There are three main issue to consider.
Range and accuracy. If you need a very high level of accuracy, then you will need to use the extended data type or even create your own, although that's unlikely.
Memory. Yes, SGL takes less than DBL, but unless you're dealing with really huge amounts of data this won't matter.
Coercion. Most built in functions work on DBL. If you wire a SGL into them, they will coerce it, possibly creating a copy of the data and increasing your memory usage.
To sum it up, most of the times it would be best to use the default DBL. It's highly unlikely you'll need one of the others.
As for your second question, it sounds to me like the data is a single organism, so I would say you should leave it in the cluster, but that really depends on whether the functions need it or not and whether you're constantly bundling and unbundling the cluster. Note that 5000 elements is far from being a large array and you shouldn't have any problems handling it.
As for the timing unit, if you really only have 5000 elements (that's 10 microseconds of data?) then you should not have a problem with using a U32 with a nanosecond as the base unit. That should give you the ability to measure more than 4 seconds.
Try to take over the world!

Subtotal in ALV grid for a particular type and Grand total in ALV

Hi,
I need to have sub total for a particular type(eg: goods, services).. and grand total at end in ALV grid..
ALV output required as below:
Type VAT registration number Country Total Gross Amounts Total Tax Amounts Total Amount, ex-tax
Goods ATU12345678 AT 222.42 0 222.42
Goods NL123456789B02 NL 3,417.00 0 3,417.00
 Goods Total 3,639.42 - 3,639.42
Services ATU12345678 AT 2,342.34 0 2,342.34
Services NL123456789B02 NL 223.33 0 223.33
 Services Total 2,565.67 - 2,565.67
 Grand Total 6,205.09 - 6,205.09
Let me as to how to achieve the above type in ALV grid...
Regards
Shiva

check this link..
Grand Totals in ALV grid disply function module
or do like this..
REPORT ZALVTESTFORSUBTOTAL.
tables:pa0008.
type-pools:slis.
types:begin of ty_pa0008,
 pernr like pa0008-pernr,
 begda like pa0008-begda,
 endda like pa0008-endda,
 ansal like pa0008-ansal,
 lga01 like pa0008-lga01,
 bet01 like pa0008-bet01,
 end of ty_pa0008.
data:it_pa0008 type standard table of ty_pa0008 with header line.
data:it_fieldcat type SLIS_T_FIELDCAT_ALV,
 wa_fieldcat type slis_fieldcat_alv,
 it_layout type slis_layout_alv,
 WA_events TYPE slis_alv_event,
 it_events TYPE slis_t_event.
select-options:s_pernr for pa0008-pernr.
start-of-selection.
perform getD_data.
perform disp_alv.
*& Form getD_data
 text
--> p1 text
<-- p2 text
form getD_data .
select pernr
 begda
 endda
 ansal
 lga01
 bet01
 from pa0008
 into table it_pa0008
 where pernr in s_pernr.
sort it_pa0008 by pernr begda descending.
endform. " getD_data
*& Form disp_alv
 text
--> p1 text
<-- p2 text
form disp_alv .
wa_fieldcat-fieldname = 'PERNR'.
wa_FIELDCAT-REPTEXT_DDIC = 'Personnel no'.
*WA_FIELDCAT-no_subtotals = 'X'.
append wa_fieldcat to it_fieldcat.
clear wa_fieldcat.
wa_fieldcat-fieldname = 'BEGDA'.
wa_FIELDCAT-REPTEXT_DDIC = 'Start date'.
append wa_fieldcat to it_fieldcat.
clear wa_fieldcat.
wa_fieldcat-fieldname = 'ENDDA'.
wa_FIELDCAT-REPTEXT_DDIC = 'End date'.
append wa_fieldcat to it_fieldcat.
clear wa_fieldcat.
wa_fieldcat-fieldname = 'ANSAL'.
wa_FIELDCAT-REPTEXT_DDIC = 'Annula salary'.
wa_fieldcat-do_sum = 'X'.
append wa_fieldcat to it_fieldcat.
clear wa_fieldcat.
wa_fieldcat-fieldname = 'LGA01'.
wa_FIELDCAT-REPTEXT_DDIC = 'Wage Type'.
append wa_fieldcat to it_fieldcat.
clear wa_fieldcat.
wa_fieldcat-fieldname = 'BET01'.
wa_FIELDCAT-REPTEXT_DDIC = 'Amount for wagetype'.
append wa_fieldcat to it_fieldcat.
clear wa_fieldcat.
DATA: sort TYPE slis_sortinfo_alv,
it_sort TYPE slis_t_sortinfo_alv.
sort-fieldname = 'PERNR'.
sort-subtot = 'X'.
SORT-UP = 'X'.
APPEND sort TO it_sort.
*sort-fieldname = 'BEGDA'.
*SORT-NO_SUBTOTS = 'X'.
*APPEND sort TO it_sort.
IT_layout-totals_text = 'total text'.
IT_layout-subtotals_text = 'Subtotal text'.
*WA_EVENTS-NAME = 'SUBTOTAL TEXT'.
*WA_EVENTS-FORM = 'SUBTOTAL TEXT'.
*APPEND WA_EVENTS TO IT_EVENTS.
CALL FUNCTION 'REUSE_ALV_LIST_DISPLAY'
EXPORTING
 I_CALLBACK_PROGRAM = sy-repid
 IS_LAYOUT = it_LAYOUT
 IT_FIELDCAT = IT_FIELDCAT
 it_sort = it_sort
it_events = it_events
 TABLES
 t_outtab = it_pa0008 .
endform. " disp_alv

Few questions - game loop, data types, speed

Hello, I have a few questions after studying some topics in this forum regarding game creation:
1) What's the fastest way to wait in the game loop? I've seen two approaches:
thread.sleep(10)andsynchronized(this) { wait(10); }2) What data types shall I use? In C++ I use to prefer int over short in all cases, because 32bit hardware works faster with integers. Is this same on cell phones?
3) Speed of applications is slow. I just wonder wheter it's my fault. I was testing application, which only cleared the buffer and outputted FPS and I got around 20 frames. It was Nokia 6300 with 240x320 display. After testing on other phones I've found out that the bigger the resolution, the slower the game is going. Is this normal?
Thanks for replies...

1) You're not going to notice any really speed difference between the two code snippets. Read up on 'Threads', and you'll see why one may be used in place of the other depending on the situation. In general there may be a slight performance loss, however unnoticable, when using the synchronized version, but when you are multithreading it is likely necessary.
sleep(int) is impossible to interrupt, so it's generally a no-no in most situations. However we are talking about devices where every bit of performance helps, so as long as it works for ya, it's not a big deal.
2) The performance difference is fairly negligable, if any. The biggest thing to consider is memory requirements, and shorts take 1/2 the data.
Also, many phones don't support floating point data types, so you'll likely need to use ints/longs to calculate your values if you want to have any accuracy beyond whole numbers. Doing something like shifting bits or using 1000x values in your calculations can get around most of the problems when you can't use floats.
3) The biggest performance killers are IO, memory allocation, screen drawing; pretty much in that order. So I imagine that you are re-creating a new String object every time you output your FPS value on screen right? Doing that every frame would destroy any hopes of getting high-performance.
Just be careful, and never allocate objects when you can avoid it. anything where you concat String objects using + will cause your performance to die a horrible painful slow death. Remove anything that says 'new' from your main loop, and all String operations, and it'll likely speed things up a lot for ya.
Does your main loop have something like this?
g.drawString("FPS: " + currentFps, 0,0,Graphics.TOP | Graphics.LEFT);
This is very bad because of the String operation. It'll create a new String every frame.
If you have any more specicif questions, or you'd just like to pick the brain of a mobile game dev, stop by my messageboard:
http://attackgames.proboards84.com
Message was edited by:
hooble

New Effective CAL essay: How do I create an abstract data type in CAL?

How do I create an abstract data type in CAL? An abstract data type is one whose internal representation can be changed without needing to modify the source code of client modules that make use of that type. For software maintainability, it is a good idea to make a type that is subject to change or enhancement into an abstract data type. Another reason to create an abstract data type is to enforce invariants for values of the type that can only be ensured by using constructor functions (i.e. functions that return values of that type). In principle it is simple to create an abstract data type in CAL. For an algebraic data type, make the type constructor public and all data constructors private. For a foreign data type, make the type constructor public and the implementation scope private. If a scope qualifier is omitted, the scope is taken to be private. For example, the Map algebraic data type has the public type constructor Map and the data constructors Tip and Bin are each private, so it is an abstract data type. /** A map from keys (of type {@code k@}) to values (of type {@code a@}). */ data public Map k a = private Tip | private Bin size :: !Int key :: !k value :: a leftMap :: !(Map k a) rightMap :: !(Map k a); There are a number of invariants of this type: the size field represents the number of elements in the map represented by its Bin value. The keys in leftMap are all less than key, which in turn is less than all the keys in rightMap. In particular, non-empty Map values can only be created if the key parameter type is a member of the Ord type class. Values of the Map type can be created outside the Cal.Collections.Map module only by using constructor functions such as insert: insert :: Ord k => k -> a -> Map k a -> Map k a; public insert !key value !map = ... The owner of the Cal.Collections.Map module must ensure that all invariants of the Map type are satisfied, but if this is done, then it will automatically hold for clients using this function. Some examples of foreign abstract data types are Color, StringNoCase and RelativeDate: data foreign unsafe import jvm private "java.awt.Color" public Color; data foreign unsafe import jvm private "java.lang.String" public StringNoCase; data foreign unsafe import jvm private "int" public RelativeDate; The private implementation scope for Color means that a foreign function whose type involves Color can only be declared in the Cal.Graphics.Color module where the Color type is defined. A foreign function declaration involving the Color type relies on the compiler knowing that the Color type corresponds to java.awt.Color to resolve the corresponding Java entity i.e. it must know about the implementation of the Color type. Having a private implementation scope means that the Color type can be changed to correspond to a different Java class, or indeed to be an algebraic type, without the risk of breaking client code. In all these three cases there are useful, and different, design reasons to adopt a private implementation scope: For RelativeDate, the Java implementation type int represents a coded Gregorian date value in the date scheme used by Crystal Reports. Not all int values correspond to valid dates, and the algorithm to map an int to a year/month/day equivalent is fairly complicated, taking into account things like Gregorian calendar reform. Thus, it is desirable to hide the implementation of this type. For StringNoCase, the implementation is more straightforward as a java.lang.String. The reason to adopt a private implementation scope is to ensure that all functions involving StringNoCase preserve the semantics of StringNoCase as representing a case-insensitive string value. Otherwise it is very easy for clients to declare a function such as: foreign unsafe import jvm "method replace" replaceChar :: StringNoCase -> Char -> Char -> StringNoCase; which does not handle case-insensitivity correctly, but is a perfectly valid declaration. This declaration results in a compilation error when it is placed outside the module in which StringNoCase is defined because of the private implementation scope of StringNoCase. For Color, the issue is somewhat more subtle. The java.awt.Color implementation type is semantically the same as the CAL Color type. The problem is that java.awt.Color is mutable (since it can be sub-classed to create a mutable type). It is preferable for a first-class CAL type to not be mutable, so we simply make the implementation scope private to ensure that this will be the case. A somewhat less encapsulated kind of abstract data type can be created using friend modules and protected scope. For example, if an algebraic type is public, and all its data constructors are protected, then the data constructors can be accessed in the friend modules of the module in which the type is defined. Effectively this means that the implementation of the semantics of the type stretches over the module in which the type is defined, and all of its friend modules. These must all be checked if the implementation of the type is modified. Given the merits of abstract data types discussed above, it is perhaps surprising that most of the core types defined in the Prelude module are not abstract data types. For example: Boolean, Char, Int, Double, String, List, Maybe, Either, Ordering, JObject, JList, and all record and tuple types are non-abstract types. There are different reasons for this, depending on the particular type involved. For example, Boolean, List, Maybe, Either and Ordering are all rather canonical algebraic data types with a long history in functional languages, with many standard functions using them. They are thus guaranteed never to change. In addition, their values have no particular design invariants that need to be enforced via constructor functions. Exposing the data constructors gives clients some additional syntactic flexibility in using values of the type. For example, they can pattern match on the values using case expressions or let patterns. Essentially the same explanation holds for record and tuple types. Although non-tuple record types are less canonical, they do correspond to the fundamental notion of an anonymous named-field product type. The "anonymous" here simply means that the programmer can create an entirely new record type simply by creating a value; the type does not have to be declared anywhere prior to use. Char, Int, Double, String, JObject and JList are foreign types where in fact part of the semantics of the type is that we want clients to know that the type is a foreign type. For example, we want clients to know that Prelude.Int is essentially the Java primitive unboxed int type, and has all the semantics you would expect of the Java int type i.e. this is quite different from RelativeDate which is using int as its implementation type in a very tactical way that we may choose to change. One can think of a public foreign type declaration with public implementation scope as simply introducing the Java type into the CAL namespace. One interesting point here is with CAL's naming convention for public foreign types. We prefix a type name by "J" (for "Java") for foreign types with public implementation type such that the underlying Java type is mutable. This is intended as mnemonic that the type is not a pure functional type and thus some caution needs to be taken when using it. For example, Prelude.JObject has public Java implementation type java.lang.Object. In the case where the underlying Java type is not mutable, we do not use the prefix, since even though the type is foreign; it is basically a first class functional type and can be freely used without concern. For example, Prelude.String has public Java implementation type java.lang.String. In the case where the implementation type is private, then the fact that the type is a foreign type, whether mutable or not, is an implementation detail and we do not hint at that detail via the name. Thus Color.Color has as its private Java implementation type the mutable Java type java.awt.Color. When creating abstract data types it is important to not inadvertently supply public API functions that conflict with the desired public semantics of the type. For example, if the type is publicly a pure-functional (i.e. immutable) type such as Color, it is important not to expose functions that mutate the internal Java representation. A more subtle case of inadvertently exposing the implementation of a type can occur with derived instances. For example, deriving the Prelude.Outputable and Prelude.Inputable type classes on a foreign type, whose implementation type is a mutable Java reference type, allows the client to gain access to the underlying Java value and mutate it
(by calling Prelude.output, mutating, and then calling Prelude.input). The solution in this case is to not derive Inputable and Outputable instances, but rather to define a custom Inputable and Outputable instance that copies the underlying values.

Hi Pandra801,
When you create a the external content type, please try to add a filter based on your select statement.
http://arsalkhatri.wordpress.com/2012/01/07/external-list-with-bcs-search-filters-finders/
Or, try to create a stored procedure based on your select statement, then create ECT using the SQL stored procedure.
A step by step guide in designing BCS entities by using a SQL stored procedure
http://blogs.msdn.com/b/sharepointdev/archive/2011/02/10/173-a-step-by-step-guide-in-designing-bcs-entities-by-using-a-sql-stored-procedure.aspx
I hope this helps.
Thanks,
Wendy
Wendy Li
TechNet Community Support

How to use Call library function node for a function in dll with VOID data type

Hi All,
I would like to ask for your kind help,
I am facing an issue with the call library node.
I have a C++ function(stdcall), which has void as data type
error code XXXX(hwnd, lID, getValue, *void data1, *void data2)
the data1 and data2 types are always changing depending upoin the value of "getValue".
Primarily i can use call library node multiple times and adapt each node according to the data types of data1 and data2, and extract the values and use in the code. Here is no issue. Real question is:
My question:
How can i just use one time call library node and make a case depending upon the "getvalue", which will control the data type of data1 and data2. Here i really looking for solutions.
My trials:
i used varaints as input to the call libray node for data1 and data2, and selected Parameters in call libraby node as " Adapt to type". here labview just crashed.
i really appreciate your feedbackand suggestions.
Thanks
Kutbuddin
Solved!
Go to Solution.
Attachments:
Clipboard02.jpg ‏103 KB

A variant is a very specific LabVIEW datatype (really a C++ type object internally) and trying to pass that to a function, which excepts a flat memory pointer there, for sure will crash very quickly.
As to endianess, yes Unflatten will be able to adjust for endianess, which in this case however is most likely exectly NOT what you want. So make sure that the you select native type for the endianess input on Unflatten from String. LabVIEW internally works with whatever is the native endianess, as will most likely your C++ DLL. The platform independent big endian format does only come into play when you receive data streams over some streaming interface like a network connection. Here it is desirable to use an endian format that is independent from the actual platform that generates and consumes the data stream. LabVIEWs default endianes is big endian here.
But as long as you pass data directly to native components like DLLs there is no difference in endianess between what LabVIEW uses and what those components use.
Rolf Kalbermatter
CIT Engineering Netherlands
a division of Test & Measurement Solutions

ALV graph dependency on Data Types.

Similar Messages

Maybe you are looking for