For my application, I created 4 components, and want to 'install' only the applications and their docs. According to https://cmake.org/Wiki/CMake:CPackConfiguration, if in my config file I have the line
set(CPACK_INSTALL_CMAKE_PROJECTS "${CMAKE_CURRENT_BINARY_DIR};${CMAKE_PROJECT_NAME};ALL;/")
then all the four components are installed. If I use the component names instead like
set(CPACK_INSTALL_CMAKE_PROJECTS "${CMAKE_CURRENT_BINARY_DIR};${CMAKE_PROJECT_NAME};apps docs;/")
I receive a package with no component name appended, and it contains the top level requested directory but no files at all. If I provide only one component like
set(CPACK_INSTALL_CMAKE_PROJECTS "${CMAKE_CURRENT_BINARY_DIR};${CMAKE_PROJECT_NAME};apps;/")
the package name does not contain the name of the component, but the files are generated in their right position.
What do I wrong?
EDIT: Just to add more strange things:
For the ALL case 4 files with names install_manifest_+comp name are created. For the second case I have install_manifest_apps docs.txt of zero length, and for the third case I have install_manifest_apps.txt, with the right content.
It seems to me that the macro can only accept one argument which can be either ALL or a component name; this means it cannot be used to install a 2-component system.
Moreover: if I use a component name, make package only produces that one package, without appending the component name, but with the right content. Till now I guessed that make install and make package are independent.
Set multiple values to CPACK_INSTALL_CMAKE_PROJECTS, one for each component you want to include:
set( CPACK_INSTALL_CMAKE_PROJECTS
"${CMAKE_CURRENT_BINARY_DIR};${CMAKE_PROJECT_NAME};apps;/"
"${CMAKE_CURRENT_BINARY_DIR};${CMAKE_PROJECT_NAME};docs;/"
)
Related
I want to access a module description variable (a custom variable) while building an android module i.e. an executable or shared-library or prebuilt and based on its value do some extra processing on the executable or shared-lib or prebuilt.
Is there a way to do it?
Add the variable to the list of variables in clear_vars.mk , so that they get cleared when an Android.mk does include $(CLEAR_VARS)
I'm trying to make two packages using cmake 2.8.12: the first one contains a shared library, the second one contains headers files. Apparently, I want two packages with names libname and libname-dev respectively, so I used command install with no specified component for the shared library and the same command with COMPONENT dev for the headers files. Before this I had defined an empty default component name. So I have a code as follows:
set(CMAKE_INSTALL_DEFAULT_COMPONENT_NAME "")
install(TARGETS libname DESTINATION /usr/local/lib)
install(FILES ${header_files} DESTINATION /usr/local/include COMPONENT dev)
But as a result, despite the empty default component name libname-Unspecified and libname-dev are generated. I find this behavior quite weird, because it seems I have no way to define two packages with previously described names. Are there any ideas how to do this in any other way?
In case of debian:
Give your main component some temporary name, e.g. "runtime".
install(TARGETS my-target
COMPONENT runtime
DESTINATION ${INSTALLDIR}
)
Then remove the postfixed component name as follows:
set(CPACK_DEBIAN_RUNTIME_PACKAGE_NAME "${CPACK_PACKAGE_NAME}")
Might work with other generators, too if they provide something equivalent to "CPACK_DEBIAN_RUNTIME_PACKAGE_NAME".
I'm trying to pass a parameter into the File List component through input port 0. All of my attempts thus far have been met with an error,
Input edge has no effect. Disconnect edge or use metadata fields as parameters in Target URL, Source path or Target path.
Ideally, I would like the Target URL to be something along the lines of http://${S3_ACCESS_KEY}:${S3_SECRET_KEY}#${MY_BUCKET}.s3.amazonaws.com/reports/${port:0.value}/*_interestingReport.csv where ${port:0.value} is the value passed in from the input port.
What is the correct way to use data coming in on input port 0?
The way how passing parameters from input edge for File List (but other file components as well) works, is that you use the name of the metadata column from an input edge and enclose it between ${ and }.
So if the metadata on the edge have a field called directory, which contains the dir you want to use, this is the way how to do it.
http://${S3_ACCESS_KEY}:${S3_SECRET_KEY}#${MY_BUCKET}.s3.amazonaws.com/reports/${directory}/*_interestingReport.csv
Let me show you an example of a very simple graph which uses a 'Data Generator' that creates the flow and sends it as the input of a 'File List' component'.
http://www.filedropper.com/inputportfilelist_1
As you can see the way the input field is referenced is '${DATA_SOURCE_DIR}/${fileDir}/', being 'fileDir' the only field contained in the metadata of the link that connects both components. It'll basically list the files located in ${PROJECT}\data\source\manifests.
I hope this helps.
I have several existing merge modules and one created by me. Lets call my module as 'A'. My merge modules and one existing merge module(say B) contains a file with the same name(customUI.resources) but with different content.
I am making an installer which uses both modules A & B. I want to use the customUI.resources file from module 'A'. When i compile the installer.wix file, it gives me the error as:
light.exe : error LGHT0204 : ICE30: The target file 'va2q4hvb.res|CustomUI.resources' is installed in '[ProgramFilesFolder]\abc\' by two different components on an LFN system: '_5AD14D2401C1408A8B1117B8B85F53E3.7C263653_7A00_4D9C_B52A_39AF1892BC49' and 'Adapter_Config_ja.BABCE01A_0CB9_4A62_B877_5E520A53D609'.
This breaks component reference counting. How can i resolve this error?
Are you in control of authoring module a and b?
If so, author the files as 2 different filenames and use the CopyFile element ( DuplicateFile table ) to duplicate the specific name to the common desired name. Also put mutually exclusive conditions on the components to make sure only one or the other gets installed.
Then in your main installer set a property to decide which one gets installed. The result will pass all ICE validation tests and give you the desired behavior.
You could use Dark to decompile the merge modules and then either combine them to a single module or re-create merge module B without the conflicting file.
My application has 5 plugins. Each plugin has a perspective of it's own and hence each perspective extension definition is under individual plugin's plugin.xml.
Now, I want to control the order in which these perspectives appear in my application. How to do it?
There is one main plugin that holds "ApplicationWorkBenchAdvisor.java". This has initialize() method in which I am iterating through the perspective registry using
PlatformUI.getWorkbench().getPerspectiveRegistry().getPerspectives();
and then appending perspective ids in a comma separated fashion to a String variable (pbar) which is used later like this.
PlatformUI.getPreferenceStore().setDefault(IWorkbenchPreferenceConstants.PERSPECTIVE_BAR_EXTRAS, pbar);
PlatformUI.getPreferenceStore().setValue(IWorkbenchPreferenceConstants.PERSPECTIVE_BAR_EXTRAS, pbar);
When iterating thourgh the perspective registry, I can compare perspective ids and sort it(when adding to 'pbar' by comparing ids) the way I want it to appear but, I don't want to do this ordering here as it appears like a dirty way.
Is there any other place where we can fix the order in which perspectives appear? (Each perspective resides in different plugin).
ADDED
1) Could we also control the ordering in the perspective switcher?
2) Is there a way to control entry into perspective registry to in inflict the desired order. If not could we write back into perspective registry?
If your application is encapsulated as an eclipse product, you may tweak the plugin.properties/plugin_customization.ini file.
(file referenced by the 'preferenceCustomization' property in your product extension point.)
This file is a java.io.Properties format file. Typically this file is used to set the values for preferences that are published as part of a plug-in's public API.
(Example of such a file for org.eclipse.platform)
So if the string representing the order of perspective can be referenced as a property, you can define the default order in there.
Since the source code of IWorkbenchPreferenceConstants mentions:
/**
* Lists the extra perspectives to show in the perspective bar.
* The value is a comma-separated list of perspective ids.
* The default is the empty string.
*
* #since 3.2
*/
public static final String JavaDoc PERSPECTIVE_BAR_EXTRAS = "PERSPECTIVE_BAR_EXTRAS"; //$NON-NLS-1$
Maybe a line in the plugin_customization.ini file:
org.eclipse.ui/PERSPECTIVE_BAR_EXTRAS=perspectiveId1,perspectiveId2,perspectiveId3
would allow you to specify that order without having to hard-code it.
Additional notes:
IPerspectiveRegistry (or PerspectiveRegistry) is not made to write anything (especially for perspective defined in an extension)
Ordering may be found in the state of the workbench (stored in the workspace and then restored when its launched again, .metadata/.plugins/org.eclipse.ui.workbench/workbench.xml)
Do you confirm that:
IPerspectiveRegistry registry = PlatformUI.getWorkbench().getPerspectiveRegistry();
IPerspectiveDescriptor[] perspectives = registry.getPerspectives();
is not in the right order when the plugin_customization.ini does define that order correctly ?
Liverpool 5 - 0 Aston Villa does confirm that (in the comments), but also indicates the (ordered) ini file entries internally get recorded into preference store, which means they can be retrieved through the preference store API:
PatformUI.getPreferenceStore().getDefault(
IWorkbenchPreferenceConstants.PERSPECTIVE_BAR_EXTRAS)
Liverpool 5 - 0 Aston Villa then add:
perspective registry (the initial "PlatformUI.getWorkbench().getPerspectiveRegistry().getPerspectives();" bit) remains unaltered (and unordered).
But, you still can "readily access to ordered list of perspectives" through preference store.
So, for other tasks, instead of iterating though perspective registry (which is still unordered), we can use the ordered variable that stores list of ordered perpective ids.
.
.
.
.
Note: another possibility is to Replace the Perspective-Switcher in RCP apps
=> to:
You can more easily define the order in a menu or in buttons there.
Extreme solution: re-implement a perspective switcher.
To sum up all the observations and findings,
1) It is not possible to alter entries in the perspective registry. It is read-only.
2) To make perspective appear in the order that we want on perspective bar, we can achieve it by adding an entry in plugin_customization.ini (or preferences.ini) as shown below.
org.eclipse.ui/PERSPECTIVE_BAR_EXTRAS=perspectiveId1,perspectiveId2,perspectiveId3
3) If we want to fetch this ordered list, we can't fetch it directly. But as this ini file entry internally gets recorded in PreferenceStore we can fetch the same value from PreferenceStore using the following API as shown below.
PlatformUI.getPreferenceStore().getDefault(
IWorkbenchPreferenceConstants.PERSPECTIVE_BAR_EXTRAS);
Why would someone need to access the entry defined in ini file at all?
Well, in my case I had a view in which i had to display links to every perspective. As my perspective bar was sorted in desired order, I also wanted to maintain the same order in my view while displaying links to perspectives.
4) There is no known way to inflict the same sort order in the display of default perspective switcher. While a new custom perspective switcher can be written to achieve the desired effect.