CTest 2.8.11 Docs: Difference between revisions

From KitwarePublic
Jump to navigationJump to search
(Created page with "<pre> ctest version 2.8.11 ------------------------------------------------------------------------------ Name ctest - Testing driver provided by CMake. ---------------------...")
 
(Replaced content with "See https://cmake.org/cmake/help/v2.8.11/ctest.html")
 
Line 1: Line 1:
<pre>
See https://cmake.org/cmake/help/v2.8.11/ctest.html
ctest version 2.8.11
------------------------------------------------------------------------------
Name
 
  ctest - Testing driver provided by CMake.
 
------------------------------------------------------------------------------
Usage
 
  ctest [options]
 
------------------------------------------------------------------------------
Description
 
The "ctest" executable is the CMake test driver program.  CMake-generated
build trees created for projects that use the ENABLE_TESTING and ADD_TEST
commands have testing support.  This program will run the tests and report
results.
 
------------------------------------------------------------------------------
Options
 
  -C <cfg>, --build-config <cfg>
      Choose configuration to test.
 
      Some CMake-generated build trees can have multiple build
      configurations in the same tree.  This option can be used to specify
      which one should be tested.  Example configurations are "Debug" and
      "Release".
 
  -V,--verbose
      Enable verbose output from tests.
 
      Test output is normally suppressed and only summary information is
      displayed.  This option will show all test output.
 
  -VV,--extra-verbose
      Enable more verbose output from tests.
 
      Test output is normally suppressed and only summary information is
      displayed.  This option will show even more test output.
 
  --debug
      Displaying more verbose internals of CTest.
 
      This feature will result in large number of output that is mostly
      useful for debugging dashboard problems.
 
  --output-on-failure
      Output anything outputted by the test program if the test should fail.
      This option can also be enabled by setting the environment variable
      CTEST_OUTPUT_ON_FAILURE
 
  -F
      Enable failover.
 
      This option allows ctest to resume a test set execution that was
      previously interrupted.  If no interruption occurred, the -F option
      will have no effect.
 
  -j <jobs>, --parallel <jobs>
      Run the tests in parallel using thegiven number of jobs.
 
      This option tells ctest to run the tests in parallel using given
      number of jobs.
 
  -Q,--quiet
      Make ctest quiet.
 
      This option will suppress all the output.  The output log file will
      still be generated if the --output-log is specified.  Options such as
      --verbose, --extra-verbose, and --debug are ignored if --quiet is
      specified.
 
  -O <file>, --output-log <file>
      Output to log file
 
      This option tells ctest to write all its output to a log file.
 
  -N,--show-only
      Disable actual execution of tests.
 
      This option tells ctest to list the tests that would be run but not
      actually run them.  Useful in conjunction with the -R and -E options.
 
  -L <regex>, --label-regex <regex>
      Run tests with labels matching regular expression.
 
      This option tells ctest to run only the tests whose labels match the
      given regular expression.
 
  -R <regex>, --tests-regex <regex>
      Run tests matching regular expression.
 
      This option tells ctest to run only the tests whose names match the
      given regular expression.
 
  -E <regex>, --exclude-regex <regex>
      Exclude tests matching regular expression.
 
      This option tells ctest to NOT run the tests whose names match the
      given regular expression.
 
  -LE <regex>, --label-exclude <regex>
      Exclude tests with labels matching regular expression.
 
      This option tells ctest to NOT run the tests whose labels match the
      given regular expression.
 
  -D <dashboard>, --dashboard <dashboard>
      Execute dashboard test
 
      This option tells ctest to perform act as a Dart client and perform a
      dashboard test.  All tests are <Mode><Test>, where Mode can be
      Experimental, Nightly, and Continuous, and Test can be Start, Update,
      Configure, Build, Test, Coverage, and Submit.
 
  -D <var>:<type>=<value>
      Define a variable for script mode
 
      Pass in variable values on the command line.  Use in conjunction with
      -S to pass variable values to a dashboard script.  Parsing -D
      arguments as variable values is only attempted if the value following
      -D does not match any of the known dashboard types.
 
  -M <model>, --test-model <model>
      Sets the model for a dashboard
 
      This option tells ctest to act as a Dart client where the TestModel
      can be Experimental, Nightly, and Continuous.  Combining -M and -T is
      similar to -D
 
  -T <action>, --test-action <action>
      Sets the dashboard action to perform
 
      This option tells ctest to act as a Dart client and perform some
      action such as start, build, test etc.  Combining -M and -T is similar
      to -D
 
  --track <track>
      Specify the track to submit dashboard to
 
      Submit dashboard to specified track instead of default one.  By
      default, the dashboard is submitted to Nightly, Experimental, or
      Continuous track, but by specifying this option, the track can be
      arbitrary.
 
  -S <script>, --script <script>
      Execute a dashboard for a configuration
 
      This option tells ctest to load in a configuration script which sets a
      number of parameters such as the binary and source directories.  Then
      ctest will do what is required to create and run a dashboard.  This
      option basically sets up a dashboard and then runs ctest -D with the
      appropriate options.
 
  -SP <script>, --script-new-process <script>
      Execute a dashboard for a configuration
 
      This option does the same operations as -S but it will do them in a
      separate process.  This is primarily useful in cases where the script
      may modify the environment and you do not want the modified
      environment to impact other -S scripts.
 
  -A <file>, --add-notes <file>
      Add a notes file with submission
 
      This option tells ctest to include a notes file when submitting
      dashboard.
 
  -I [Start,End,Stride,test#,test#|Test file], --tests-information
      Run a specific number of tests by number.
 
      This option causes ctest to run tests starting at number Start, ending
      at number End, and incrementing by Stride.  Any additional numbers
      after Stride are considered individual test numbers.  Start, End,or
      stride can be empty.  Optionally a file can be given that contains the
      same syntax as the command line.
 
  -U, --union
      Take the Union of -I and -R
 
      When both -R and -I are specified by default the intersection of tests
      are run.  By specifying -U the union of tests is run instead.
 
  --max-width <width>
      Set the max width for a test name to output
 
      Set the maximum width for each test name to show in the output.  This
      allows the user to widen the output to avoid clipping the test name
      which can be very annoying.
 
  --interactive-debug-mode [0|1]
      Set the interactive mode to 0 or 1.
 
      This option causes ctest to run tests in either an interactive mode or
      a non-interactive mode.  On Windows this means that in non-interactive
      mode, all system debug pop up windows are blocked.  In dashboard mode
      (Experimental, Nightly, Continuous), the default is non-interactive.
      When just running tests not for a dashboard the default is to allow
      popups and interactive debugging.
 
  --no-label-summary
      Disable timing summary information for labels.
 
      This option tells ctest not to print summary information for each
      label associated with the tests run.  If there are no labels on the
      tests, nothing extra is printed.
 
  --build-and-test
      Configure, build and run a test.
 
      This option tells ctest to configure (i.e.  run cmake on), build, and
      or execute a test.  The configure and test steps are optional.  The
      arguments to this command line are the source and binary directories.
      By default this will run CMake on the Source/Bin directories specified
      unless --build-nocmake is specified.  Both --build-makeprogram and
      --build-generator MUST be provided to use --built-and-test.  If
      --test-command is specified then that will be run after the build is
      complete.  Other options that affect this mode are --build-target
      --build-nocmake, --build-run-dir, --build-two-config, --build-exe-dir,
      --build-project,--build-noclean, --build-options
 
  --build-target
      Specify a specific target to build.
 
      This option goes with the --build-and-test option, if left out the all
      target is built.
 
  --build-nocmake
      Run the build without running cmake first.
 
      Skip the cmake step.
 
  --build-run-dir
      Specify directory to run programs from.
 
      Directory where programs will be after it has been compiled.
 
  --build-two-config
      Run CMake twice
 
  --build-exe-dir
      Specify the directory for the executable.
 
  --build-generator
      Specify the generator to use.
 
  --build-generator-toolset
      Specify the generator-specific toolset.
 
  --build-project
      Specify the name of the project to build.
 
  --build-makeprogram
      Specify the make program to use.
 
  --build-noclean
      Skip the make clean step.
 
  --build-config-sample
      A sample executable to use to determine the configuration
 
      A sample executable to use to determine the configuration that should
      be used.  e.g.  Debug/Release/etc
 
  --build-options
      Add extra options to the build step.
 
      This option must be the last option with the exception of
      --test-command
 
  --test-command
      The test to run with the --build-and-test option.
 
  --test-timeout
      The time limit in seconds, internal use only.
 
  --tomorrow-tag
      Nightly or experimental starts with next day tag.
 
      This is useful if the build will not finish in one day.
 
  --ctest-config
      The configuration file used to initialize CTest state when submitting
      dashboards.
 
      This option tells CTest to use different initialization file instead
      of CTestConfiguration.tcl.  This way multiple initialization files can
      be used for example to submit to multiple dashboards.
 
  --overwrite
      Overwrite CTest configuration option.
 
      By default ctest uses configuration options from configuration file.
      This option will overwrite the configuration option.
 
  --extra-submit <file>[;<file>]
      Submit extra files to the dashboard.
 
      This option will submit extra files to the dashboard.
 
  --force-new-ctest-process
      Run child CTest instances as new processes
 
      By default CTest will run child CTest instances within the same
      process.  If this behavior is not desired, this argument will enforce
      new processes for child CTest processes.
 
  --schedule-random
      Use a random order for scheduling tests
 
      This option will run the tests in a random order.  It is commonly used
      to detect implicit dependencies in a test suite.
 
  --submit-index
      Submit individual dashboard tests with specific index
 
      This option allows performing the same CTest action (such as test)
      multiple times and submit all stages to the same dashboard (Dart2
      required).  Each execution requires different index.
 
  --timeout <seconds>
      Set a global timeout on all tests.
 
      This option will set a global timeout on all tests that do not already
      have a timeout set on them.
 
  --stop-time <time>
      Set a time at which all tests should stop running.
 
      Set a real time of day at which all tests should timeout.  Example:
      7:00:00 -0400.  Any time format understood by the curl date parser is
      accepted.  Local time is assumed if no timezone is specified.
 
  --http1.0
      Submit using HTTP 1.0.
 
      This option will force CTest to use HTTP 1.0 to submit files to the
      dashboard, instead of HTTP 1.1.
 
  --no-compress-output
      Do not compress test output when submitting.
 
      This flag will turn off automatic compression of test output.  Use
      this to maintain compatibility with an older version of CDash which
      doesn't support compressed test output.
 
  --print-labels
      Print all available test labels.
 
      This option will not run any tests, it will simply print the list of
      all labels associated with the test set.
 
  --help-command <cmd> [<file>]
      Show help for a single command and exit.
 
      Prints the help for the command to stdout or to the specified file.
 
  --help-command-list [<file>]
      List available commands and exit.
 
      Prints the list of all available listfile commands to stdout or the
      specified file.
 
  --help-commands [<file>]
      Print help for all commands and exit.
 
      Prints the help for all commands to stdout or to the specified file.
 
  --copyright [file]
      Print the CMake copyright and exit.
 
      If a file is specified, the copyright is written into it.
 
  --help,-help,-usage,-h,-H,/?
      Print usage information and exit.
 
      Usage describes the basic command line interface and its options.
 
  --help-full [file]
      Print full help and exit.
 
      Full help displays most of the documentation provided by the UNIX man
      page.  It is provided for use on non-UNIX platforms, but is also
      convenient if the man page is not installed.  If a file is specified,
      the help is written into it.
 
  --help-html [file]
      Print full help in HTML format.
 
      This option is used by CMake authors to help produce web pages.  If a
      file is specified, the help is written into it.
 
  --help-man [file]
      Print full help as a UNIX man page and exit.
 
      This option is used by the cmake build to generate the UNIX man page.
      If a file is specified, the help is written into it.
 
  --version,-version,/V [file]
      Show program name/version banner and exit.
 
      If a file is specified, the version is written into it.
 
------------------------------------------------------------------------------
Generators
 
The following generators are available on this platform:
 
------------------------------------------------------------------------------
Commands
 
  break
      Break from an enclosing foreach or while loop.
 
        break()
 
      Breaks from an enclosing foreach loop or while loop
 
  build_name
      Deprecated.  Use ${CMAKE_SYSTEM} and ${CMAKE_CXX_COMPILER} instead.
 
        build_name(variable)
 
      Sets the specified variable to a string representing the platform and
      compiler settings.  These values are now available through the
      CMAKE_SYSTEM and CMAKE_CXX_COMPILER variables.
 
  cmake_minimum_required
      Set the minimum required version of cmake for a project.
 
        cmake_minimum_required(VERSION major[.minor[.patch[.tweak]]]
                                [FATAL_ERROR])
 
      If the current version of CMake is lower than that required it will
      stop processing the project and report an error.  When a version
      higher than 2.4 is specified the command implicitly invokes
 
        cmake_policy(VERSION major[.minor[.patch[.tweak]]])
 
      which sets the cmake policy version level to the version specified.
      When version 2.4 or lower is given the command implicitly invokes
 
        cmake_policy(VERSION 2.4)
 
      which enables compatibility features for CMake 2.4 and lower.
 
      The FATAL_ERROR option is accepted but ignored by CMake 2.6 and
      higher.  It should be specified so CMake versions 2.4 and lower fail
      with an error instead of just a warning.
 
  cmake_policy
      Manage CMake Policy settings.
 
      As CMake evolves it is sometimes necessary to change existing behavior
      in order to fix bugs or improve implementations of existing features.
      The CMake Policy mechanism is designed to help keep existing projects
      building as new versions of CMake introduce changes in behavior.  Each
      new policy (behavioral change) is given an identifier of the form
      "CMP<NNNN>" where "<NNNN>" is an integer index.  Documentation
      associated with each policy describes the OLD and NEW behavior and the
      reason the policy was introduced.  Projects may set each policy to
      select the desired behavior.  When CMake needs to know which behavior
      to use it checks for a setting specified by the project.  If no
      setting is available the OLD behavior is assumed and a warning is
      produced requesting that the policy be set.
 
      The cmake_policy command is used to set policies to OLD or NEW
      behavior.  While setting policies individually is supported, we
      encourage projects to set policies based on CMake versions.
 
        cmake_policy(VERSION major.minor[.patch[.tweak]])
 
      Specify that the current CMake list file is written for the given
      version of CMake.  All policies introduced in the specified version or
      earlier will be set to use NEW behavior.  All policies introduced
      after the specified version will be unset (unless variable
      CMAKE_POLICY_DEFAULT_CMP<NNNN> sets a default).  This effectively
      requests behavior preferred as of a given CMake version and tells
      newer CMake versions to warn about their new policies.  The policy
      version specified must be at least 2.4 or the command will report an
      error.  In order to get compatibility features supporting versions
      earlier than 2.4 see documentation of policy CMP0001.
 
        cmake_policy(SET CMP<NNNN> NEW)
        cmake_policy(SET CMP<NNNN> OLD)
 
      Tell CMake to use the OLD or NEW behavior for a given policy.
      Projects depending on the old behavior of a given policy may silence a
      policy warning by setting the policy state to OLD.  Alternatively one
      may fix the project to work with the new behavior and set the policy
      state to NEW.
 
        cmake_policy(GET CMP<NNNN> <variable>)
 
      Check whether a given policy is set to OLD or NEW behavior.  The
      output variable value will be "OLD" or "NEW" if the policy is set, and
      empty otherwise.
 
      CMake keeps policy settings on a stack, so changes made by the
      cmake_policy command affect only the top of the stack.  A new entry on
      the policy stack is managed automatically for each subdirectory to
      protect its parents and siblings.  CMake also manages a new entry for
      scripts loaded by include() and find_package() commands except when
      invoked with the NO_POLICY_SCOPE option (see also policy CMP0011).
      The cmake_policy command provides an interface to manage custom
      entries on the policy stack:
 
        cmake_policy(PUSH)
        cmake_policy(POP)
 
      Each PUSH must have a matching POP to erase any changes.  This is
      useful to make temporary changes to policy settings.
 
      Functions and macros record policy settings when they are created and
      use the pre-record policies when they are invoked.  If the function or
      macro implementation sets policies, the changes automatically
      propagate up through callers until they reach the closest nested
      policy stack entry.
 
  configure_file
      Copy a file to another location and modify its contents.
 
        configure_file(<input> <output>
                        [COPYONLY] [ESCAPE_QUOTES] [@ONLY]
                        [NEWLINE_STYLE [UNIX|DOS|WIN32|LF|CRLF] ])
 
      Copies a file <input> to file <output> and substitutes variable values
      referenced in the file content.  If <input> is a relative path it is
      evaluated with respect to the current source directory.  The <input>
      must be a file, not a directory.  If <output> is a relative path it is
      evaluated with respect to the current binary directory.  If <output>
      names an existing directory the input file is placed in that directory
      with its original name.
 
      If the <input> file is modified the build system will re-run CMake to
      re-configure the file and generate the build system again.
 
      This command replaces any variables in the input file referenced as
      ${VAR} or @VAR@ with their values as determined by CMake.  If a
      variable is not defined, it will be replaced with nothing.  If
      COPYONLY is specified, then no variable expansion will take place.  If
      ESCAPE_QUOTES is specified then any substituted quotes will be C-style
      escaped.  The file will be configured with the current values of CMake
      variables.  If @ONLY is specified, only variables of the form @VAR@
      will be replaced and ${VAR} will be ignored.  This is useful for
      configuring scripts that use ${VAR}.
 
      Input file lines of the form "#cmakedefine VAR ..." will be replaced
      with either "#define VAR ..." or "/* #undef VAR */" depending on
      whether VAR is set in CMake to any value not considered a false
      constant by the if() command.  (Content of "...", if any, is processed
      as above.) Input file lines of the form "#cmakedefine01 VAR" will be
      replaced with either "#define VAR 1" or "#define VAR 0" similarly.
 
      With NEWLINE_STYLE the line ending could be adjusted:
 
          'UNIX' or 'LF' for \n, 'DOS', 'WIN32' or 'CRLF' for \r\n.
 
      COPYONLY must not be used with NEWLINE_STYLE.
 
 
  ctest_build
      Build the project.
 
        ctest_build([BUILD build_dir] [TARGET target] [RETURN_VALUE res]
                    [APPEND][NUMBER_ERRORS val] [NUMBER_WARNINGS val])
 
      Builds the given build directory and stores results in Build.xml.  If
      no BUILD is given, the CTEST_BINARY_DIRECTORY variable is used.
 
      The TARGET variable can be used to specify a build target.  If none is
      specified, the "all" target will be built.
 
      The RETURN_VALUE option specifies a variable in which to store the
      return value of the native build tool.  The NUMBER_ERRORS and
      NUMBER_WARNINGS options specify variables in which to store the number
      of build errors and warnings detected.
 
      The APPEND option marks results for append to those previously
      submitted to a dashboard server since the last ctest_start.  Append
      semantics are defined by the dashboard server in use.
 
  ctest_configure
      Configure the project build tree.
 
        ctest_configure([BUILD build_dir] [SOURCE source_dir] [APPEND]
                        [OPTIONS options] [RETURN_VALUE res])
 
      Configures the given build directory and stores results in
      Configure.xml.  If no BUILD is given, the CTEST_BINARY_DIRECTORY
      variable is used.  If no SOURCE is given, the CTEST_SOURCE_DIRECTORY
      variable is used.  The OPTIONS argument specifies command line
      arguments to pass to the configuration tool.  The RETURN_VALUE option
      specifies a variable in which to store the return value of the native
      build tool.
 
      The APPEND option marks results for append to those previously
      submitted to a dashboard server since the last ctest_start.  Append
      semantics are defined by the dashboard server in use.
 
  ctest_coverage
      Collect coverage tool results.
 
        ctest_coverage([BUILD build_dir] [RETURN_VALUE res] [APPEND]
                        [LABELS label1 [label2 [...]]])
 
      Perform the coverage of the given build directory and stores results
      in Coverage.xml.  The second argument is a variable that will hold
      value.
 
      The LABELS option filters the coverage report to include only source
      files labeled with at least one of the labels specified.
 
      The APPEND option marks results for append to those previously
      submitted to a dashboard server since the last ctest_start.  Append
      semantics are defined by the dashboard server in use.
 
  ctest_empty_binary_directory
      empties the binary directory
 
        ctest_empty_binary_directory( directory )
 
      Removes a binary directory.  This command will perform some checks
      prior to deleting the directory in an attempt to avoid malicious or
      accidental directory deletion.
 
  ctest_memcheck
      Run tests with a dynamic analysis tool.
 
        ctest_memcheck([BUILD build_dir] [RETURN_VALUE res] [APPEND]
                    [START start number] [END end number]
                    [STRIDE stride number] [EXCLUDE exclude regex ]
                    [INCLUDE include regex]
                    [EXCLUDE_LABEL exclude regex]
                    [INCLUDE_LABEL label regex]
                    [PARALLEL_LEVEL level] )
 
      Tests the given build directory and stores results in MemCheck.xml.
      The second argument is a variable that will hold value.  Optionally,
      you can specify the starting test number START, the ending test number
      END, the number of tests to skip between each test STRIDE, a regular
      expression for tests to run INCLUDE, or a regular expression for tests
      not to run EXCLUDE.  EXCLUDE_LABEL and INCLUDE_LABEL are regular
      expressions for tests to be included or excluded by the test property
      LABEL.  PARALLEL_LEVEL should be set to a positive number representing
      the number of tests to be run in parallel.
 
      The APPEND option marks results for append to those previously
      submitted to a dashboard server since the last ctest_start.  Append
      semantics are defined by the dashboard server in use.
 
  ctest_read_custom_files
      read CTestCustom files.
 
        ctest_read_custom_files( directory ... )
 
      Read all the CTestCustom.ctest or CTestCustom.cmake files from the
      given directory.
 
  ctest_run_script
      runs a ctest -S script
 
        ctest_run_script([NEW_PROCESS] script_file_name script_file_name1
                    script_file_name2 ... [RETURN_VALUE var])
 
      Runs a script or scripts much like if it was run from ctest -S.  If no
      argument is provided then the current script is run using the current
      settings of the variables.  If NEW_PROCESS is specified then each
      script will be run in a separate process.If RETURN_VALUE is specified
      the return value of the last script run will be put into var.
 
  ctest_sleep
      sleeps for some amount of time
 
        ctest_sleep(<seconds>)
 
      Sleep for given number of seconds.
 
        ctest_sleep(<time1> <duration> <time2>)
 
      Sleep for t=(time1 + duration - time2) seconds if t > 0.
 
  ctest_start
      Starts the testing for a given model
 
        ctest_start(Model [TRACK <track>] [APPEND] [source [binary]])
 
      Starts the testing for a given model.  The command should be called
      after the binary directory is initialized.  If the 'source' and
      'binary' directory are not specified, it reads the
      CTEST_SOURCE_DIRECTORY and CTEST_BINARY_DIRECTORY.  If the track is
      specified, the submissions will go to the specified track.  If APPEND
      is used, the existing TAG is used rather than creating a new one based
      on the current time stamp.
 
  ctest_submit
      Submit results to a dashboard server.
 
        ctest_submit([PARTS ...] [FILES ...] [RETRY_COUNT count]                [RETRY_DELAY delay][RETURN_VALUE res])
 
      By default all available parts are submitted if no PARTS or FILES are
      specified.  The PARTS option lists a subset of parts to be submitted.
      Valid part names are:
 
        Start      = nothing
        Update    = ctest_update results, in Update.xml
        Configure  = ctest_configure results, in Configure.xml
        Build      = ctest_build results, in Build.xml
        Test      = ctest_test results, in Test.xml
        Coverage  = ctest_coverage results, in Coverage.xml
        MemCheck  = ctest_memcheck results, in DynamicAnalysis.xml
        Notes      = Files listed by CTEST_NOTES_FILES, in Notes.xml
        ExtraFiles = Files listed by CTEST_EXTRA_SUBMIT_FILES
        Submit    = nothing
 
      The FILES option explicitly lists specific files to be submitted.
      Each individual file must exist at the time of the call.
 
      The RETRY_DELAY option specifies how long in seconds to wait after a
      timed-out submission before attempting to re-submit.
 
      The RETRY_COUNT option specifies how many times to retry a timed-out
      submission.
 
 
  ctest_test
      Run tests in the project build tree.
 
        ctest_test([BUILD build_dir] [APPEND]
                    [START start number] [END end number]
                    [STRIDE stride number] [EXCLUDE exclude regex ]
                    [INCLUDE include regex] [RETURN_VALUE res]
                    [EXCLUDE_LABEL exclude regex]
                    [INCLUDE_LABEL label regex]
                    [PARALLEL_LEVEL level]
                    [SCHEDULE_RANDOM on]
                    [STOP_TIME time of day])
 
      Tests the given build directory and stores results in Test.xml.  The
      second argument is a variable that will hold value.  Optionally, you
      can specify the starting test number START, the ending test number
      END, the number of tests to skip between each test STRIDE, a regular
      expression for tests to run INCLUDE, or a regular expression for tests
      to not run EXCLUDE.  EXCLUDE_LABEL and INCLUDE_LABEL are regular
      expression for test to be included or excluded by the test property
      LABEL.  PARALLEL_LEVEL should be set to a positive number representing
      the number of tests to be run in parallel.  SCHEDULE_RANDOM will
      launch tests in a random order, and is typically used to detect
      implicit test dependencies.  STOP_TIME is the time of day at which the
      tests should all stop running.
 
      The APPEND option marks results for append to those previously
      submitted to a dashboard server since the last ctest_start.  Append
      semantics are defined by the dashboard server in use.
 
  ctest_update
      Update the work tree from version control.
 
        ctest_update([SOURCE source] [RETURN_VALUE res])
 
      Updates the given source directory and stores results in Update.xml.
      If no SOURCE is given, the CTEST_SOURCE_DIRECTORY variable is used.
      The RETURN_VALUE option specifies a variable in which to store the
      result, which is the number of files updated or -1 on error.
 
  ctest_upload
      Upload files to a dashboard server.
 
        ctest_upload(FILES ...)
 
      Pass a list of files to be sent along with the build results to the
      dashboard server.
 
 
  else
      Starts the else portion of an if block.
 
        else(expression)
 
      See the if command.
 
  elseif
      Starts the elseif portion of an if block.
 
        elseif(expression)
 
      See the if command.
 
  endforeach
      Ends a list of commands in a FOREACH block.
 
        endforeach(expression)
 
      See the FOREACH command.
 
  endfunction
      Ends a list of commands in a function block.
 
        endfunction(expression)
 
      See the function command.
 
  endif
      Ends a list of commands in an if block.
 
        endif(expression)
 
      See the if command.
 
  endmacro
      Ends a list of commands in a macro block.
 
        endmacro(expression)
 
      See the macro command.
 
  endwhile
      Ends a list of commands in a while block.
 
        endwhile(expression)
 
      See the while command.
 
  exec_program
      Deprecated.  Use the execute_process() command instead.
 
      Run an executable program during the processing of the CMakeList.txt
      file.
 
        exec_program(Executable [directory in which to run]
                      [ARGS <arguments to executable>]
                      [OUTPUT_VARIABLE <var>]
                      [RETURN_VALUE <var>])
 
      The executable is run in the optionally specified directory.  The
      executable can include arguments if it is double quoted, but it is
      better to use the optional ARGS argument to specify arguments to the
      program.  This is because cmake will then be able to escape spaces in
      the executable path.  An optional argument OUTPUT_VARIABLE specifies a
      variable in which to store the output.  To capture the return value of
      the execution, provide a RETURN_VALUE.  If OUTPUT_VARIABLE is
      specified, then no output will go to the stdout/stderr of the console
      running cmake.
 
 
  execute_process
      Execute one or more child processes.
 
        execute_process(COMMAND <cmd1> [args1...]]
                        [COMMAND <cmd2> [args2...] [...]]
                        [WORKING_DIRECTORY <directory>]
                        [TIMEOUT <seconds>]
                        [RESULT_VARIABLE <variable>]
                        [OUTPUT_VARIABLE <variable>]
                        [ERROR_VARIABLE <variable>]
                        [INPUT_FILE <file>]
                        [OUTPUT_FILE <file>]
                        [ERROR_FILE <file>]
                        [OUTPUT_QUIET]
                        [ERROR_QUIET]
                        [OUTPUT_STRIP_TRAILING_WHITESPACE]
                        [ERROR_STRIP_TRAILING_WHITESPACE])
 
      Runs the given sequence of one or more commands with the standard
      output of each process piped to the standard input of the next.  A
      single standard error pipe is used for all processes.  If
      WORKING_DIRECTORY is given the named directory will be set as the
      current working directory of the child processes.  If TIMEOUT is given
      the child processes will be terminated if they do not finish in the
      specified number of seconds (fractions are allowed).  If
      RESULT_VARIABLE is given the variable will be set to contain the
      result of running the processes.  This will be an integer return code
      from the last child or a string describing an error condition.  If
      OUTPUT_VARIABLE or ERROR_VARIABLE are given the variable named will be
      set with the contents of the standard output and standard error pipes
      respectively.  If the same variable is named for both pipes their
      output will be merged in the order produced.  If INPUT_FILE,
      OUTPUT_FILE, or ERROR_FILE is given the file named will be attached to
      the standard input of the first process, standard output of the last
      process, or standard error of all processes respectively.  If
      OUTPUT_QUIET or ERROR_QUIET is given then the standard output or
      standard error results will be quietly ignored.  If more than one
      OUTPUT_* or ERROR_* option is given for the same pipe the precedence
      is not specified.  If no OUTPUT_* or ERROR_* options are given the
      output will be shared with the corresponding pipes of the CMake
      process itself.
 
      The execute_process command is a newer more powerful version of
      exec_program, but the old command has been kept for compatibility.
 
  file
      File manipulation command.
 
        file(WRITE filename "message to write"... )
        file(APPEND filename "message to write"... )
        file(READ filename variable [LIMIT numBytes] [OFFSET offset] [HEX])
        file(<MD5|SHA1|SHA224|SHA256|SHA384|SHA512> filename variable)
        file(STRINGS filename variable [LIMIT_COUNT num]
              [LIMIT_INPUT numBytes] [LIMIT_OUTPUT numBytes]
              [LENGTH_MINIMUM numBytes] [LENGTH_MAXIMUM numBytes]
              [NEWLINE_CONSUME] [REGEX regex]
              [NO_HEX_CONVERSION])
        file(GLOB variable [RELATIVE path] [globbing expressions]...)
        file(GLOB_RECURSE variable [RELATIVE path]
              [FOLLOW_SYMLINKS] [globbing expressions]...)
        file(RENAME <oldname> <newname>)
        file(REMOVE [file1 ...])
        file(REMOVE_RECURSE [file1 ...])
        file(MAKE_DIRECTORY [directory1 directory2 ...])
        file(RELATIVE_PATH variable directory file)
        file(TO_CMAKE_PATH path result)
        file(TO_NATIVE_PATH path result)
        file(DOWNLOAD url file [INACTIVITY_TIMEOUT timeout]
              [TIMEOUT timeout] [STATUS status] [LOG log] [SHOW_PROGRESS]
              [EXPECTED_HASH ALGO=value] [EXPECTED_MD5 sum]
              [TLS_VERIFY on|off] [TLS_CAINFO file])
        file(UPLOAD filename url [INACTIVITY_TIMEOUT timeout]
              [TIMEOUT timeout] [STATUS status] [LOG log] [SHOW_PROGRESS])
        file(TIMESTAMP filename variable [<format string>] [UTC])
 
      WRITE will write a message into a file called 'filename'.  It
      overwrites the file if it already exists, and creates the file if it
      does not exist.  (If the file is a build input, use configure_file to
      update the file only when its content changes.)
 
      APPEND will write a message into a file same as WRITE, except it will
      append it to the end of the file
 
      READ will read the content of a file and store it into the variable.
      It will start at the given offset and read up to numBytes.  If the
      argument HEX is given, the binary data will be converted to
      hexadecimal representation and this will be stored in the variable.
 
      MD5, SHA1, SHA224, SHA256, SHA384, and SHA512 will compute a
      cryptographic hash of the content of a file.
 
      STRINGS will parse a list of ASCII strings from a file and store it in
      a variable.  Binary data in the file are ignored.  Carriage return
      (CR) characters are ignored.  It works also for Intel Hex and Motorola
      S-record files, which are automatically converted to binary format
      when reading them.  Disable this using NO_HEX_CONVERSION.
 
      LIMIT_COUNT sets the maximum number of strings to return.  LIMIT_INPUT
      sets the maximum number of bytes to read from the input file.
      LIMIT_OUTPUT sets the maximum number of bytes to store in the output
      variable.  LENGTH_MINIMUM sets the minimum length of a string to
      return.  Shorter strings are ignored.  LENGTH_MAXIMUM sets the maximum
      length of a string to return.  Longer strings are split into strings
      no longer than the maximum length.  NEWLINE_CONSUME allows newlines to
      be included in strings instead of terminating them.
 
      REGEX specifies a regular expression that a string must match to be
      returned.  Typical usage
 
        file(STRINGS myfile.txt myfile)
 
      stores a list in the variable "myfile" in which each item is a line
      from the input file.
 
      GLOB will generate a list of all files that match the globbing
      expressions and store it into the variable.  Globbing expressions are
      similar to regular expressions, but much simpler.  If RELATIVE flag is
      specified for an expression, the results will be returned as a
      relative path to the given path.  (We do not recommend using GLOB to
      collect a list of source files from your source tree.  If no
      CMakeLists.txt file changes when a source is added or removed then the
      generated build system cannot know when to ask CMake to regenerate.)
 
      Examples of globbing expressions include:
 
          *.cxx      - match all files with extension cxx
          *.vt?      - match all files with extension vta,...,vtz
          f[3-5].txt - match files f3.txt, f4.txt, f5.txt
 
      GLOB_RECURSE will generate a list similar to the regular GLOB, except
      it will traverse all the subdirectories of the matched directory and
      match the files.  Subdirectories that are symlinks are only traversed
      if FOLLOW_SYMLINKS is given or cmake policy CMP0009 is not set to NEW.
      See cmake --help-policy CMP0009 for more information.
 
      Examples of recursive globbing include:
 
          /dir/*.py  - match all python files in /dir and subdirectories
 
      MAKE_DIRECTORY will create the given directories, also if their parent
      directories don't exist yet
 
      RENAME moves a file or directory within a filesystem, replacing the
      destination atomically.
 
      REMOVE will remove the given files, also in subdirectories
 
      REMOVE_RECURSE will remove the given files and directories, also
      non-empty directories
 
      RELATIVE_PATH will determine relative path from directory to the given
      file.
 
      TO_CMAKE_PATH will convert path into a cmake style path with unix /.
      The input can be a single path or a system path like "$ENV{PATH}".
      Note the double quotes around the ENV call TO_CMAKE_PATH only takes
      one argument.  This command will also convert the native list
      delimiters for a list of paths like the PATH environment variable.
 
      TO_NATIVE_PATH works just like TO_CMAKE_PATH, but will convert from a
      cmake style path into the native path style \ for windows and / for
      UNIX.
 
      DOWNLOAD will download the given URL to the given file.  If LOG var is
      specified a log of the download will be put in var.  If STATUS var is
      specified the status of the operation will be put in var.  The status
      is returned in a list of length 2.  The first element is the numeric
      return value for the operation, and the second element is a string
      value for the error.  A 0 numeric error means no error in the
      operation.  If TIMEOUT time is specified, the operation will timeout
      after time seconds, time should be specified as an integer.  The
      INACTIVITY_TIMEOUT specifies an integer number of seconds of
      inactivity after which the operation should terminate.  If
      EXPECTED_HASH ALGO=value is specified, the operation will verify that
      the downloaded file's actual hash matches the expected value, where
      ALGO is one of MD5, SHA1, SHA224, SHA256, SHA384, or SHA512.  If it
      does not match, the operation fails with an error.  ("EXPECTED_MD5
      sum" is short-hand for "EXPECTED_HASH MD5=sum".) If SHOW_PROGRESS is
      specified, progress information will be printed as status messages
      until the operation is complete.  For https URLs CMake must be built
      with OpenSSL.  TLS/SSL certificates are not checked by default.  Set
      TLS_VERIFY to ON to check certificates and/or use EXPECTED_HASH to
      verify downloaded content.  Set TLS_CAINFO to specify a custom
      Certificate Authority file.  If either TLS option is not given CMake
      will check variables CMAKE_TLS_VERIFY and CMAKE_TLS_CAINFO,
      respectively.
 
      UPLOAD will upload the given file to the given URL.  If LOG var is
      specified a log of the upload will be put in var.  If STATUS var is
      specified the status of the operation will be put in var.  The status
      is returned in a list of length 2.  The first element is the numeric
      return value for the operation, and the second element is a string
      value for the error.  A 0 numeric error means no error in the
      operation.  If TIMEOUT time is specified, the operation will timeout
      after time seconds, time should be specified as an integer.  The
      INACTIVITY_TIMEOUT specifies an integer number of seconds of
      inactivity after which the operation should terminate.  If
      SHOW_PROGRESS is specified, progress information will be printed as
      status messages until the operation is complete.
 
      TIMESTAMP will write a string representation of the modification time
      of filename to variable.
 
      Should the command be unable to obtain a timestamp variable will be
      set to the empty string "".
 
      See documentation of the string TIMESTAMP sub-command for more
      details.
 
      The file() command also provides COPY and INSTALL signatures:
 
        file(<COPY|INSTALL> files... DESTINATION <dir>
              [FILE_PERMISSIONS permissions...]
              [DIRECTORY_PERMISSIONS permissions...]
              [NO_SOURCE_PERMISSIONS] [USE_SOURCE_PERMISSIONS]
              [FILES_MATCHING]
              [[PATTERN <pattern> | REGEX <regex>]
              [EXCLUDE] [PERMISSIONS permissions...]] [...])
 
      The COPY signature copies files, directories, and symlinks to a
      destination folder.  Relative input paths are evaluated with respect
      to the current source directory, and a relative destination is
      evaluated with respect to the current build directory.  Copying
      preserves input file timestamps, and optimizes out a file if it exists
      at the destination with the same timestamp.  Copying preserves input
      permissions unless explicit permissions or NO_SOURCE_PERMISSIONS are
      given (default is USE_SOURCE_PERMISSIONS).  See the install(DIRECTORY)
      command for documentation of permissions, PATTERN, REGEX, and EXCLUDE
      options.
 
      The INSTALL signature differs slightly from COPY: it prints status
      messages, and NO_SOURCE_PERMISSIONS is default.  Installation scripts
      generated by the install() command use this signature (with some
      undocumented options for internal use).
 
  find_file
      Find the full path to a file.
 
          find_file(<VAR> name1 [path1 path2 ...])
 
      This is the short-hand signature for the command that is sufficient in
      many cases.  It is the same as find_file(<VAR> name1 [PATHS path1
      path2 ...])
 
          find_file(
                    <VAR>
                    name | NAMES name1 [name2 ...]
                    [HINTS path1 [path2 ... ENV var]]
                    [PATHS path1 [path2 ... ENV var]]
                    [PATH_SUFFIXES suffix1 [suffix2 ...]]
                    [DOC "cache documentation string"]
                    [NO_DEFAULT_PATH]
                    [NO_CMAKE_ENVIRONMENT_PATH]
                    [NO_CMAKE_PATH]
                    [NO_SYSTEM_ENVIRONMENT_PATH]
                    [NO_CMAKE_SYSTEM_PATH]
                    [CMAKE_FIND_ROOT_PATH_BOTH |
                    ONLY_CMAKE_FIND_ROOT_PATH |
                    NO_CMAKE_FIND_ROOT_PATH]
                  )
 
      This command is used to find a full path to named file.  A cache entry
      named by <VAR> is created to store the result of this command.  If the
      full path to a file is found the result is stored in the variable and
      the search will not be repeated unless the variable is cleared.  If
      nothing is found, the result will be <VAR>-NOTFOUND, and the search
      will be attempted again the next time find_file is invoked with the
      same variable.  The name of the full path to a file that is searched
      for is specified by the names listed after the NAMES argument.
      Additional search locations can be specified after the PATHS argument.
      If ENV var is found in the HINTS or PATHS section the environment
      variable var will be read and converted from a system environment
      variable to a cmake style list of paths.  For example ENV PATH would
      be a way to list the system path variable.  The argument after DOC
      will be used for the documentation string in the cache.  PATH_SUFFIXES
      specifies additional subdirectories to check below each search path.
 
      If NO_DEFAULT_PATH is specified, then no additional paths are added to
      the search.  If NO_DEFAULT_PATH is not specified, the search process
      is as follows:
 
      1.  Search paths specified in cmake-specific cache variables.  These
      are intended to be used on the command line with a -DVAR=value.  This
      can be skipped if NO_CMAKE_PATH is passed.
 
          <prefix>/include/<arch> if CMAKE_LIBRARY_ARCHITECTURE is set, and
          <prefix>/include for each <prefix> in CMAKE_PREFIX_PATH
          CMAKE_INCLUDE_PATH
          CMAKE_FRAMEWORK_PATH
 
      2.  Search paths specified in cmake-specific environment variables.
      These are intended to be set in the user's shell configuration.  This
      can be skipped if NO_CMAKE_ENVIRONMENT_PATH is passed.
 
          <prefix>/include/<arch> if CMAKE_LIBRARY_ARCHITECTURE is set, and
          <prefix>/include for each <prefix> in CMAKE_PREFIX_PATH
          CMAKE_INCLUDE_PATH
          CMAKE_FRAMEWORK_PATH
 
      3.  Search the paths specified by the HINTS option.  These should be
      paths computed by system introspection, such as a hint provided by the
      location of another item already found.  Hard-coded guesses should be
      specified with the PATHS option.
 
      4.  Search the standard system environment variables.  This can be
      skipped if NO_SYSTEM_ENVIRONMENT_PATH is an argument.
 
          PATH
          INCLUDE
 
      5.  Search cmake variables defined in the Platform files for the
      current system.  This can be skipped if NO_CMAKE_SYSTEM_PATH is
      passed.
 
          <prefix>/include/<arch> if CMAKE_LIBRARY_ARCHITECTURE is set, and
          <prefix>/include for each <prefix> in CMAKE_SYSTEM_PREFIX_PATH
          CMAKE_SYSTEM_INCLUDE_PATH
          CMAKE_SYSTEM_FRAMEWORK_PATH
 
      6.  Search the paths specified by the PATHS option or in the
      short-hand version of the command.  These are typically hard-coded
      guesses.
 
      On Darwin or systems supporting OS X Frameworks, the cmake variable
      CMAKE_FIND_FRAMEWORK can be set to empty or one of the following:
 
          "FIRST"  - Try to find frameworks before standard
                    libraries or headers. This is the default on Darwin.
          "LAST"  - Try to find frameworks after standard
                    libraries or headers.
          "ONLY"  - Only try to find frameworks.
          "NEVER" - Never try to find frameworks.
 
      On Darwin or systems supporting OS X Application Bundles, the cmake
      variable CMAKE_FIND_APPBUNDLE can be set to empty or one of the
      following:
 
          "FIRST"  - Try to find application bundles before standard
                    programs. This is the default on Darwin.
          "LAST"  - Try to find application bundles after standard
                    programs.
          "ONLY"  - Only try to find application bundles.
          "NEVER" - Never try to find application bundles.
 
      The CMake variable CMAKE_FIND_ROOT_PATH specifies one or more
      directories to be prepended to all other search directories.  This
      effectively "re-roots" the entire search under given locations.  By
      default it is empty.  It is especially useful when cross-compiling to
      point to the root directory of the target environment and CMake will
      search there too.  By default at first the directories listed in
      CMAKE_FIND_ROOT_PATH and then the non-rooted directories will be
      searched.  The default behavior can be adjusted by setting
      CMAKE_FIND_ROOT_PATH_MODE_INCLUDE.  This behavior can be manually
      overridden on a per-call basis.  By using CMAKE_FIND_ROOT_PATH_BOTH
      the search order will be as described above.  If
      NO_CMAKE_FIND_ROOT_PATH is used then CMAKE_FIND_ROOT_PATH will not be
      used.  If ONLY_CMAKE_FIND_ROOT_PATH is used then only the re-rooted
      directories will be searched.
 
      The default search order is designed to be most-specific to
      least-specific for common use cases.  Projects may override the order
      by simply calling the command multiple times and using the NO_*
      options:
 
          find_file(<VAR> NAMES name PATHS paths... NO_DEFAULT_PATH)
          find_file(<VAR> NAMES name)
 
      Once one of the calls succeeds the result variable will be set and
      stored in the cache so that no call will search again.
 
  find_library
      Find a library.
 
          find_library(<VAR> name1 [path1 path2 ...])
 
      This is the short-hand signature for the command that is sufficient in
      many cases.  It is the same as find_library(<VAR> name1 [PATHS path1
      path2 ...])
 
          find_library(
                    <VAR>
                    name | NAMES name1 [name2 ...] [NAMES_PER_DIR]
                    [HINTS path1 [path2 ... ENV var]]
                    [PATHS path1 [path2 ... ENV var]]
                    [PATH_SUFFIXES suffix1 [suffix2 ...]]
                    [DOC "cache documentation string"]
                    [NO_DEFAULT_PATH]
                    [NO_CMAKE_ENVIRONMENT_PATH]
                    [NO_CMAKE_PATH]
                    [NO_SYSTEM_ENVIRONMENT_PATH]
                    [NO_CMAKE_SYSTEM_PATH]
                    [CMAKE_FIND_ROOT_PATH_BOTH |
                    ONLY_CMAKE_FIND_ROOT_PATH |
                    NO_CMAKE_FIND_ROOT_PATH]
                  )
 
      This command is used to find a library.  A cache entry named by <VAR>
      is created to store the result of this command.  If the library is
      found the result is stored in the variable and the search will not be
      repeated unless the variable is cleared.  If nothing is found, the
      result will be <VAR>-NOTFOUND, and the search will be attempted again
      the next time find_library is invoked with the same variable.  The
      name of the library that is searched for is specified by the names
      listed after the NAMES argument.  Additional search locations can be
      specified after the PATHS argument.  If ENV var is found in the HINTS
      or PATHS section the environment variable var will be read and
      converted from a system environment variable to a cmake style list of
      paths.  For example ENV PATH would be a way to list the system path
      variable.  The argument after DOC will be used for the documentation
      string in the cache.  PATH_SUFFIXES specifies additional
      subdirectories to check below each search path.
 
      If NO_DEFAULT_PATH is specified, then no additional paths are added to
      the search.  If NO_DEFAULT_PATH is not specified, the search process
      is as follows:
 
      1.  Search paths specified in cmake-specific cache variables.  These
      are intended to be used on the command line with a -DVAR=value.  This
      can be skipped if NO_CMAKE_PATH is passed.
 
          <prefix>/lib/<arch> if CMAKE_LIBRARY_ARCHITECTURE is set, and
          <prefix>/lib for each <prefix> in CMAKE_PREFIX_PATH
          CMAKE_LIBRARY_PATH
          CMAKE_FRAMEWORK_PATH
 
      2.  Search paths specified in cmake-specific environment variables.
      These are intended to be set in the user's shell configuration.  This
      can be skipped if NO_CMAKE_ENVIRONMENT_PATH is passed.
 
          <prefix>/lib/<arch> if CMAKE_LIBRARY_ARCHITECTURE is set, and
          <prefix>/lib for each <prefix> in CMAKE_PREFIX_PATH
          CMAKE_LIBRARY_PATH
          CMAKE_FRAMEWORK_PATH
 
      3.  Search the paths specified by the HINTS option.  These should be
      paths computed by system introspection, such as a hint provided by the
      location of another item already found.  Hard-coded guesses should be
      specified with the PATHS option.
 
      4.  Search the standard system environment variables.  This can be
      skipped if NO_SYSTEM_ENVIRONMENT_PATH is an argument.
 
          PATH
          LIB
 
      5.  Search cmake variables defined in the Platform files for the
      current system.  This can be skipped if NO_CMAKE_SYSTEM_PATH is
      passed.
 
          <prefix>/lib/<arch> if CMAKE_LIBRARY_ARCHITECTURE is set, and
          <prefix>/lib for each <prefix> in CMAKE_SYSTEM_PREFIX_PATH
          CMAKE_SYSTEM_LIBRARY_PATH
          CMAKE_SYSTEM_FRAMEWORK_PATH
 
      6.  Search the paths specified by the PATHS option or in the
      short-hand version of the command.  These are typically hard-coded
      guesses.
 
      On Darwin or systems supporting OS X Frameworks, the cmake variable
      CMAKE_FIND_FRAMEWORK can be set to empty or one of the following:
 
          "FIRST"  - Try to find frameworks before standard
                    libraries or headers. This is the default on Darwin.
          "LAST"  - Try to find frameworks after standard
                    libraries or headers.
          "ONLY"  - Only try to find frameworks.
          "NEVER" - Never try to find frameworks.
 
      On Darwin or systems supporting OS X Application Bundles, the cmake
      variable CMAKE_FIND_APPBUNDLE can be set to empty or one of the
      following:
 
          "FIRST"  - Try to find application bundles before standard
                    programs. This is the default on Darwin.
          "LAST"  - Try to find application bundles after standard
                    programs.
          "ONLY"  - Only try to find application bundles.
          "NEVER" - Never try to find application bundles.
 
      The CMake variable CMAKE_FIND_ROOT_PATH specifies one or more
      directories to be prepended to all other search directories.  This
      effectively "re-roots" the entire search under given locations.  By
      default it is empty.  It is especially useful when cross-compiling to
      point to the root directory of the target environment and CMake will
      search there too.  By default at first the directories listed in
      CMAKE_FIND_ROOT_PATH and then the non-rooted directories will be
      searched.  The default behavior can be adjusted by setting
      CMAKE_FIND_ROOT_PATH_MODE_LIBRARY.  This behavior can be manually
      overridden on a per-call basis.  By using CMAKE_FIND_ROOT_PATH_BOTH
      the search order will be as described above.  If
      NO_CMAKE_FIND_ROOT_PATH is used then CMAKE_FIND_ROOT_PATH will not be
      used.  If ONLY_CMAKE_FIND_ROOT_PATH is used then only the re-rooted
      directories will be searched.
 
      The default search order is designed to be most-specific to
      least-specific for common use cases.  Projects may override the order
      by simply calling the command multiple times and using the NO_*
      options:
 
          find_library(<VAR> NAMES name PATHS paths... NO_DEFAULT_PATH)
          find_library(<VAR> NAMES name)
 
      Once one of the calls succeeds the result variable will be set and
      stored in the cache so that no call will search again.
 
      When more than one value is given to the NAMES option this command by
      default will consider one name at a time and search every directory
      for it.  The NAMES_PER_DIR option tells this command to consider one
      directory at a time and search for all names in it.
 
      If the library found is a framework, then VAR will be set to the full
      path to the framework <fullPath>/A.framework.  When a full path to a
      framework is used as a library, CMake will use a -framework A, and a
      -F<fullPath> to link the framework to the target.
 
      If the global property FIND_LIBRARY_USE_LIB64_PATHS is set all search
      paths will be tested as normal, with "64/" appended, and with all
      matches of "lib/" replaced with "lib64/".  This property is
      automatically set for the platforms that are known to need it if at
      least one of the languages supported by the PROJECT command is
      enabled.
 
  find_package
      Load settings for an external project.
 
        find_package(<package> [version] [EXACT] [QUIET] [MODULE]
                      [REQUIRED] [[COMPONENTS] [components...]]
                      [OPTIONAL_COMPONENTS components...]
                      [NO_POLICY_SCOPE])
 
      Finds and loads settings from an external project.  <package>_FOUND
      will be set to indicate whether the package was found.  When the
      package is found package-specific information is provided through
      variables and imported targets documented by the package itself.  The
      QUIET option disables messages if the package cannot be found.  The
      MODULE option disables the second signature documented below.  The
      REQUIRED option stops processing with an error message if the package
      cannot be found.
 
      A package-specific list of required components may be listed after the
      COMPONENTS option (or after the REQUIRED option if present).
      Additional optional components may be listed after
      OPTIONAL_COMPONENTS.  Available components and their influence on
      whether a package is considered to be found are defined by the target
      package.
 
      The [version] argument requests a version with which the package found
      should be compatible (format is major[.minor[.patch[.tweak]]]).  The
      EXACT option requests that the version be matched exactly.  If no
      [version] and/or component list is given to a recursive invocation
      inside a find-module, the corresponding arguments are forwarded
      automatically from the outer call (including the EXACT flag for
      [version]).  Version support is currently provided only on a
      package-by-package basis (details below).
 
      User code should generally look for packages using the above simple
      signature.  The remainder of this command documentation specifies the
      full command signature and details of the search process.  Project
      maintainers wishing to provide a package to be found by this command
      are encouraged to read on.
 
      The command has two modes by which it searches for packages: "Module"
      mode and "Config" mode.  Module mode is available when the command is
      invoked with the above reduced signature.  CMake searches for a file
      called "Find<package>.cmake" in the CMAKE_MODULE_PATH followed by the
      CMake installation.  If the file is found, it is read and processed by
      CMake.  It is responsible for finding the package, checking the
      version, and producing any needed messages.  Many find-modules provide
      limited or no support for versioning; check the module documentation.
      If no module is found and the MODULE option is not given the command
      proceeds to Config mode.
 
      The complete Config mode command signature is:
 
        find_package(<package> [version] [EXACT] [QUIET]
                      [REQUIRED] [[COMPONENTS] [components...]]
                      [CONFIG|NO_MODULE]
                      [NO_POLICY_SCOPE]
                      [NAMES name1 [name2 ...]]
                      [CONFIGS config1 [config2 ...]]
                      [HINTS path1 [path2 ... ]]
                      [PATHS path1 [path2 ... ]]
                      [PATH_SUFFIXES suffix1 [suffix2 ...]]
                      [NO_DEFAULT_PATH]
                      [NO_CMAKE_ENVIRONMENT_PATH]
                      [NO_CMAKE_PATH]
                      [NO_SYSTEM_ENVIRONMENT_PATH]
                      [NO_CMAKE_PACKAGE_REGISTRY]
                      [NO_CMAKE_BUILDS_PATH]
                      [NO_CMAKE_SYSTEM_PATH]
                      [NO_CMAKE_SYSTEM_PACKAGE_REGISTRY]
                      [CMAKE_FIND_ROOT_PATH_BOTH |
                      ONLY_CMAKE_FIND_ROOT_PATH |
                      NO_CMAKE_FIND_ROOT_PATH])
 
      The CONFIG option may be used to skip Module mode explicitly and
      switch to Config mode.  It is synonymous to using NO_MODULE.  Config
      mode is also implied by use of options not specified in the reduced
      signature.
 
      Config mode attempts to locate a configuration file provided by the
      package to be found.  A cache entry called <package>_DIR is created to
      hold the directory containing the file.  By default the command
      searches for a package with the name <package>.  If the NAMES option
      is given the names following it are used instead of <package>.  The
      command searches for a file called "<name>Config.cmake" or
      "<lower-case-name>-config.cmake" for each name specified.  A
      replacement set of possible configuration file names may be given
      using the CONFIGS option.  The search procedure is specified below.
      Once found, the configuration file is read and processed by CMake.
      Since the file is provided by the package it already knows the
      location of package contents.  The full path to the configuration file
      is stored in the cmake variable <package>_CONFIG.
 
      All configuration files which have been considered by CMake while
      searching for an installation of the package with an appropriate
      version are stored in the cmake variable <package>_CONSIDERED_CONFIGS,
      the associated versions in <package>_CONSIDERED_VERSIONS.
 
      If the package configuration file cannot be found CMake will generate
      an error describing the problem unless the QUIET argument is
      specified.  If REQUIRED is specified and the package is not found a
      fatal error is generated and the configure step stops executing.  If
      <package>_DIR has been set to a directory not containing a
      configuration file CMake will ignore it and search from scratch.
 
      When the [version] argument is given Config mode will only find a
      version of the package that claims compatibility with the requested
      version (format is major[.minor[.patch[.tweak]]]).  If the EXACT
      option is given only a version of the package claiming an exact match
      of the requested version may be found.  CMake does not establish any
      convention for the meaning of version numbers.  Package version
      numbers are checked by "version" files provided by the packages
      themselves.  For a candidate package configuration file
      "<config-file>.cmake" the corresponding version file is located next
      to it and named either "<config-file>-version.cmake" or
      "<config-file>Version.cmake".  If no such version file is available
      then the configuration file is assumed to not be compatible with any
      requested version.  A basic version file containing generic version
      matching code can be created using the macro
      write_basic_package_version_file(), see its documentation for more
      details.  When a version file is found it is loaded to check the
      requested version number.  The version file is loaded in a nested
      scope in which the following variables have been defined:
 
        PACKAGE_FIND_NAME          = the <package> name
        PACKAGE_FIND_VERSION      = full requested version string
        PACKAGE_FIND_VERSION_MAJOR = major version if requested, else 0
        PACKAGE_FIND_VERSION_MINOR = minor version if requested, else 0
        PACKAGE_FIND_VERSION_PATCH = patch version if requested, else 0
        PACKAGE_FIND_VERSION_TWEAK = tweak version if requested, else 0
        PACKAGE_FIND_VERSION_COUNT = number of version components, 0 to 4
 
      The version file checks whether it satisfies the requested version and
      sets these variables:
 
        PACKAGE_VERSION            = full provided version string
        PACKAGE_VERSION_EXACT      = true if version is exact match
        PACKAGE_VERSION_COMPATIBLE = true if version is compatible
        PACKAGE_VERSION_UNSUITABLE = true if unsuitable as any version
 
      These variables are checked by the find_package command to determine
      whether the configuration file provides an acceptable version.  They
      are not available after the find_package call returns.  If the version
      is acceptable the following variables are set:
 
        <package>_VERSION      = full provided version string
        <package>_VERSION_MAJOR = major version if provided, else 0
        <package>_VERSION_MINOR = minor version if provided, else 0
        <package>_VERSION_PATCH = patch version if provided, else 0
        <package>_VERSION_TWEAK = tweak version if provided, else 0
        <package>_VERSION_COUNT = number of version components, 0 to 4
 
      and the corresponding package configuration file is loaded.  When
      multiple package configuration files are available whose version files
      claim compatibility with the version requested it is unspecified which
      one is chosen.  No attempt is made to choose a highest or closest
      version number.
 
      Config mode provides an elaborate interface and search procedure.
      Much of the interface is provided for completeness and for use
      internally by find-modules loaded by Module mode.  Most user code
      should simply call
 
        find_package(<package> [major[.minor]] [EXACT] [REQUIRED|QUIET])
 
      in order to find a package.  Package maintainers providing CMake
      package configuration files are encouraged to name and install them
      such that the procedure outlined below will find them without
      requiring use of additional options.
 
      CMake constructs a set of possible installation prefixes for the
      package.  Under each prefix several directories are searched for a
      configuration file.  The tables below show the directories searched.
      Each entry is meant for installation trees following Windows (W), UNIX
      (U), or Apple (A) conventions.
 
        <prefix>/                                              (W)
        <prefix>/(cmake|CMake)/                                (W)
        <prefix>/<name>*/                                      (W)
        <prefix>/<name>*/(cmake|CMake)/                        (W)
        <prefix>/(lib/<arch>|lib|share)/cmake/<name>*/          (U)
        <prefix>/(lib/<arch>|lib|share)/<name>*/                (U)
        <prefix>/(lib/<arch>|lib|share)/<name>*/(cmake|CMake)/  (U)
 
      On systems supporting OS X Frameworks and Application Bundles the
      following directories are searched for frameworks or bundles
      containing a configuration file:
 
        <prefix>/<name>.framework/Resources/                    (A)
        <prefix>/<name>.framework/Resources/CMake/              (A)
        <prefix>/<name>.framework/Versions/*/Resources/        (A)
        <prefix>/<name>.framework/Versions/*/Resources/CMake/  (A)
        <prefix>/<name>.app/Contents/Resources/                (A)
        <prefix>/<name>.app/Contents/Resources/CMake/          (A)
 
      In all cases the <name> is treated as case-insensitive and corresponds
      to any of the names specified (<package> or names given by NAMES).
      Paths with lib/<arch> are enabled if CMAKE_LIBRARY_ARCHITECTURE is
      set.  If PATH_SUFFIXES is specified the suffixes are appended to each
      (W) or (U) directory entry one-by-one.
 
      This set of directories is intended to work in cooperation with
      projects that provide configuration files in their installation trees.
      Directories above marked with (W) are intended for installations on
      Windows where the prefix may point at the top of an application's
      installation directory.  Those marked with (U) are intended for
      installations on UNIX platforms where the prefix is shared by multiple
      packages.  This is merely a convention, so all (W) and (U) directories
      are still searched on all platforms.  Directories marked with (A) are
      intended for installations on Apple platforms.  The cmake variables
      CMAKE_FIND_FRAMEWORK and CMAKE_FIND_APPBUNDLE determine the order of
      preference as specified below.
 
      The set of installation prefixes is constructed using the following
      steps.  If NO_DEFAULT_PATH is specified all NO_* options are enabled.
 
      1.  Search paths specified in cmake-specific cache variables.  These
      are intended to be used on the command line with a -DVAR=value.  This
      can be skipped if NO_CMAKE_PATH is passed.
 
          CMAKE_PREFIX_PATH
          CMAKE_FRAMEWORK_PATH
          CMAKE_APPBUNDLE_PATH
 
      2.  Search paths specified in cmake-specific environment variables.
      These are intended to be set in the user's shell configuration.  This
      can be skipped if NO_CMAKE_ENVIRONMENT_PATH is passed.
 
          <package>_DIR
          CMAKE_PREFIX_PATH
          CMAKE_FRAMEWORK_PATH
          CMAKE_APPBUNDLE_PATH
 
      3.  Search paths specified by the HINTS option.  These should be paths
      computed by system introspection, such as a hint provided by the
      location of another item already found.  Hard-coded guesses should be
      specified with the PATHS option.
 
      4.  Search the standard system environment variables.  This can be
      skipped if NO_SYSTEM_ENVIRONMENT_PATH is passed.  Path entries ending
      in "/bin" or "/sbin" are automatically converted to their parent
      directories.
 
          PATH
 
      5.  Search project build trees recently configured in a CMake GUI.
      This can be skipped if NO_CMAKE_BUILDS_PATH is passed.  It is intended
      for the case when a user is building multiple dependent projects one
      after another.
 
      6.  Search paths stored in the CMake user package registry.  This can
      be skipped if NO_CMAKE_PACKAGE_REGISTRY is passed.  On Windows a
      <package> may appear under registry key
 
        HKEY_CURRENT_USER\Software\Kitware\CMake\Packages\<package>
 
      as a REG_SZ value, with arbitrary name, that specifies the directory
      containing the package configuration file.  On UNIX platforms a
      <package> may appear under the directory
 
        ~/.cmake/packages/<package>
 
      as a file, with arbitrary name, whose content specifies the directory
      containing the package configuration file. See the export(PACKAGE)
      command to create user package registry entries for project build
      trees.
 
      7.  Search cmake variables defined in the Platform files for the
      current system.  This can be skipped if NO_CMAKE_SYSTEM_PATH is
      passed.
 
          CMAKE_SYSTEM_PREFIX_PATH
          CMAKE_SYSTEM_FRAMEWORK_PATH
          CMAKE_SYSTEM_APPBUNDLE_PATH
 
      8. Search paths stored in the CMake system package registry.  This
      can be skipped if NO_CMAKE_SYSTEM_PACKAGE_REGISTRY is passed.  On
      Windows a <package> may appear under registry key
 
        HKEY_LOCAL_MACHINE\Software\Kitware\CMake\Packages\<package>
 
      as a REG_SZ value, with arbitrary name, that specifies the directory
      containing the package configuration file.  There is no system package
      registry on non-Windows platforms.
 
      9.  Search paths specified by the PATHS option.  These are typically
      hard-coded guesses.
 
      On Darwin or systems supporting OS X Frameworks, the cmake variable
      CMAKE_FIND_FRAMEWORK can be set to empty or one of the following:
 
          "FIRST"  - Try to find frameworks before standard
                    libraries or headers. This is the default on Darwin.
          "LAST"  - Try to find frameworks after standard
                    libraries or headers.
          "ONLY"  - Only try to find frameworks.
          "NEVER" - Never try to find frameworks.
 
      On Darwin or systems supporting OS X Application Bundles, the cmake
      variable CMAKE_FIND_APPBUNDLE can be set to empty or one of the
      following:
 
          "FIRST"  - Try to find application bundles before standard
                    programs. This is the default on Darwin.
          "LAST"  - Try to find application bundles after standard
                    programs.
          "ONLY"  - Only try to find application bundles.
          "NEVER" - Never try to find application bundles.
 
      The CMake variable CMAKE_FIND_ROOT_PATH specifies one or more
      directories to be prepended to all other search directories.  This
      effectively "re-roots" the entire search under given locations.  By
      default it is empty.  It is especially useful when cross-compiling to
      point to the root directory of the target environment and CMake will
      search there too.  By default at first the directories listed in
      CMAKE_FIND_ROOT_PATH and then the non-rooted directories will be
      searched.  The default behavior can be adjusted by setting
      CMAKE_FIND_ROOT_PATH_MODE_PACKAGE.  This behavior can be manually
      overridden on a per-call basis.  By using CMAKE_FIND_ROOT_PATH_BOTH
      the search order will be as described above.  If
      NO_CMAKE_FIND_ROOT_PATH is used then CMAKE_FIND_ROOT_PATH will not be
      used.  If ONLY_CMAKE_FIND_ROOT_PATH is used then only the re-rooted
      directories will be searched.
 
      The default search order is designed to be most-specific to
      least-specific for common use cases.  Projects may override the order
      by simply calling the command multiple times and using the NO_*
      options:
 
          find_package(<package> PATHS paths... NO_DEFAULT_PATH)
          find_package(<package>)
 
      Once one of the calls succeeds the result variable will be set and
      stored in the cache so that no call will search again.
 
      Every non-REQUIRED find_package() call can be disabled by setting the
      variable CMAKE_DISABLE_FIND_PACKAGE_<package> to TRUE.  See the
      documentation for the CMAKE_DISABLE_FIND_PACKAGE_<package> variable
      for more information.
 
      When loading a find module or package configuration file find_package
      defines variables to provide information about the call arguments (and
      restores their original state before returning):
 
        <package>_FIND_REQUIRED      = true if REQUIRED option was given
        <package>_FIND_QUIETLY      = true if QUIET option was given
        <package>_FIND_VERSION      = full requested version string
        <package>_FIND_VERSION_MAJOR = major version if requested, else 0
        <package>_FIND_VERSION_MINOR = minor version if requested, else 0
        <package>_FIND_VERSION_PATCH = patch version if requested, else 0
        <package>_FIND_VERSION_TWEAK = tweak version if requested, else 0
        <package>_FIND_VERSION_COUNT = number of version components, 0 to 4
        <package>_FIND_VERSION_EXACT = true if EXACT option was given
        <package>_FIND_COMPONENTS    = list of requested components
        <package>_FIND_REQUIRED_<c>  = true if component <c> is required
                                      false if component <c> is optional
 
      In Module mode the loaded find module is responsible to honor the
      request detailed by these variables; see the find module for details.
      In Config mode find_package handles REQUIRED, QUIET, and version
      options automatically but leaves it to the package configuration file
      to handle components in a way that makes sense for the package.  The
      package configuration file may set <package>_FOUND to false to tell
      find_package that component requirements are not satisfied.
 
      See the cmake_policy() command documentation for discussion of the
      NO_POLICY_SCOPE option.
 
  find_path
      Find the directory containing a file.
 
          find_path(<VAR> name1 [path1 path2 ...])
 
      This is the short-hand signature for the command that is sufficient in
      many cases.  It is the same as find_path(<VAR> name1 [PATHS path1
      path2 ...])
 
          find_path(
                    <VAR>
                    name | NAMES name1 [name2 ...]
                    [HINTS path1 [path2 ... ENV var]]
                    [PATHS path1 [path2 ... ENV var]]
                    [PATH_SUFFIXES suffix1 [suffix2 ...]]
                    [DOC "cache documentation string"]
                    [NO_DEFAULT_PATH]
                    [NO_CMAKE_ENVIRONMENT_PATH]
                    [NO_CMAKE_PATH]
                    [NO_SYSTEM_ENVIRONMENT_PATH]
                    [NO_CMAKE_SYSTEM_PATH]
                    [CMAKE_FIND_ROOT_PATH_BOTH |
                    ONLY_CMAKE_FIND_ROOT_PATH |
                    NO_CMAKE_FIND_ROOT_PATH]
                  )
 
      This command is used to find a directory containing the named file.  A
      cache entry named by <VAR> is created to store the result of this
      command.  If the file in a directory is found the result is stored in
      the variable and the search will not be repeated unless the variable
      is cleared.  If nothing is found, the result will be <VAR>-NOTFOUND,
      and the search will be attempted again the next time find_path is
      invoked with the same variable.  The name of the file in a directory
      that is searched for is specified by the names listed after the NAMES
      argument.  Additional search locations can be specified after the
      PATHS argument.  If ENV var is found in the HINTS or PATHS section the
      environment variable var will be read and converted from a system
      environment variable to a cmake style list of paths.  For example ENV
      PATH would be a way to list the system path variable.  The argument
      after DOC will be used for the documentation string in the cache.
      PATH_SUFFIXES specifies additional subdirectories to check below each
      search path.
 
      If NO_DEFAULT_PATH is specified, then no additional paths are added to
      the search.  If NO_DEFAULT_PATH is not specified, the search process
      is as follows:
 
      1.  Search paths specified in cmake-specific cache variables.  These
      are intended to be used on the command line with a -DVAR=value.  This
      can be skipped if NO_CMAKE_PATH is passed.
 
          <prefix>/include/<arch> if CMAKE_LIBRARY_ARCHITECTURE is set, and
          <prefix>/include for each <prefix> in CMAKE_PREFIX_PATH
          CMAKE_INCLUDE_PATH
          CMAKE_FRAMEWORK_PATH
 
      2.  Search paths specified in cmake-specific environment variables.
      These are intended to be set in the user's shell configuration.  This
      can be skipped if NO_CMAKE_ENVIRONMENT_PATH is passed.
 
          <prefix>/include/<arch> if CMAKE_LIBRARY_ARCHITECTURE is set, and
          <prefix>/include for each <prefix> in CMAKE_PREFIX_PATH
          CMAKE_INCLUDE_PATH
          CMAKE_FRAMEWORK_PATH
 
      3.  Search the paths specified by the HINTS option.  These should be
      paths computed by system introspection, such as a hint provided by the
      location of another item already found.  Hard-coded guesses should be
      specified with the PATHS option.
 
      4.  Search the standard system environment variables.  This can be
      skipped if NO_SYSTEM_ENVIRONMENT_PATH is an argument.
 
          PATH
          INCLUDE
 
      5.  Search cmake variables defined in the Platform files for the
      current system.  This can be skipped if NO_CMAKE_SYSTEM_PATH is
      passed.
 
          <prefix>/include/<arch> if CMAKE_LIBRARY_ARCHITECTURE is set, and
          <prefix>/include for each <prefix> in CMAKE_SYSTEM_PREFIX_PATH
          CMAKE_SYSTEM_INCLUDE_PATH
          CMAKE_SYSTEM_FRAMEWORK_PATH
 
      6.  Search the paths specified by the PATHS option or in the
      short-hand version of the command.  These are typically hard-coded
      guesses.
 
      On Darwin or systems supporting OS X Frameworks, the cmake variable
      CMAKE_FIND_FRAMEWORK can be set to empty or one of the following:
 
          "FIRST"  - Try to find frameworks before standard
                    libraries or headers. This is the default on Darwin.
          "LAST"  - Try to find frameworks after standard
                    libraries or headers.
          "ONLY"  - Only try to find frameworks.
          "NEVER" - Never try to find frameworks.
 
      On Darwin or systems supporting OS X Application Bundles, the cmake
      variable CMAKE_FIND_APPBUNDLE can be set to empty or one of the
      following:
 
          "FIRST"  - Try to find application bundles before standard
                    programs. This is the default on Darwin.
          "LAST"  - Try to find application bundles after standard
                    programs.
          "ONLY"  - Only try to find application bundles.
          "NEVER" - Never try to find application bundles.
 
      The CMake variable CMAKE_FIND_ROOT_PATH specifies one or more
      directories to be prepended to all other search directories.  This
      effectively "re-roots" the entire search under given locations.  By
      default it is empty.  It is especially useful when cross-compiling to
      point to the root directory of the target environment and CMake will
      search there too.  By default at first the directories listed in
      CMAKE_FIND_ROOT_PATH and then the non-rooted directories will be
      searched.  The default behavior can be adjusted by setting
      CMAKE_FIND_ROOT_PATH_MODE_INCLUDE.  This behavior can be manually
      overridden on a per-call basis.  By using CMAKE_FIND_ROOT_PATH_BOTH
      the search order will be as described above.  If
      NO_CMAKE_FIND_ROOT_PATH is used then CMAKE_FIND_ROOT_PATH will not be
      used.  If ONLY_CMAKE_FIND_ROOT_PATH is used then only the re-rooted
      directories will be searched.
 
      The default search order is designed to be most-specific to
      least-specific for common use cases.  Projects may override the order
      by simply calling the command multiple times and using the NO_*
      options:
 
          find_path(<VAR> NAMES name PATHS paths... NO_DEFAULT_PATH)
          find_path(<VAR> NAMES name)
 
      Once one of the calls succeeds the result variable will be set and
      stored in the cache so that no call will search again.
 
      When searching for frameworks, if the file is specified as A/b.h, then
      the framework search will look for A.framework/Headers/b.h.  If that
      is found the path will be set to the path to the framework.  CMake
      will convert this to the correct -F option to include the file.
 
  find_program
      Find an executable program.
 
          find_program(<VAR> name1 [path1 path2 ...])
 
      This is the short-hand signature for the command that is sufficient in
      many cases.  It is the same as find_program(<VAR> name1 [PATHS path1
      path2 ...])
 
          find_program(
                    <VAR>
                    name | NAMES name1 [name2 ...]
                    [HINTS path1 [path2 ... ENV var]]
                    [PATHS path1 [path2 ... ENV var]]
                    [PATH_SUFFIXES suffix1 [suffix2 ...]]
                    [DOC "cache documentation string"]
                    [NO_DEFAULT_PATH]
                    [NO_CMAKE_ENVIRONMENT_PATH]
                    [NO_CMAKE_PATH]
                    [NO_SYSTEM_ENVIRONMENT_PATH]
                    [NO_CMAKE_SYSTEM_PATH]
                    [CMAKE_FIND_ROOT_PATH_BOTH |
                    ONLY_CMAKE_FIND_ROOT_PATH |
                    NO_CMAKE_FIND_ROOT_PATH]
                  )
 
      This command is used to find a program.  A cache entry named by <VAR>
      is created to store the result of this command.  If the program is
      found the result is stored in the variable and the search will not be
      repeated unless the variable is cleared.  If nothing is found, the
      result will be <VAR>-NOTFOUND, and the search will be attempted again
      the next time find_program is invoked with the same variable.  The
      name of the program that is searched for is specified by the names
      listed after the NAMES argument.  Additional search locations can be
      specified after the PATHS argument.  If ENV var is found in the HINTS
      or PATHS section the environment variable var will be read and
      converted from a system environment variable to a cmake style list of
      paths.  For example ENV PATH would be a way to list the system path
      variable.  The argument after DOC will be used for the documentation
      string in the cache.  PATH_SUFFIXES specifies additional
      subdirectories to check below each search path.
 
      If NO_DEFAULT_PATH is specified, then no additional paths are added to
      the search.  If NO_DEFAULT_PATH is not specified, the search process
      is as follows:
 
      1.  Search paths specified in cmake-specific cache variables.  These
      are intended to be used on the command line with a -DVAR=value.  This
      can be skipped if NO_CMAKE_PATH is passed.
 
          <prefix>/[s]bin for each <prefix> in CMAKE_PREFIX_PATH
          CMAKE_PROGRAM_PATH
          CMAKE_APPBUNDLE_PATH
 
      2.  Search paths specified in cmake-specific environment variables.
      These are intended to be set in the user's shell configuration.  This
      can be skipped if NO_CMAKE_ENVIRONMENT_PATH is passed.
 
          <prefix>/[s]bin for each <prefix> in CMAKE_PREFIX_PATH
          CMAKE_PROGRAM_PATH
          CMAKE_APPBUNDLE_PATH
 
      3.  Search the paths specified by the HINTS option.  These should be
      paths computed by system introspection, such as a hint provided by the
      location of another item already found.  Hard-coded guesses should be
      specified with the PATHS option.
 
      4.  Search the standard system environment variables.  This can be
      skipped if NO_SYSTEM_ENVIRONMENT_PATH is an argument.
 
          PATH
         
 
      5.  Search cmake variables defined in the Platform files for the
      current system.  This can be skipped if NO_CMAKE_SYSTEM_PATH is
      passed.
 
          <prefix>/[s]bin for each <prefix> in CMAKE_SYSTEM_PREFIX_PATH
          CMAKE_SYSTEM_PROGRAM_PATH
          CMAKE_SYSTEM_APPBUNDLE_PATH
 
      6.  Search the paths specified by the PATHS option or in the
      short-hand version of the command.  These are typically hard-coded
      guesses.
 
      On Darwin or systems supporting OS X Frameworks, the cmake variable
      CMAKE_FIND_FRAMEWORK can be set to empty or one of the following:
 
          "FIRST"  - Try to find frameworks before standard
                    libraries or headers. This is the default on Darwin.
          "LAST"  - Try to find frameworks after standard
                    libraries or headers.
          "ONLY"  - Only try to find frameworks.
          "NEVER" - Never try to find frameworks.
 
      On Darwin or systems supporting OS X Application Bundles, the cmake
      variable CMAKE_FIND_APPBUNDLE can be set to empty or one of the
      following:
 
          "FIRST"  - Try to find application bundles before standard
                    programs. This is the default on Darwin.
          "LAST"  - Try to find application bundles after standard
                    programs.
          "ONLY"  - Only try to find application bundles.
          "NEVER" - Never try to find application bundles.
 
      The CMake variable CMAKE_FIND_ROOT_PATH specifies one or more
      directories to be prepended to all other search directories.  This
      effectively "re-roots" the entire search under given locations.  By
      default it is empty.  It is especially useful when cross-compiling to
      point to the root directory of the target environment and CMake will
      search there too.  By default at first the directories listed in
      CMAKE_FIND_ROOT_PATH and then the non-rooted directories will be
      searched.  The default behavior can be adjusted by setting
      CMAKE_FIND_ROOT_PATH_MODE_PROGRAM.  This behavior can be manually
      overridden on a per-call basis.  By using CMAKE_FIND_ROOT_PATH_BOTH
      the search order will be as described above.  If
      NO_CMAKE_FIND_ROOT_PATH is used then CMAKE_FIND_ROOT_PATH will not be
      used.  If ONLY_CMAKE_FIND_ROOT_PATH is used then only the re-rooted
      directories will be searched.
 
      The default search order is designed to be most-specific to
      least-specific for common use cases.  Projects may override the order
      by simply calling the command multiple times and using the NO_*
      options:
 
          find_program(<VAR> NAMES name PATHS paths... NO_DEFAULT_PATH)
          find_program(<VAR> NAMES name)
 
      Once one of the calls succeeds the result variable will be set and
      stored in the cache so that no call will search again.
 
  foreach
      Evaluate a group of commands for each value in a list.
 
        foreach(loop_var arg1 arg2 ...)
          COMMAND1(ARGS ...)
          COMMAND2(ARGS ...)
          ...
        endforeach(loop_var)
 
      All commands between foreach and the matching endforeach are recorded
      without being invoked.  Once the endforeach is evaluated, the recorded
      list of commands is invoked once for each argument listed in the
      original foreach command.  Before each iteration of the loop
      "${loop_var}" will be set as a variable with the current value in the
      list.
 
        foreach(loop_var RANGE total)
        foreach(loop_var RANGE start stop [step])
 
      Foreach can also iterate over a generated range of numbers.  There are
      three types of this iteration:
 
      * When specifying single number, the range will have elements 0 to
      "total".
 
      * When specifying two numbers, the range will have elements from the
      first number to the second number.
 
      * The third optional number is the increment used to iterate from the
      first number to the second number.
 
        foreach(loop_var IN [LISTS [list1 [...]]]
                            [ITEMS [item1 [...]]])
 
      Iterates over a precise list of items.  The LISTS option names
      list-valued variables to be traversed, including empty elements (an
      empty string is a zero-length list).  The ITEMS option ends argument
      parsing and includes all arguments following it in the iteration.
 
  function
      Start recording a function for later invocation as a command.
 
        function(<name> [arg1 [arg2 [arg3 ...]]])
          COMMAND1(ARGS ...)
          COMMAND2(ARGS ...)
          ...
        endfunction(<name>)
 
      Define a function named <name> that takes arguments named arg1 arg2
      arg3 (...).  Commands listed after function, but before the matching
      endfunction, are not invoked until the function is invoked.  When it
      is invoked, the commands recorded in the function are first modified
      by replacing formal parameters (${arg1}) with the arguments passed,
      and then invoked as normal commands.  In addition to referencing the
      formal parameters you can reference the variable ARGC which will be
      set to the number of arguments passed into the function as well as
      ARGV0 ARGV1 ARGV2 ...  which will have the actual values of the
      arguments passed in.  This facilitates creating functions with
      optional arguments.  Additionally ARGV holds the list of all arguments
      given to the function and ARGN holds the list of arguments past the
      last expected argument.
 
      A function opens a new scope: see set(var PARENT_SCOPE) for details.
 
      See the cmake_policy() command documentation for the behavior of
      policies inside functions.
 
  get_cmake_property
      Get a property of the CMake instance.
 
        get_cmake_property(VAR property)
 
      Get a property from the CMake instance.  The value of the property is
      stored in the variable VAR.  If the property is not found, VAR will be
      set to "NOTFOUND".  Some supported properties include: VARIABLES,
      CACHE_VARIABLES, COMMANDS, MACROS, and COMPONENTS.
 
      See also the more general get_property() command.
 
  get_directory_property
      Get a property of DIRECTORY scope.
 
        get_directory_property(<variable> [DIRECTORY <dir>] <prop-name>)
 
      Store a property of directory scope in the named variable.  If the
      property is not defined the empty-string is returned.  The DIRECTORY
      argument specifies another directory from which to retrieve the
      property value.  The specified directory must have already been
      traversed by CMake.
 
        get_directory_property(<variable> [DIRECTORY <dir>]
                                DEFINITION <var-name>)
 
      Get a variable definition from a directory.  This form is useful to
      get a variable definition from another directory.
 
      See also the more general get_property() command.
 
  get_filename_component
      Get a specific component of a full filename.
 
        get_filename_component(<VAR> <FileName> <COMP> [CACHE])
 
      Set <VAR> to a component of <FileName>, where <COMP> is one of:
 
        PATH      = Directory without file name
        NAME      = File name without directory
        EXT      = File name longest extension (.b.c from d/a.b.c)
        NAME_WE  = File name without directory or longest extension
        ABSOLUTE  = Full path to file
        REALPATH  = Full path to existing file with symlinks resolved
 
      Paths are returned with forward slashes and have no trailing slahes.
      The longest file extension is always considered.  If the optional
      CACHE argument is specified, the result variable is added to the
      cache.
 
        get_filename_component(<VAR> FileName
                                PROGRAM [PROGRAM_ARGS <ARG_VAR>]
                                [CACHE])
 
      The program in FileName will be found in the system search path or
      left as a full path.  If PROGRAM_ARGS is present with PROGRAM, then
      any command-line arguments present in the FileName string are split
      from the program name and stored in <ARG_VAR>.  This is used to
      separate a program name from its arguments in a command line string.
 
  get_property
      Get a property.
 
        get_property(<variable>
                      <GLOBAL            |
                      DIRECTORY [dir]    |
                      TARGET    <target> |
                      SOURCE    <source> |
                      TEST      <test>  |
                      CACHE    <entry>  |
                      VARIABLE>
                      PROPERTY <name>
                      [SET | DEFINED | BRIEF_DOCS | FULL_DOCS])
 
      Get one property from one object in a scope.  The first argument
      specifies the variable in which to store the result.  The second
      argument determines the scope from which to get the property.  It must
      be one of the following:
 
      GLOBAL scope is unique and does not accept a name.
 
      DIRECTORY scope defaults to the current directory but another
      directory (already processed by CMake) may be named by full or
      relative path.
 
      TARGET scope must name one existing target.
 
      SOURCE scope must name one source file.
 
      TEST scope must name one existing test.
 
      CACHE scope must name one cache entry.
 
      VARIABLE scope is unique and does not accept a name.
 
      The required PROPERTY option is immediately followed by the name of
      the property to get.  If the property is not set an empty value is
      returned.  If the SET option is given the variable is set to a boolean
      value indicating whether the property has been set.  If the DEFINED
      option is given the variable is set to a boolean value indicating
      whether the property has been defined such as with define_property.
      If BRIEF_DOCS or FULL_DOCS is given then the variable is set to a
      string containing documentation for the requested property.  If
      documentation is requested for a property that has not been defined
      NOTFOUND is returned.
 
  if
      Conditionally execute a group of commands.
 
        if(expression)
          # then section.
          COMMAND1(ARGS ...)
          COMMAND2(ARGS ...)
          ...
        elseif(expression2)
          # elseif section.
          COMMAND1(ARGS ...)
          COMMAND2(ARGS ...)
          ...
        else(expression)
          # else section.
          COMMAND1(ARGS ...)
          COMMAND2(ARGS ...)
          ...
        endif(expression)
 
      Evaluates the given expression.  If the result is true, the commands
      in the THEN section are invoked.  Otherwise, the commands in the else
      section are invoked.  The elseif and else sections are optional.  You
      may have multiple elseif clauses.  Note that the expression in the
      else and endif clause is optional.  Long expressions can be used and
      there is a traditional order of precedence.  Parenthetical expressions
      are evaluated first followed by unary operators such as EXISTS,
      COMMAND, and DEFINED.  Then any EQUAL, LESS, GREATER, STRLESS,
      STRGREATER, STREQUAL, MATCHES will be evaluated.  Then NOT operators
      and finally AND, OR operators will be evaluated.  Possible expressions
      are:
 
        if(<constant>)
 
      True if the constant is 1, ON, YES, TRUE, Y, or a non-zero number.
      False if the constant is 0, OFF, NO, FALSE, N, IGNORE, NOTFOUND, '',
      or ends in the suffix '-NOTFOUND'.  Named boolean constants are
      case-insensitive.  If the argument is not one of these constants, it
      is treated as a variable:
 
        if(<variable>)
 
      True if the variable is defined to a value that is not a false
      constant.  False otherwise.  (Note macro arguments are not variables.)
 
        if(NOT <expression>)
 
      True if the expression is not true.
 
        if(<expr1> AND <expr2>)
 
      True if both expressions would be considered true individually.
 
        if(<expr1> OR <expr2>)
 
      True if either expression would be considered true individually.
 
        if(COMMAND command-name)
 
      True if the given name is a command, macro or function that can be
      invoked.
 
        if(POLICY policy-id)
 
      True if the given name is an existing policy (of the form CMP<NNNN>).
 
        if(TARGET target-name)
 
      True if the given name is an existing target, built or imported.
 
        if(EXISTS file-name)
        if(EXISTS directory-name)
 
      True if the named file or directory exists.  Behavior is well-defined
      only for full paths.
 
        if(file1 IS_NEWER_THAN file2)
 
      True if file1 is newer than file2 or if one of the two files doesn't
      exist.  Behavior is well-defined only for full paths.  If the file
      time stamps are exactly the same, an IS_NEWER_THAN comparison returns
      true, so that any dependent build operations will occur in the event
      of a tie.  This includes the case of passing the same file name for
      both file1 and file2.
 
        if(IS_DIRECTORY directory-name)
 
      True if the given name is a directory.  Behavior is well-defined only
      for full paths.
 
        if(IS_SYMLINK file-name)
 
      True if the given name is a symbolic link.  Behavior is well-defined
      only for full paths.
 
        if(IS_ABSOLUTE path)
 
      True if the given path is an absolute path.
 
        if(<variable|string> MATCHES regex)
 
      True if the given string or variable's value matches the given regular
      expression.
 
        if(<variable|string> LESS <variable|string>)
        if(<variable|string> GREATER <variable|string>)
        if(<variable|string> EQUAL <variable|string>)
 
      True if the given string or variable's value is a valid number and the
      inequality or equality is true.
 
        if(<variable|string> STRLESS <variable|string>)
        if(<variable|string> STRGREATER <variable|string>)
        if(<variable|string> STREQUAL <variable|string>)
 
      True if the given string or variable's value is lexicographically less
      (or greater, or equal) than the string or variable on the right.
 
        if(<variable|string> VERSION_LESS <variable|string>)
        if(<variable|string> VERSION_EQUAL <variable|string>)
        if(<variable|string> VERSION_GREATER <variable|string>)
 
      Component-wise integer version number comparison (version format is
      major[.minor[.patch[.tweak]]]).
 
        if(DEFINED <variable>)
 
      True if the given variable is defined.  It does not matter if the
      variable is true or false just if it has been set.
 
        if((expression) AND (expression OR (expression)))
 
      The expressions inside the parenthesis are evaluated first and then
      the remaining expression is evaluated as in the previous examples.
      Where there are nested parenthesis the innermost are evaluated as part
      of evaluating the expression that contains them.
 
      The if command was written very early in CMake's history, predating
      the ${} variable evaluation syntax, and for convenience evaluates
      variables named by its arguments as shown in the above signatures.
      Note that normal variable evaluation with ${} applies before the if
      command even receives the arguments.  Therefore code like
 
        set(var1 OFF)
        set(var2 "var1")
        if(${var2})
 
      appears to the if command as
 
        if(var1)
 
      and is evaluated according to the if(<variable>) case documented
      above.  The result is OFF which is false.  However, if we remove the
      ${} from the example then the command sees
 
        if(var2)
 
      which is true because var2 is defined to "var1" which is not a false
      constant.
 
      Automatic evaluation applies in the other cases whenever the
      above-documented signature accepts <variable|string>:
 
      1) The left hand argument to MATCHES is first checked to see if it is
      a defined variable, if so the variable's value is used, otherwise the
      original value is used.
 
      2) If the left hand argument to MATCHES is missing it returns false
      without error
 
      3) Both left and right hand arguments to LESS GREATER EQUAL are
      independently tested to see if they are defined variables, if so their
      defined values are used otherwise the original value is used.
 
      4) Both left and right hand arguments to STRLESS STREQUAL STRGREATER
      are independently tested to see if they are defined variables, if so
      their defined values are used otherwise the original value is used.
 
      5) Both left and right hand argumemnts to VERSION_LESS VERSION_EQUAL
      VERSION_GREATER are independently tested to see if they are defined
      variables, if so their defined values are used otherwise the original
      value is used.
 
      6) The right hand argument to NOT is tested to see if it is a boolean
      constant, if so the value is used, otherwise it is assumed to be a
      variable and it is dereferenced.
 
      7) The left and right hand arguments to AND OR are independently
      tested to see if they are boolean constants, if so they are used as
      such, otherwise they are assumed to be variables and are dereferenced.
 
 
 
  include
      Read CMake listfile code from the given file.
 
        include(<file|module> [OPTIONAL] [RESULT_VARIABLE <VAR>]
                              [NO_POLICY_SCOPE])
 
      Reads CMake listfile code from the given file.  Commands in the file
      are processed immediately as if they were written in place of the
      include command.  If OPTIONAL is present, then no error is raised if
      the file does not exist.  If RESULT_VARIABLE is given the variable
      will be set to the full filename which has been included or NOTFOUND
      if it failed.
 
      If a module is specified instead of a file, the file with name
      <modulename>.cmake is searched first in CMAKE_MODULE_PATH, then in the
      CMake module directory.  There is one exception to this: if the file
      which calls include() is located itself in the CMake module directory,
      then first the CMake module directory is searched and
      CMAKE_MODULE_PATH afterwards.  See also policy CMP0017.
 
      See the cmake_policy() command documentation for discussion of the
      NO_POLICY_SCOPE option.
 
  list
      List operations.
 
        list(LENGTH <list> <output variable>)
        list(GET <list> <element index> [<element index> ...]
              <output variable>)
        list(APPEND <list> <element> [<element> ...])
        list(FIND <list> <value> <output variable>)
        list(INSERT <list> <element_index> <element> [<element> ...])
        list(REMOVE_ITEM <list> <value> [<value> ...])
        list(REMOVE_AT <list> <index> [<index> ...])
        list(REMOVE_DUPLICATES <list>)
        list(REVERSE <list>)
        list(SORT <list>)
 
      LENGTH will return a given list's length.
 
      GET will return list of elements specified by indices from the list.
 
      APPEND will append elements to the list.
 
      FIND will return the index of the element specified in the list or -1
      if it wasn't found.
 
      INSERT will insert elements to the list to the specified location.
 
      REMOVE_AT and REMOVE_ITEM will remove items from the list.  The
      difference is that REMOVE_ITEM will remove the given items, while
      REMOVE_AT will remove the items at the given indices.
 
      REMOVE_DUPLICATES will remove duplicated items in the list.
 
      REVERSE reverses the contents of the list in-place.
 
      SORT sorts the list in-place alphabetically.
 
      The list subcommands APPEND, INSERT, REMOVE_AT, REMOVE_ITEM,
      REMOVE_DUPLICATES, REVERSE and SORT may create new values for the list
      within the current CMake variable scope.  Similar to the SET command,
      the LIST command creates new variable values in the current scope,
      even if the list itself is actually defined in a parent scope.  To
      propagate the results of these operations upwards, use SET with
      PARENT_SCOPE, SET with CACHE INTERNAL, or some other means of value
      propagation.
 
      NOTES: A list in cmake is a ; separated group of strings.  To create a
      list the set command can be used.  For example, set(var a b c d e)
      creates a list with a;b;c;d;e, and set(var "a b c d e") creates a
      string or a list with one item in it.
 
      When specifying index values, if <element index> is 0 or greater, it
      is indexed from the beginning of the list, with 0 representing the
      first list element.  If <element index> is -1 or lesser, it is indexed
      from the end of the list, with -1 representing the last list element.
      Be careful when counting with negative indices: they do not start from
      0.  -0 is equivalent to 0, the first list element.
 
 
  macro
      Start recording a macro for later invocation as a command.
 
        macro(<name> [arg1 [arg2 [arg3 ...]]])
          COMMAND1(ARGS ...)
          COMMAND2(ARGS ...)
          ...
        endmacro(<name>)
 
      Define a macro named <name> that takes arguments named arg1 arg2 arg3
      (...).  Commands listed after macro, but before the matching endmacro,
      are not invoked until the macro is invoked.  When it is invoked, the
      commands recorded in the macro are first modified by replacing formal
      parameters (${arg1}) with the arguments passed, and then invoked as
      normal commands.  In addition to referencing the formal parameters you
      can reference the values ${ARGC} which will be set to the number of
      arguments passed into the function as well as ${ARGV0} ${ARGV1}
      ${ARGV2} ...  which will have the actual values of the arguments
      passed in.  This facilitates creating macros with optional arguments.
      Additionally ${ARGV} holds the list of all arguments given to the
      macro and ${ARGN} holds the list of arguments past the last expected
      argument.  Note that the parameters to a macro and values such as ARGN
      are not variables in the usual CMake sense.  They are string
      replacements much like the C preprocessor would do with a macro.  If
      you want true CMake variables and/or better CMake scope control you
      should look at the function command.
 
      See the cmake_policy() command documentation for the behavior of
      policies inside macros.
 
  make_directory
      Deprecated.  Use the file(MAKE_DIRECTORY ) command instead.
 
        make_directory(directory)
 
      Creates the specified directory.  Full paths should be given.  Any
      parent directories that do not exist will also be created.  Use with
      care.
 
  mark_as_advanced
      Mark cmake cached variables as advanced.
 
        mark_as_advanced([CLEAR|FORCE] VAR VAR2 VAR...)
 
      Mark the named cached variables as advanced.  An advanced variable
      will not be displayed in any of the cmake GUIs unless the show
      advanced option is on.  If CLEAR is the first argument advanced
      variables are changed back to unadvanced.  If FORCE is the first
      argument, then the variable is made advanced.  If neither FORCE nor
      CLEAR is specified, new values will be marked as advanced, but if the
      variable already has an advanced/non-advanced state, it will not be
      changed.
 
      It does nothing in script mode.
 
  math
      Mathematical expressions.
 
        math(EXPR <output variable> <math expression>)
 
      EXPR evaluates mathematical expression and returns result in the
      output variable.  Example mathematical expression is '5 * ( 10 + 13
      )'.  Supported operators are + - * / % | & ^ ~ << >> * / %.  They have
      the same meaning as they do in C code.
 
  message
      Display a message to the user.
 
        message([STATUS|WARNING|AUTHOR_WARNING|FATAL_ERROR|SEND_ERROR]
                "message to display" ...)
 
      The optional keyword determines the type of message:
 
        (none)        = Important information
        STATUS        = Incidental information
        WARNING        = CMake Warning, continue processing
        AUTHOR_WARNING = CMake Warning (dev), continue processing
        SEND_ERROR    = CMake Error, continue processing,
                                      but skip generation
        FATAL_ERROR    = CMake Error, stop processing and generation
 
      The CMake command-line tool displays STATUS messages on stdout and all
      other message types on stderr.  The CMake GUI displays all messages in
      its log area.  The interactive dialogs (ccmake and CMakeSetup) show
      STATUS messages one at a time on a status line and other messages in
      interactive pop-up boxes.
 
      CMake Warning and Error message text displays using a simple markup
      language.  Non-indented text is formatted in line-wrapped paragraphs
      delimited by newlines.  Indented text is considered pre-formatted.
 
  option
      Provides an option that the user can optionally select.
 
        option(<option_variable> "help string describing option"
                [initial value])
 
      Provide an option for the user to select as ON or OFF.  If no initial
      value is provided, OFF is used.
 
      If you have options that depend on the values of other options, see
      the module help for CMakeDependentOption.
 
  remove
      Deprecated.  Use the list(REMOVE_ITEM ) command instead.
 
        remove(VAR VALUE VALUE ...)
 
      Removes VALUE from the variable VAR.  This is typically used to remove
      entries from a vector (e.g.  semicolon separated list).  VALUE is
      expanded.
 
  return
      Return from a file, directory or function.
 
        return()
 
      Returns from a file, directory or function.  When this command is
      encountered in an included file (via include() or find_package()), it
      causes processing of the current file to stop and control is returned
      to the including file.  If it is encountered in a file which is not
      included by another file, e.g.  a CMakeLists.txt, control is returned
      to the parent directory if there is one.  If return is called in a
      function, control is returned to the caller of the function.  Note
      that a macro is not a function and does not handle return like a
      function does.
 
  separate_arguments
      Parse space-separated arguments into a semicolon-separated list.
 
        separate_arguments(<var> <UNIX|WINDOWS>_COMMAND "<args>")
 
      Parses a unix- or windows-style command-line string "<args>" and
      stores a semicolon-separated list of the arguments in <var>.  The
      entire command line must be given in one "<args>" argument.
 
      The UNIX_COMMAND mode separates arguments by unquoted whitespace.  It
      recognizes both single-quote and double-quote pairs.  A backslash
      escapes the next literal character (\" is "); there are no special
      escapes (\n is just n).
 
      The WINDOWS_COMMAND mode parses a windows command-line using the same
      syntax the runtime library uses to construct argv at startup.  It
      separates arguments by whitespace that is not double-quoted.
      Backslashes are literal unless they precede double-quotes.  See the
      MSDN article "Parsing C Command-Line Arguments" for details.
 
        separate_arguments(VARIABLE)
 
      Convert the value of VARIABLE to a semi-colon separated list.  All
      spaces are replaced with ';'.  This helps with generating command
      lines.
 
  set
      Set a CMake, cache or environment variable to a given value.
 
        set(<variable> <value>
            [[CACHE <type> <docstring> [FORCE]] | PARENT_SCOPE])
 
      Within CMake sets <variable> to the value <value>.  <value> is
      expanded before <variable> is set to it.  Normally, set will set a
      regular CMake variable.  If CACHE is present, then the <variable> is
      put in the cache instead, unless it is already in the cache.  See
      section 'Variable types in CMake' below for details of regular and
      cache variables and their interactions.  If CACHE is used, <type> and
      <docstring> are required.  <type> is used by the CMake GUI to choose a
      widget with which the user sets a value.  The value for <type> may be
      one of
 
        FILEPATH = File chooser dialog.
        PATH    = Directory chooser dialog.
        STRING  = Arbitrary string.
        BOOL    = Boolean ON/OFF checkbox.
        INTERNAL = No GUI entry (used for persistent variables).
 
      If <type> is INTERNAL, the cache variable is marked as internal, and
      will not be shown to the user in tools like cmake-gui.  This is
      intended for values that should be persisted in the cache, but which
      users should not normally change.  INTERNAL implies FORCE.
 
      Normally, set(...CACHE...) creates cache variables, but does not
      modify them.  If FORCE is specified, the value of the cache variable
      is set, even if the variable is already in the cache.  This should
      normally be avoided, as it will remove any changes to the cache
      variable's value by the user.
 
      If PARENT_SCOPE is present, the variable will be set in the scope
      above the current scope.  Each new directory or function creates a new
      scope.  This command will set the value of a variable into the parent
      directory or calling function (whichever is applicable to the case at
      hand).  PARENT_SCOPE cannot be combined with CACHE.
 
      If <value> is not specified then the variable is removed instead of
      set.  See also: the unset() command.
 
        set(<variable> <value1> ... <valueN>)
 
      In this case <variable> is set to a semicolon separated list of
      values.
 
      <variable> can be an environment variable such as:
 
        set( ENV{PATH} /home/martink )
 
      in which case the environment variable will be set.
 
      *** Variable types in CMake ***
 
      In CMake there are two types of variables: normal variables and cache
      variables.  Normal variables are meant for the internal use of the
      script (just like variables in most programming languages); they are
      not persisted across CMake runs.  Cache variables (unless set with
      INTERNAL) are mostly intended for configuration settings where the
      first CMake run determines a suitable default value, which the user
      can then override, by editing the cache with tools such as ccmake or
      cmake-gui.  Cache variables are stored in the CMake cache file, and
      are persisted across CMake runs.
 
      Both types can exist at the same time with the same name but different
      values.  When ${FOO} is evaluated, CMake first looks for a normal
      variable 'FOO' in scope and uses it if set.  If and only if no normal
      variable exists then it falls back to the cache variable 'FOO'.
 
      Some examples:
 
      The code 'set(FOO "x")' sets the normal variable 'FOO'.  It does not
      touch the cache, but it will hide any existing cache value 'FOO'.
 
      The code 'set(FOO "x" CACHE ...)' checks for 'FOO' in the cache,
      ignoring any normal variable of the same name.  If 'FOO' is in the
      cache then nothing happens to either the normal variable or the cache
      variable.  If 'FOO' is not in the cache, then it is added to the
      cache.
 
      Finally, whenever a cache variable is added or modified by a command,
      CMake also *removes* the normal variable of the same name from the
      current scope so that an immediately following evaluation of it will
      expose the newly cached value.
 
      Normally projects should avoid using normal and cache variables of the
      same name, as this interaction can be hard to follow.  However, in
      some situations it can be useful.  One example (used by some
      projects):
 
      A project has a subproject in its source tree.  The child project has
      its own CMakeLists.txt, which is included from the parent
      CMakeLists.txt using add_subdirectory().  Now, if the parent and the
      child project provide the same option (for example a compiler option),
      the parent gets the first chance to add a user-editable option to the
      cache.  Normally, the child would then use the same value that the
      parent uses.  However, it may be necessary to hard-code the value for
      the child project's option while still allowing the user to edit the
      value used by the parent project.  The parent project can achieve this
      simply by setting a normal variable with the same name as the option
      in a scope sufficient to hide the option's cache variable from the
      child completely.  The parent has already set the cache variable, so
      the child's set(...CACHE...) will do nothing, and evaluating the
      option variable will use the value from the normal variable, which
      hides the cache variable.
 
  set_directory_properties
      Set a property of the directory.
 
        set_directory_properties(PROPERTIES prop1 value1 prop2 value2)
 
      Set a property for the current directory and subdirectories.  If the
      property is not found, CMake will report an error.  The properties
      include: INCLUDE_DIRECTORIES, LINK_DIRECTORIES,
      INCLUDE_REGULAR_EXPRESSION, and ADDITIONAL_MAKE_CLEAN_FILES.
      ADDITIONAL_MAKE_CLEAN_FILES is a list of files that will be cleaned as
      a part of "make clean" stage.
 
  set_property
      Set a named property in a given scope.
 
        set_property(<GLOBAL                            |
                      DIRECTORY [dir]                  |
                      TARGET    [target1 [target2 ...]] |
                      SOURCE    [src1 [src2 ...]]      |
                      TEST      [test1 [test2 ...]]    |
                      CACHE    [entry1 [entry2 ...]]>
                      [APPEND] [APPEND_STRING]
                      PROPERTY <name> [value1 [value2 ...]])
 
      Set one property on zero or more objects of a scope.  The first
      argument determines the scope in which the property is set.  It must
      be one of the following:
 
      GLOBAL scope is unique and does not accept a name.
 
      DIRECTORY scope defaults to the current directory but another
      directory (already processed by CMake) may be named by full or
      relative path.
 
      TARGET scope may name zero or more existing targets.
 
      SOURCE scope may name zero or more source files.  Note that source
      file properties are visible only to targets added in the same
      directory (CMakeLists.txt).
 
      TEST scope may name zero or more existing tests.
 
      CACHE scope must name zero or more cache existing entries.
 
      The required PROPERTY option is immediately followed by the name of
      the property to set.  Remaining arguments are used to compose the
      property value in the form of a semicolon-separated list.  If the
      APPEND option is given the list is appended to any existing property
      value.If the APPEND_STRING option is given the string is append to any
      existing property value as string, i.e.  it results in a longer string
      and not a list of strings.
 
  site_name
      Set the given variable to the name of the computer.
 
        site_name(variable)
 
 
  string
      String operations.
 
        string(REGEX MATCH <regular_expression>
                <output variable> <input> [<input>...])
        string(REGEX MATCHALL <regular_expression>
                <output variable> <input> [<input>...])
        string(REGEX REPLACE <regular_expression>
                <replace_expression> <output variable>
                <input> [<input>...])
        string(REPLACE <match_string>
                <replace_string> <output variable>
                <input> [<input>...])
        string(<MD5|SHA1|SHA224|SHA256|SHA384|SHA512>
                <output variable> <input>)
        string(COMPARE EQUAL <string1> <string2> <output variable>)
        string(COMPARE NOTEQUAL <string1> <string2> <output variable>)
        string(COMPARE LESS <string1> <string2> <output variable>)
        string(COMPARE GREATER <string1> <string2> <output variable>)
        string(ASCII <number> [<number> ...] <output variable>)
        string(CONFIGURE <string1> <output variable>
                [@ONLY] [ESCAPE_QUOTES])
        string(TOUPPER <string1> <output variable>)
        string(TOLOWER <string1> <output variable>)
        string(LENGTH <string> <output variable>)
        string(SUBSTRING <string> <begin> <length> <output variable>)
        string(STRIP <string> <output variable>)
        string(RANDOM [LENGTH <length>] [ALPHABET <alphabet>]
                [RANDOM_SEED <seed>] <output variable>)
        string(FIND <string> <substring> <output variable> [REVERSE])
        string(TIMESTAMP <output variable> [<format string>] [UTC])
 
      REGEX MATCH will match the regular expression once and store the match
      in the output variable.
 
      REGEX MATCHALL will match the regular expression as many times as
      possible and store the matches in the output variable as a list.
 
      REGEX REPLACE will match the regular expression as many times as
      possible and substitute the replacement expression for the match in
      the output.  The replace expression may refer to paren-delimited
      subexpressions of the match using \1, \2, ..., \9.  Note that two
      backslashes (\\1) are required in CMake code to get a backslash
      through argument parsing.
 
      REPLACE will replace all occurrences of match_string in the input with
      replace_string and store the result in the output.
 
      MD5, SHA1, SHA224, SHA256, SHA384, and SHA512 will compute a
      cryptographic hash of the input string.
 
      COMPARE EQUAL/NOTEQUAL/LESS/GREATER will compare the strings and store
      true or false in the output variable.
 
      ASCII will convert all numbers into corresponding ASCII characters.
 
      CONFIGURE will transform a string like CONFIGURE_FILE transforms a
      file.
 
      TOUPPER/TOLOWER will convert string to upper/lower characters.
 
      LENGTH will return a given string's length.
 
      SUBSTRING will return a substring of a given string.  If length is -1
      the remainder of the string starting at begin will be returned.
 
      STRIP will return a substring of a given string with leading and
      trailing spaces removed.
 
      RANDOM will return a random string of given length consisting of
      characters from the given alphabet.  Default length is 5 characters
      and default alphabet is all numbers and upper and lower case letters.
      If an integer RANDOM_SEED is given, its value will be used to seed the
      random number generator.
 
      FIND will return the position where the given substring was found in
      the supplied string.  If the REVERSE flag was used, the command will
      search for the position of the last occurrence of the specified
      substring.
 
      The following characters have special meaning in regular expressions:
 
          ^        Matches at beginning of input
          $        Matches at end of input
          .        Matches any single character
          [ ]      Matches any character(s) inside the brackets
          [^ ]      Matches any character(s) not inside the brackets
          -        Inside brackets, specifies an inclusive range between
                    characters on either side e.g. [a-f] is [abcdef]
                    To match a literal - using brackets, make it the first
                    or the last character e.g. [+*/-] matches basic
                    mathematical operators.
          *        Matches preceding pattern zero or more times
          +        Matches preceding pattern one or more times
          ?        Matches preceding pattern zero or once only
          |        Matches a pattern on either side of the |
          ()        Saves a matched subexpression, which can be referenced
                    in the REGEX REPLACE operation. Additionally it is saved
                    by all regular expression-related commands, including
                    e.g. if( MATCHES ), in the variables CMAKE_MATCH_(0..9).
 
      *, + and ? have higher precedence than concatenation.  | has lower
      precedence than concatenation.  This means that the regular expression
      "^ab+d$" matches "abbd" but not "ababd", and the regular expression
      "^(ab|cd)$" matches "ab" but not "abd".
 
      TIMESTAMP will write a string representation of the current date
      and/or time to the output variable.
 
      Should the command be unable to obtain a timestamp the output variable
      will be set to the empty string "".
 
      The optional UTC flag requests the current date/time representation to
      be in Coordinated Universal Time (UTC) rather than local time.
 
      The optional <format string> may contain the following format
      specifiers:
 
          %d        The day of the current month (01-31).
          %H        The hour on a 24-hour clock (00-23).
          %I        The hour on a 12-hour clock (01-12).
          %j        The day of the current year (001-366).
          %m        The month of the current year (01-12).
          %M        The minute of the current hour (00-59).
          %S        The second of the current minute.
                    60 represents a leap second. (00-60)
          %U        The week number of the current year (00-53).
          %w        The day of the current week. 0 is Sunday. (0-6)
          %y        The last two digits of the current year (00-99)
          %Y        The current year.
 
      Unknown format specifiers will be ignored and copied to the output
      as-is.
 
      If no explicit <format string> is given it will default to:
 
          %Y-%m-%dT%H:%M:%S    for local time.
          %Y-%m-%dT%H:%M:%SZ  for UTC.
 
  unset
      Unset a variable, cache variable, or environment variable.
 
        unset(<variable> [CACHE])
 
      Removes the specified variable causing it to become undefined.  If
      CACHE is present then the variable is removed from the cache instead
      of the current scope.
 
      <variable> can be an environment variable such as:
 
        unset(ENV{LD_LIBRARY_PATH})
 
      in which case the variable will be removed from the current
      environment.
 
  use_mangled_mesa
      Copy mesa headers for use in combination with system GL.
 
        use_mangled_mesa(PATH_TO_MESA OUTPUT_DIRECTORY)
 
      The path to mesa includes, should contain gl_mangle.h.  The mesa
      headers are copied to the specified output directory.  This allows
      mangled mesa headers to override other GL headers by being added to
      the include directory path earlier.
 
  variable_watch
      Watch the CMake variable for change.
 
        variable_watch(<variable name> [<command to execute>])
 
      If the specified variable changes, the message will be printed about
      the variable being changed.  If the command is specified, the command
      will be executed.  The command will receive the following arguments:
      COMMAND(<variable> <access> <value> <current list file> <stack>)
 
  while
      Evaluate a group of commands while a condition is true
 
        while(condition)
          COMMAND1(ARGS ...)
          COMMAND2(ARGS ...)
          ...
        endwhile(condition)
 
      All commands between while and the matching endwhile are recorded
      without being invoked.  Once the endwhile is evaluated, the recorded
      list of commands is invoked as long as the condition is true.  The
      condition is evaluated using the same logic as the if command.
 
  write_file
      Deprecated.  Use the file(WRITE ) command instead.
 
        write_file(filename "message to write"... [APPEND])
 
      The first argument is the file name, the rest of the arguments are
      messages to write.  If the argument APPEND is specified, then the
      message will be appended.
 
      NOTE 1: file(WRITE ...  and file(APPEND ...  do exactly the same as
      this one but add some more functionality.
 
      NOTE 2: When using write_file the produced file cannot be used as an
      input to CMake (CONFIGURE_FILE, source file ...) because it will lead
      to an infinite loop.  Use configure_file if you want to generate input
      files to CMake.
 
------------------------------------------------------------------------------
Properties
 
  CMake Properties - Properties supported by CMake, the Cross-Platform Makefile Generator.
 
This is the documentation for the properties supported by CMake.  Properties
can have different scopes.  They can either be assigned to a source file, a
directory, a target or globally to CMake.  By modifying the values of
properties the behaviour of the build system can be customized.
 
------------------------------------------------------------------------------
Compatibility Commands
 
  CMake Compatibility Listfile Commands - Obsolete commands supported by CMake for compatibility.
 
This is the documentation for now obsolete listfile commands from previous
CMake versions, which are still supported for compatibility reasons.  You
should instead use the newer, faster and shinier new commands.  ;-)
 
------------------------------------------------------------------------------
Standard CMake Modules
 
------------------------------------------------------------------------------
Copyright
 
Copyright 2000-2012 Kitware, Inc., Insight Software Consortium.  All rights
reserved.
 
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:
 
Redistributions of source code must retain the above copyright notice, this
list of conditions and the following disclaimer.
 
Redistributions in binary form must reproduce the above copyright notice,
this list of conditions and the following disclaimer in the documentation
and/or other materials provided with the distribution.
 
Neither the names of Kitware, Inc., the Insight Software Consortium, nor the
names of their contributors may be used to endorse or promote products
derived from this software without specific prior written permission.
 
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE
LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR
CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF
SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS
INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN
CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE)
ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
POSSIBILITY OF SUCH DAMAGE.
 
------------------------------------------------------------------------------
See Also
 
The following resources are available to get help using CMake:
 
  Home Page
      http://www.cmake.org
 
      The primary starting point for learning about CMake.
 
  Frequently Asked Questions
      http://www.cmake.org/Wiki/CMake_FAQ
 
      A Wiki is provided containing answers to frequently asked questions.
 
  Online Documentation
      http://www.cmake.org/HTML/Documentation.html
 
      Links to available documentation may be found on this web page.
 
  Mailing List
      http://www.cmake.org/HTML/MailingLists.html
 
      For help and discussion about using cmake, a mailing list is provided
      at cmake@cmake.org.  The list is member-post-only but one may sign up
      on the CMake web page.  Please first read the full documentation at
      http://www.cmake.org before posting questions to the list.
</pre>

Latest revision as of 13:03, 24 April 2018