CTest 2.8.0 Docs: Difference between revisions

From KitwarePublic
Jump to navigationJump to search
(New page: <pre> ctest version 2.8.0 ------------------------------------------------------------------------------ Name ctest - Testing driver provided by CMake. --------------------------------...)
 
(Replaced content with "See https://cmake.org/cmake/help/v2.8.0/ctest.html")
 
Line 1: Line 1:
<pre>
See https://cmake.org/cmake/help/v2.8.0/ctest.html
ctest version 2.8.0
------------------------------------------------------------------------------
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.
 
  -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.
 
  -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
      seperate process.  This is primarily useful in cases where the script
      may modify the environment and you do not want the modified enviroment
      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 cliping 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 to not 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-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 configuraiton
 
      A sample executable to use to determine the configuraiton 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.
 
  --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.
 
  --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.
 
------------------------------------------------------------------------------
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]]
                                [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]])
 
      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])
 
      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.  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])
 
      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.
 
      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 replaces and ${VAR} will be ignored.  This is useful for
      configuring scripts that use ${VAR}.  Any occurrences of #cmakedefine
      VAR will be replaced with either #define VAR or /* #undef VAR */
      depending on the setting of VAR in CMake.  Any occurrences of
      #cmakedefine01 VAR will be replaced with either #define VAR 1 or
      #define VAR 0 depending on whether VAR evaluates to TRUE or FALSE in
      CMake
 
  ctest_build
      Build the project.
 
        ctest_build([BUILD build_dir] [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
      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 seperate 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 )
        ctest_sleep( time1 duration time2 )
 
      With one argument it will sleep for a given number of seconds.  With
      three arguments it will wait for time2 - time1 - duration seconds.
 
  ctest_start
      Starts the testing for a given model
 
        ctest_start(Model [TRACK <track>] [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.
 
  ctest_submit
      Submit results to a dashboard server.
 
        ctest_submit([PARTS ...] [FILES ...] [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.
 
 
  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])
 
      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.
 
      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.
 
  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(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 [TIMEOUT timeout] [STATUS status] [LOG log])
 
      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.
 
      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.
 
      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.
 
      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.
 
      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 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 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 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 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 ...]
                    [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 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 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 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.
 
      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.
 
  find_package
      Load settings for an external project.
 
        find_package(<package> [version] [EXACT] [QUIET]
                      [[REQUIRED|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 documented by the package itself.  The QUIET option disables
      messages if the package cannot be found.  The REQUIRED option stops
      processing with an error message if the package cannot be found.  A
      package-specific list of components may be listed after the REQUIRED
      option or after the COMPONENTS option if no REQUIRED option is given.
      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] is given to a recursive invocation inside a find-module, the
      [version] and EXACT arguments are forwarded automatically from the
      outer call.  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 the command proceeds to Config mode.
 
      The complete Config mode command signature is:
 
        find_package(<package> [version] [EXACT] [QUIET]
                      [[REQUIRED|COMPONENTS] [components...]] [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]
                      [CMAKE_FIND_ROOT_PATH_BOTH |
                      ONLY_CMAKE_FIND_ROOT_PATH |
                      NO_CMAKE_FIND_ROOT_PATH])
 
      The NO_MODULE option may be used to skip Module mode explicitly.  It
      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.
 
      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.  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>/(share|lib)/cmake/<name>*/                    (U)
        <prefix>/(share|lib)/<name>*/                          (U)
        <prefix>/(share|lib)/<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).  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.
 
          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.  Paths are stored
      in the registry when CMake configures a project that invokes
      export(PACKAGE <name>).  See the export(PACKAGE) command documentation
      for more details.
 
      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 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.
 
      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 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 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 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 argument past the
      last expected argument.
 
      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, CMake will
      report an error.  Some supported properties include: VARIABLES,
      CACHE_VARIABLES, COMMANDS, MACROS, and COMPONENTS.
 
  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.
 
  get_filename_component
      Get a specific component of a full filename.
 
        get_filename_component(VarName FileName
                                PATH|ABSOLUTE|NAME|EXT|NAME_WE|REALPATH
                                [CACHE])
 
      Set VarName to be the path (PATH), file name (NAME), file extension
      (EXT), file name without extension (NAME_WE) of FileName, the full
      path (ABSOLUTE), or the full path with all symlinks resolved
      (REALPATH).  Note that the path is converted to Unix slashes format
      and has no trailing slashes.  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(VarName FileName
                                PROGRAM [PROGRAM_ARGS ArgVar]
                                [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 ArgVar.  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, "", or ends in
      the suffix '-NOTFOUND'.  Named boolean constants are case-insensitive.
 
        if(<variable>)
 
      True if the variable's value is not a false constant.
 
        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(IS_DIRECTORY directory-name)
 
      True if the given name is a directory.  Behavior is well-defined only
      for full paths.
 
        if(IS_ABSOLUTE path)
 
      True if the given path is an absolute path.
 
        if(variable MATCHES regex)
        if(string MATCHES regex)
 
      True if the given string or variable's value matches the given regular
      expression.
 
        if(variable LESS number)
        if(string LESS number)
        if(variable GREATER number)
        if(string GREATER number)
        if(variable EQUAL number)
        if(string EQUAL number)
 
      True if the given string or variable's value is a valid number and the
      inequality or equality is true.
 
        if(variable STRLESS string)
        if(string STRLESS string)
        if(variable STRGREATER string)
        if(string STRGREATER string)
        if(variable STREQUAL string)
        if(string STREQUAL 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(version1 VERSION_LESS version2)
        if(version1 VERSION_EQUAL version2)
        if(version1 VERSION_GREATER version2)
 
      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 statement was written fairly early in CMake's history and it
      has some convenience features that are worth covering.  The if
      statement reduces operations until there is a single remaining value,
      at that point if the case insensitive value is: ON, 1, YES, TRUE, Y it
      returns true, if it is OFF, 0, NO, FALSE, N, NOTFOUND, *-NOTFOUND,
      IGNORE it will return false.
 
      This is fairly reasonable.  The convenience feature that sometimes
      throws new authors is how CMake handles values that do not match the
      true or false list.  Those values are treated as variables and are
      dereferenced even though they do not have the required ${} syntax.
      This means that if you write
 
        if (boobah)
 
      CMake will treat it as if you wrote
 
        if (${boobah})
 
      likewise if you write
 
        if (fubar AND sol)
 
      CMake will conveniently treat it as
 
        if ("${fubar}" AND "${sol}")
 
      The later is really the correct way to write it, but the former will
      work as well.  Only some operations in the if statement have this
      special handling of arguments.  The specific details follow:
 
      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 in the CMAKE_MODULE_PATH.
 
      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.
 
      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 argument 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 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 return 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 but skip generation
        FATAL_ERROR    = CMake Error, stop all processing
 
      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.
 
  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 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.  If CACHE is present, then
      the <variable> is put in the cache.  <type> and <docstring> are then
      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, then the <value> is always written into the
      cache, replacing any values existing in the cache.  If it is not a
      cache variable, then this always writes into the current makefile.
      The FORCE option will overwrite the cache value removing any changes
      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).
 
      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.
 
  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]
                      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.
 
      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.
 
  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(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>)
 
      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.
 
      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.
 
      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.
 
      The following characters have special meaning in regular expressions:
 
          ^        Matches at beginning of a line
          $        Matches at end of a line
          .        Matches any single character
          [ ]      Matches any character(s) inside the brackets
          [^ ]      Matches any character(s) not inside the brackets
          -        Matches any character in range on either side of a dash
          *        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).
 
  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.
 
  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 2000-2009 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.
 
Summary of helpful links:
 
  Home: http://www.cmake.org
  Docs: http://www.cmake.org/HTML/Documentation.html
  Mail: http://www.cmake.org/HTML/MailingLists.html
  FAQ:  http://www.cmake.org/Wiki/CMake_FAQ
 
 
 
</pre>

Latest revision as of 13:02, 24 April 2018