CTestCoverageCollectGCOV: Add TARBALL_COMPRESSION option

For large projects, the default bzip2 compression of gcov data is
prohibitively expensively (could take several hours).  Introduce options
that allow the consumer to choose between file size and compression time.

Add a new optional argument `TARBALL_COMPRESSION` for the macro
`ctest_coverage_collect_gcov`. This option accepts the values `GZIP`,
`BZIP2`, `XZ`, `FROM_EXT`, or an expression that evaluates to `FALSE`.
The default value is `BZIP2` to preserve prior behavior.

Fixes: #20593

3 files changed, 191 insertions, 14 deletions

diff --git a/Help/release/dev/CTestCoverageCollectGCOV-compress-opts.rst b/Help/release/dev/CTestCoverageCollectGCOV-compress-opts.rst
new file mode 100644
index 0000000000..eb8532d44c
--- /dev/null
+++ b/Help/release/dev/CTestCoverageCollectGCOV-compress-opts.rst
@@ -0,0 +1,7 @@
+CTestCoverageCollectGCOV-compress-opts
+--------------------------------------
+
+* The :module:`CTestCoverageCollectGCOV` module
+  :command:`ctest_coverage_collect_gcov` function gained a
+  ``TARBALL_COMPRESSION`` option to control compression of the
+  tarball of collected results.
diff --git a/Modules/CTestCoverageCollectGCOV.cmake b/Modules/CTestCoverageCollectGCOV.cmake
index b498086c9c..b1268bebf2 100644
--- a/Modules/CTestCoverageCollectGCOV.cmake
+++ b/Modules/CTestCoverageCollectGCOV.cmake
@@ -37,6 +37,17 @@ After generating this tar file, it can be sent to CDash for display with the
     upload to CDash.  Relative paths will be interpreted with respect
     to the top-level build directory.
 
+  ``TARBALL_COMPRESSION <option>`` Specify a compression algorithm for the
+    ``TARBALL`` data file.  Using this option reduces the size of the data file
+    before it is submitted to CDash.  ``<option>`` must be one of ``GZIP``,
+    ``BZIP2``, ``XZ``, ``ZSTD``, ``FROM_EXT``, or an expression that CMake
+    evaluates as ``FALSE``. The default value is ``BZIP2``.
+
+    If ``FROM_EXT`` is specified, the resulting file will be compressed based on
+    the file extension of the ``<tarfile>`` (i.e. ``.tar.gz`` will use ``GZIP``
+    compression). File extensions that will produce compressed output include
+    ``.tar.gz``, ``.tgz``, ``.tar.bzip2``, ``.tbz``, ``.tar.xz``, and ``.txz``.
+
   ``SOURCE <source_dir>``
     Specify the top-level source directory for the build.
     Default is the value of :variable:`CTEST_SOURCE_DIRECTORY`.
@@ -68,7 +79,7 @@ After generating this tar file, it can be sent to CDash for display with the
 
 function(ctest_coverage_collect_gcov)
   set(options QUIET GLOB DELETE)
-  set(oneValueArgs TARBALL SOURCE BUILD GCOV_COMMAND)
+  set(oneValueArgs TARBALL SOURCE BUILD GCOV_COMMAND TARBALL_COMPRESSION)
   set(multiValueArgs GCOV_OPTIONS)
   cmake_parse_arguments(GCOV  "${options}" "${oneValueArgs}"
     "${multiValueArgs}" "" ${ARGN} )
@@ -91,6 +102,13 @@ function(ctest_coverage_collect_gcov)
   else()
     set(gcov_command "${GCOV_GCOV_COMMAND}")
   endif()
+  if(NOT DEFINED GCOV_TARBALL_COMPRESSION)
+    set(GCOV_TARBALL_COMPRESSION "BZIP2")
+  elseif( GCOV_TARBALL_COMPRESSION AND
+      NOT GCOV_TARBALL_COMPRESSION MATCHES "^(GZIP|BZIP2|XZ|ZSTD|FROM_EXT)$")
+    message(FATAL_ERROR "TARBALL_COMPRESSION must be one of OFF, GZIP, "
+      "BZIP2, XZ, ZSTD, or FROM_EXT for ctest_coverage_collect_gcov")
+  endif()
   # run gcov on each gcda file in the binary tree
   set(gcda_files)
   set(label_files)
@@ -270,14 +288,38 @@ ${label_files}
 ${uncovered_files_for_tar}
 ")
 
-  if (GCOV_QUIET)
-    set(tar_opts "cfj")
-  else()
-    set(tar_opts "cvfj")
+  # Prepare tar command line arguments
+
+  set(tar_opts "")
+  # Select data compression mode
+  if( GCOV_TARBALL_COMPRESSION STREQUAL "FROM_EXT")
+    if( GCOV_TARBALL MATCHES [[\.(tgz|tar.gz)$]] )
+      string(APPEND tar_opts "z")
+    elseif( GCOV_TARBALL MATCHES [[\.(txz|tar.xz)$]] )
+      string(APPEND tar_opts "J")
+    elseif( GCOV_TARBALL MATCHES [[\.(tbz|tar.bz)$]] )
+      string(APPEND tar_opts "j")
+    endif()
+  elseif(GCOV_TARBALL_COMPRESSION STREQUAL "GZIP")
+    string(APPEND tar_opts "z")
+  elseif(GCOV_TARBALL_COMPRESSION STREQUAL "XZ")
+    string(APPEND tar_opts "J")
+  elseif(GCOV_TARBALL_COMPRESSION STREQUAL "BZIP2")
+    string(APPEND tar_opts "j")
+  elseif(GCOV_TARBALL_COMPRESSION STREQUAL "ZSTD")
+    set(zstd_tar_opt "--zstd")
+  endif()
+  # Verbosity options
+  if(NOT GCOV_QUIET AND NOT tar_opts MATCHES v)
+    string(APPEND tar_opts "v")
   endif()
+  # Prepend option 'c' specifying 'create'
+  string(PREPEND tar_opts "c")
+  # Append option 'f' so that the next argument is the filename
+  string(APPEND tar_opts "f")
 
   execute_process(COMMAND
-    ${CMAKE_COMMAND} -E tar ${tar_opts} ${GCOV_TARBALL}
+    ${CMAKE_COMMAND} -E tar ${tar_opts} ${GCOV_TARBALL} ${zstd_tar_opt}
     "--mtime=1970-01-01 0:0:0 UTC"
     "--format=gnutar"
     --files-from=${coverage_dir}/coverage_file_list.txt
diff --git a/Tests/CTestCoverageCollectGCOV/test.cmake.in b/Tests/CTestCoverageCollectGCOV/test.cmake.in
index 9fe25a7552..1818888755 100644
--- a/Tests/CTestCoverageCollectGCOV/test.cmake.in
+++ b/Tests/CTestCoverageCollectGCOV/test.cmake.in
@@ -8,6 +8,10 @@ ctest_configure()
 ctest_build()
 ctest_test()
 
+#------------------------------------------------------------------------------#
+# Common setup for all tests.
+#------------------------------------------------------------------------------#
+
 list(APPEND CTEST_CUSTOM_COVERAGE_EXCLUDE
   "/foo/something"
   "/3rdparty/"
@@ -15,15 +19,118 @@ list(APPEND CTEST_CUSTOM_COVERAGE_EXCLUDE
   "/CMakeFiles/"
 )
 list(APPEND CTEST_EXTRA_COVERAGE_GLOB "*.cpp")
-
 include(CTestCoverageCollectGCOV)
+set(expected_out
+  CMakeFiles/myexecutable.dir/Labels.json
+  Testing/CoverageInfo/data.json
+  Testing/CoverageInfo/extra.cpp.gcov
+  Testing/CoverageInfo/main.cpp.gcov
+  uncovered/extra/uncovered1.cpp
+  uncovered/uncovered2.cpp
+)
+
+#------------------------------------------------------------------------------#
+# Test 1: with standard arguments
+#------------------------------------------------------------------------------#
+
+set(tar_file ${CTEST_BINARY_DIRECTORY}/gcov.tbz)
+ctest_coverage_collect_gcov(
+  TARBALL "${tar_file}"
+  SOURCE "${CTEST_SOURCE_DIRECTORY}"
+  BUILD "${CTEST_BINARY_DIRECTORY}"
+  GCOV_COMMAND "${CMAKE_COMMAND}"
+  GCOV_OPTIONS -P "@CMake_SOURCE_DIR@/Tests/CTestCoverageCollectGCOV/fakegcov.cmake")
+file(REMOVE_RECURSE "${CTEST_BINARY_DIRECTORY}/uncovered")
+
+execute_process(COMMAND
+  ${CMAKE_COMMAND} -E tar tf ${tar_file}
+  OUTPUT_VARIABLE out
+  WORKING_DIRECTORY ${CMAKE_CURRENT_LIST_DIR}
+  OUTPUT_STRIP_TRAILING_WHITESPACE
+)
+
+string(REPLACE "\n" ";" out "${out}")
+list(SORT out)
+
+if("${out}" STREQUAL "${expected_out}")
+  message("PASSED with correct output: ${out}")
+else()
+  message(FATAL_ERROR "FAILED: expected:\n${expected_out}\nGot:\n${out}")
+endif()
+
+#------------------------------------------------------------------------------#
+# Test 2: with optional argument: TARBALL_COMPRESSION "GZIP"
+#------------------------------------------------------------------------------#
+
+set(tar_file ${CTEST_BINARY_DIRECTORY}/gcov.tgz)
+ctest_coverage_collect_gcov(
+  TARBALL "${tar_file}"
+  TARBALL_COMPRESSION "GZIP"
+  SOURCE "${CTEST_SOURCE_DIRECTORY}"
+  BUILD "${CTEST_BINARY_DIRECTORY}"
+  GCOV_COMMAND "${CMAKE_COMMAND}"
+  GCOV_OPTIONS -P "@CMake_SOURCE_DIR@/Tests/CTestCoverageCollectGCOV/fakegcov.cmake")
+file(REMOVE_RECURSE "${CTEST_BINARY_DIRECTORY}/uncovered")
+
+execute_process(COMMAND
+  ${CMAKE_COMMAND} -E tar tf ${tar_file}
+  OUTPUT_VARIABLE out
+  WORKING_DIRECTORY ${CMAKE_CURRENT_LIST_DIR}
+  OUTPUT_STRIP_TRAILING_WHITESPACE
+)
+
+string(REPLACE "\n" ";" out "${out}")
+list(SORT out)
+
+if("${out}" STREQUAL "${expected_out}")
+  message("PASSED with correct output: ${out}")
+else()
+  message(FATAL_ERROR "FAILED: expected:\n${expected_out}\nGot:\n${out}")
+endif()
+
+#------------------------------------------------------------------------------#
+# Test 3: with optional argument: TARBALL_COMPRESSION "FROM_EXT"
+#------------------------------------------------------------------------------#
+
+set(tar_file ${CTEST_BINARY_DIRECTORY}/gcov.txz)
+ctest_coverage_collect_gcov(
+  TARBALL "${tar_file}"
+  TARBALL_COMPRESSION "FROM_EXT"
+  SOURCE "${CTEST_SOURCE_DIRECTORY}"
+  BUILD "${CTEST_BINARY_DIRECTORY}"
+  GCOV_COMMAND "${CMAKE_COMMAND}"
+  GCOV_OPTIONS -P "@CMake_SOURCE_DIR@/Tests/CTestCoverageCollectGCOV/fakegcov.cmake")
+file(REMOVE_RECURSE "${CTEST_BINARY_DIRECTORY}/uncovered")
+
+execute_process(COMMAND
+  ${CMAKE_COMMAND} -E tar tf ${tar_file}
+  OUTPUT_VARIABLE out
+  WORKING_DIRECTORY ${CMAKE_CURRENT_LIST_DIR}
+  OUTPUT_STRIP_TRAILING_WHITESPACE
+)
+
+string(REPLACE "\n" ";" out "${out}")
+list(SORT out)
+
+if("${out}" STREQUAL "${expected_out}")
+  message("PASSED with correct output: ${out}")
+else()
+  message(FATAL_ERROR "FAILED: expected:\n${expected_out}\nGot:\n${out}")
+endif()
+
+#------------------------------------------------------------------------------#
+# Test 4: with optional argument: TARBALL_COMPRESSION "FALSE"
+#------------------------------------------------------------------------------#
+
 set(tar_file ${CTEST_BINARY_DIRECTORY}/gcov.tar)
 ctest_coverage_collect_gcov(
   TARBALL "${tar_file}"
+  TARBALL_COMPRESSION "FALSE"
   SOURCE "${CTEST_SOURCE_DIRECTORY}"
   BUILD "${CTEST_BINARY_DIRECTORY}"
   GCOV_COMMAND "${CMAKE_COMMAND}"
   GCOV_OPTIONS -P "@CMake_SOURCE_DIR@/Tests/CTestCoverageCollectGCOV/fakegcov.cmake")
+file(REMOVE_RECURSE "${CTEST_BINARY_DIRECTORY}/uncovered")
 
 execute_process(COMMAND
   ${CMAKE_COMMAND} -E tar tf ${tar_file}
@@ -35,15 +142,36 @@ execute_process(COMMAND
 string(REPLACE "\n" ";" out "${out}")
 list(SORT out)
 
-set(expected_out
-  CMakeFiles/myexecutable.dir/Labels.json
-  Testing/CoverageInfo/data.json
-  Testing/CoverageInfo/extra.cpp.gcov
-  Testing/CoverageInfo/main.cpp.gcov
-  uncovered/extra/uncovered1.cpp
-  uncovered/uncovered2.cpp
+if("${out}" STREQUAL "${expected_out}")
+  message("PASSED with correct output: ${out}")
+else()
+  message(FATAL_ERROR "FAILED: expected:\n${expected_out}\nGot:\n${out}")
+endif()
+
+#------------------------------------------------------------------------------#
+# Test 5: with optional argument: TARBALL_COMPRESSION "ZSTD"
+#------------------------------------------------------------------------------#
+
+set(tar_file ${CTEST_BINARY_DIRECTORY}/gcov.zstd)
+ctest_coverage_collect_gcov(
+  TARBALL "${tar_file}"
+  TARBALL_COMPRESSION "ZSTD"
+  SOURCE "${CTEST_SOURCE_DIRECTORY}"
+  BUILD "${CTEST_BINARY_DIRECTORY}"
+  GCOV_COMMAND "${CMAKE_COMMAND}"
+  GCOV_OPTIONS -P "@CMake_SOURCE_DIR@/Tests/CTestCoverageCollectGCOV/fakegcov.cmake")
+file(REMOVE_RECURSE "${CTEST_BINARY_DIRECTORY}/uncovered")
+
+execute_process(COMMAND
+  ${CMAKE_COMMAND} -E tar tf ${tar_file}
+  OUTPUT_VARIABLE out
+  WORKING_DIRECTORY ${CMAKE_CURRENT_LIST_DIR}
+  OUTPUT_STRIP_TRAILING_WHITESPACE
 )
 
+string(REPLACE "\n" ";" out "${out}")
+list(SORT out)
+
 if("${out}" STREQUAL "${expected_out}")
   message("PASSED with correct output: ${out}")
 else()