[R-package] Suppresses unknown pragma warnings during CRAN build (#3460)

* Modified script to also remove pragma warnings.

This also includes modifying the scope of the pragma removal.
Previously this script only searched inside the ./src/include/LightGMB directory.
This was not inclusive enough to remove the warnings in files shown in the ticket.

* Adds CI test that exits if unknown pragma warnings are present.

* Expanding pragma removal to cpp and hpp files.

* Update .ci/test_r_package.sh

Removing unneeded conditions since this script will only run on Linux and Mac builds anyway.

Co-authored-by: James Lamb <jaylamb20@gmail.com>

* Update .ci/test_r_package.sh

Fixes typo

Co-authored-by: James Lamb <jaylamb20@gmail.com>

* replacing double quotes with single quotes

* Using a more portable find syntax so it works on macOS and Linux

Co-authored-by: James Lamb <jaylamb20@gmail.com>

.ci/test_r_package.sh

@@ -216,3 +216,17 @@ if [[ $OS_NAME == "macos" ]] && [[ $R_BUILD_TYPE == "cran" ]]; then
         exit -1
     fi
 fi
+
+# this check makes sure that no "warning: unknown pragma ignored" logs
+# reach the user leading them to believe that something went wrong
+if [[ $R_BUILD_TYPE == "cran" ]]; then
+    pragma_warning_present=$(
+        cat $BUILD_LOG_FILE \
+        | grep -E "warning: unknown pragma ignored" \
+        | wc -l
+    )
+    if [[ $pragma_warning_present -ne 0 ]]; then
+        echo "Unknown pragma warning is present, pragmas should have been removed before build"
+        exit -1
+    fi
+fi

build-cran-package.sh

@@ -50,19 +50,20 @@ cd ${TEMP_R_DIR}
     sed -i.bak -e "s/~~VERSION~~/${LGB_VERSION}/" DESCRIPTION
     sed -i.bak -e "s/~~DATE~~/${CURRENT_DATE}/" DESCRIPTION
 
-    # Remove 'region' and 'endregion' pragmas. This won't change
-    # the correctness of the code. CRAN does not allow you
-    # to use compiler flag '-Wno-unknown-pragmas' or
+    # Remove 'region', 'endregion', and 'warning' pragmas.
+    # This won't change the correctness of the code. CRAN does
+    # not allow you to use compiler flag '-Wno-unknown-pragmas' or
     # pragmas that suppress warnings.
     echo "Removing unknown pragmas in headers"
-    for file in src/include/LightGBM/*.h; do
+    for file in $(find . -name '*.h' -o -name '*.hpp' -o -name '*.cpp'); do
       sed \
         -i.bak \
         -e 's/^.*#pragma region.*$//' \
         -e 's/^.*#pragma endregion.*$//' \
+        -e 's/^.*#pragma warning.*$//' \
         "${file}"
     done
-    rm src/include/LightGBM/*.h.bak
+    find . -name '*.h.bak' -o -name '*.hpp.bak' -o -name '*.cpp.bak' -exec rm {} \;
 
     # When building an R package with 'configure', it seems
     # you're guaranteed to get a shared library called