diff --git a/cmake/scripts/Baseset.cmake b/cmake/scripts/Baseset.cmake
new file mode 100644
--- /dev/null
+++ b/cmake/scripts/Baseset.cmake
@@ -0,0 +1,53 @@
+cmake_minimum_required(VERSION 3.5)
+
+#
+# Create a single baseset meta file with the correct translations.
+#
+
+set(ARGC 1)
+set(ARG_READ NO)
+
+# Read all the arguments given to CMake; we are looking for -- and everything
+# that follows. Those are our language files.
+while(ARGC LESS CMAKE_ARGC)
+    set(ARG ${CMAKE_ARGV${ARGC}})
+
+    if (ARG_READ)
+        list(APPEND LANG_SOURCE_FILES "${ARG}")
+    endif (ARG_READ)
+
+    if (ARG STREQUAL "--")
+        set(ARG_READ YES)
+    endif (ARG STREQUAL "--")
+
+    math(EXPR ARGC "${ARGC} + 1")
+endwhile()
+
+# Place holder format is @<ini_key>_<str_id>@
+file(STRINGS "${BASESET_SOURCE_FILE}" PLACE_HOLDER REGEX "^@")
+string(REGEX REPLACE "@([^_]+).*@" "\\1" INI_KEY "${PLACE_HOLDER}")
+string(REGEX REPLACE "@[^_]+_(.*)@" "\\1" STR_ID "${PLACE_HOLDER}")
+string(REGEX REPLACE "@(.*)@" "\\1" PLACE_HOLDER "${PLACE_HOLDER}")
+
+# Get the translations
+foreach(LANGFILE IN LISTS LANG_SOURCE_FILES)
+    file(STRINGS "${LANGFILE}" LANGLINES REGEX "^(##isocode|${STR_ID})" ENCODING UTF-8)
+    string(FIND "${LANGLINES}" "${STR_ID}" HAS_STR_ID)
+    if (HAS_STR_ID LESS 0)
+        continue()
+    endif (HAS_STR_ID LESS 0)
+    string(REGEX REPLACE "##isocode ([^;]+).*" "\\1" ISOCODE "${LANGLINES}")
+    if ("${ISOCODE}" STREQUAL "en_GB")
+        string(REGEX REPLACE "[^:]*:(.*)" "${INI_KEY}       = \\1" LANGLINES "${LANGLINES}")
+    else()
+        string(REGEX REPLACE "[^:]*:(.*)" "${INI_KEY}.${ISOCODE} = \\1" LANGLINES "${LANGLINES}")
+    endif()
+    list(APPEND ${PLACE_HOLDER} ${LANGLINES})
+endforeach(LANGFILE)
+list(SORT ${PLACE_HOLDER})
+string(REPLACE ";" "\n" ${PLACE_HOLDER} "${${PLACE_HOLDER}}")
+
+# Get the grf md5
+file(MD5 ${BASESET_EXTRAGRF_FILE} ORIG_EXTRA_GRF_MD5)
+
+configure_file(${BASESET_SOURCE_FILE} ${BASESET_BINARY_FILE})