diff --git a/dbFactory.sh b/dbFactory.sh
index cf7292a..203f785 100644
--- a/dbFactory.sh
+++ b/dbFactory.sh
@@ -12,8 +12,11 @@ fi
 target=$1
 
 cat lib/Compose.pre \
-	| grep -i '^<multi' \
-	| grep -v '<dead' \
+	| grep '^<Multi_key>' \
+	| grep -v '<dead_' \
+	| grep -v '<Cyrillic_' \
+	| grep -v '<Greek_' \
+	| grep -v '<kana_' \
 	| grep -v '<U....>' \
 	| grep -v '<U.....>' \
 	| sed  --regexp-extended 's/(.*>)\s*:\s\"(.*)\"\s*(\S*)\s*#\s*(.*)/\1+\2+\3+\L\4/' \
diff --git a/lib/README.txt b/lib/README.txt
index 698882c..5f6af81 100644
--- a/lib/README.txt
+++ b/lib/README.txt
@@ -3,22 +3,20 @@ How to get the list of usable key-sequences?
 Download the X11 compose-sequences for en_US.UTF-8¹ and get rid of some things:
 - Comments (XCOMM)
 - Sequences not started my Multi_key
-- Sequences containing non-standard characters
+- Sequences containing non-standard characters (U.... and U.....)
+- Sequences containing non-international keys (cyrillic, greek, kana)
 - Sequences containing other dead keys
+This is done by grep in dbFactory.sh
 
 ¹) en_US.UTF-8 seems to be quite complete and also available on other locales.
 
-$ wget https://cgit.freedesktop.org/xorg/lib/libX11/plain/nls/en_US.UTF-8/Compose.pre
-$ grep -i '^<multi' Compose.pre | grep -v '<dead' | grep -v '<U....>' | grep -v '<U.....>' > Compose.usable
-
 
 Convert to psv
 --------------
 - Plus-separated-values because + is not used elsewhere in the file.
 - Regex to extract the fields into psv.
 - The second description is all-uppper, we'll convert it to all lower.
-
-$ sed --regexp-extended 's/(.*>)\s*:\s\"(.*)\"\s*(\S*)\s*#\s*(.*)/\1+\2+\3+\L\4/' Compose.usable > Compose.psv
+This is done by sed in dbFactory.sh
 
 Load into SQLite
 ----------------