From b34f701b1e36a804b156c714edd601c4b9bc5bf2 Mon Sep 17 00:00:00 2001
From: Eric House <xwords@eehouse.org>
Date: Mon, 23 Jun 2014 07:29:05 -0700
Subject: [PATCH] use comm instead of huge grep loop to filter wordlist

---
 xwords4/dawg/English/Makefile.top5000 | 9 ++++-----
 1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/xwords4/dawg/English/Makefile.top5000 b/xwords4/dawg/English/Makefile.top5000
index 394605258..dad5d2cae 100644
--- a/xwords4/dawg/English/Makefile.top5000
+++ b/xwords4/dawg/English/Makefile.top5000
@@ -32,11 +32,10 @@ SOURCEDICT ?= $(ENGLISH)/CSW.dict.gz
 # word from CSW that's also in top5000 makes the grade
 
 $(XWLANG)Main.dict.gz: $(SOURCEDICT) $(ENGLISH)/Top_5000.dict.gz
-	zcat $(ENGLISH)/Top_5000.dict.gz | tr [a-z] [A-Z] > /tmp/Top_5000.dict
-	rm -f $@
-	for WORD in $$(zcat $< | tr -d '\r' | tr [a-z] [A-Z] | grep -e "^[A-Z]\{2,15\}$$"); do \
-		grep -q "^$$WORD$$" /tmp/Top_5000.dict && echo $$WORD; \
-	done | gzip -c > $@
+	zcat $(ENGLISH)/Top_5000.dict.gz | tr [a-z] [A-Z] | sort > /tmp/Top_5000.dict
+	zcat $< | tr -d '\r' | tr [a-z] [A-Z] | grep -e "^[A-Z]\{2,15\}$$" | sort > /tmp/$$(basename $<)
+	rm -f $@ 
+	comm -12 /tmp/$$(basename $<) /tmp/Top_5000.dict | gzip -c > $@
 
 # Everything but creating of the Main.dict file is inherited from the
 # "parent" Makefile.langcommon in the parent directory.