[PATCH 6/6] HACK: fix broken messages in the perf test corpus

Jani Nikula jani at nikula.org
Wed Oct 16 12:00:13 PDT 2013


The gmime header parser rejects a lot of messages in the perf test
corpus which have this in the middle of headers:

Microsoft Mail Internet Headers Version 2.0

The header parsing stops right there. This illustrates a change in the
parsing. The message is clearly broken, but previously notmuch
accepted it anyway.

This patch "fixes" the messages in the perf test corpus to be able to
do fair comparisons of the parsers.

NOT TO BE MERGED, if that isn't obvious. This is just a quick hack.
---
 performance-test/perf-test-lib.sh | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/performance-test/perf-test-lib.sh b/performance-test/perf-test-lib.sh
index 9ee7661..caec0d0 100644
--- a/performance-test/perf-test-lib.sh
+++ b/performance-test/perf-test-lib.sh
@@ -84,7 +84,11 @@ add_email_corpus ()
 	    "${args[@]}"
 
 	printf "\n"
+	printf "Fix broken messages in corpus..."
 
+	find "${TEST_DIRECTORY}/corpus" -type f -print0 | xargs -0 sed -i -e 's/^Microsoft Mail Internet Headers Version 2\.0/X-Crap: &/'
+
+	printf "\n"
     fi
 
     cp -lr $TAG_CORPUS $TMP_DIRECTORY/corpus.tags
-- 
1.8.4.rc3



More information about the notmuch mailing list