From 2de8ce9b37bfaed7c5a962a5316dbc9c9c3cb2c8 Mon Sep 17 00:00:00 2001 From: David Bremner Date: Sat, 30 Nov 2013 23:00:03 -0400 Subject: perf-test: bump version to 0.4, use manifest files The new revision of the performance test includes manifests for each corpus, so update the support library to use these manifests at the same time. --- performance-test/perf-test-lib.sh | 72 ++++++++++++++++++++++++--------------- 1 file changed, 45 insertions(+), 27 deletions(-) (limited to 'performance-test/perf-test-lib.sh') diff --git a/performance-test/perf-test-lib.sh b/performance-test/perf-test-lib.sh index 9ee76613..44708cfd 100644 --- a/performance-test/perf-test-lib.sh +++ b/performance-test/perf-test-lib.sh @@ -41,52 +41,70 @@ add_email_corpus () { rm -rf ${MAIL_DIR} - case "$corpus_size" in - small) - mail_subdir="mail/enron/bailey-s" - check_for="${TEST_DIRECTORY}/corpus/$mail_subdir" - ;; - medium) - mail_subdir="mail/notmuch-archive" - check_for="${TEST_DIRECTORY}/corpus/$mail_subdir" - ;; - *) - mail_subdir=mail - check_for="${TEST_DIRECTORY}/corpus/$mail_subdir/enron/wolfe-j" - esac + CORPUS_DIR=${TEST_DIRECTORY}/corpus + mkdir -p "${CORPUS_DIR}" - MAIL_CORPUS="${TEST_DIRECTORY}/corpus/$mail_subdir" - TAG_CORPUS="${TEST_DIRECTORY}/corpus/tags" + MAIL_CORPUS="${CORPUS_DIR}/mail.${corpus_size}" + TAG_CORPUS="${CORPUS_DIR}/tags" - args=() - if [ ! -d "$TAG_CORPUS" ] ; then - args+=("notmuch-email-corpus/tags") + if command -v pixz > /dev/null; then + XZ=pixz + else + XZ=xz fi - if [ ! -d "$check_for" ] ; then - args+=("notmuch-email-corpus/$mail_subdir") + if [ ! -d "${CORPUS_DIR}/manifest" ]; then + + printf "Unpacking manifests\n" + tar --extract --use-compress-program ${XZ} --strip-components=1 \ + --directory ${TEST_DIRECTORY}/corpus \ + --wildcards --file ../download/notmuch-email-corpus-${PERFTEST_VERSION}.tar.xz \ + 'notmuch-email-corpus/manifest/*' fi - if [[ ${#args[@]} > 0 ]]; then - if command -v pixz > /dev/null; then - XZ=pixz + file_list=$(mktemp file_listXXXXXX) + if [ ! -d "$TAG_CORPUS" ] ; then + echo "notmuch-email-corpus/tags" >> $file_list + fi + + if [ ! -d "$MAIL_CORPUS" ] ; then + if [[ "$corpus_size" != "large" ]]; then + sed s,^,notmuch-email-corpus/, < \ + ${TEST_DIRECTORY}/corpus/manifest/MANIFEST.${corpus_size} >> $file_list else - XZ=xz + echo "notmuch-email-corpus/mail" >> $file_list fi + fi - printf "Unpacking corpus\n" - mkdir -p "${TEST_DIRECTORY}/corpus" + if [[ -s $file_list ]]; then + printf "Unpacking corpus\n" tar --checkpoint=.5000 --extract --strip-components=1 \ --directory ${TEST_DIRECTORY}/corpus \ --use-compress-program ${XZ} \ --file ../download/notmuch-email-corpus-${PERFTEST_VERSION}.tar.xz \ - "${args[@]}" + --anchored --recursion \ + --files-from $file_list printf "\n" + if [[ ! -d ${MAIL_CORPUS} ]]; then + printf "creating link farm\n" + + if [[ "$corpus_size" = large ]]; then + cp -rl ${TEST_DIRECTORY}/corpus/mail ${MAIL_CORPUS} + else + while read -r file; do + tdir=${MAIL_CORPUS}/$(dirname $file) + mkdir -p $tdir + ln ${TEST_DIRECTORY}/corpus/$file $tdir + done <${TEST_DIRECTORY}/corpus/manifest/MANIFEST.${corpus_size} + fi + fi + fi + rm $file_list cp -lr $TAG_CORPUS $TMP_DIRECTORY/corpus.tags cp -lr $MAIL_CORPUS $MAIL_DIR } -- cgit v1.2.3