diff options
author | Craig Tiller <ctiller@google.com> | 2017-03-10 17:24:45 -0800 |
---|---|---|
committer | Craig Tiller <ctiller@google.com> | 2017-03-10 17:24:45 -0800 |
commit | bbfb25bf708688748fd9bf37c03ea7cfc5bcc410 (patch) | |
tree | b6172e96d586a39c4e45109393316920f0f8fb69 /tools/profiling/microbenchmarks | |
parent | 56cea8f2eea9c8cc607bdc2915046d7ba187cd16 (diff) |
Sanitize before bigquery upload
Diffstat (limited to 'tools/profiling/microbenchmarks')
-rwxr-xr-x | tools/profiling/microbenchmarks/bm2bq.py | 19 |
1 files changed, 15 insertions, 4 deletions
diff --git a/tools/profiling/microbenchmarks/bm2bq.py b/tools/profiling/microbenchmarks/bm2bq.py index ae59332f1b..ffb11f57d8 100755 --- a/tools/profiling/microbenchmarks/bm2bq.py +++ b/tools/profiling/microbenchmarks/bm2bq.py @@ -73,6 +73,14 @@ columns = [ ('framing_bytes_per_iteration', 'float'), ] +SANITIZE = { + 'integer': int, + 'float': float, + 'boolean': bool, + 'string': str, + 'timestamp': str, +} + if sys.argv[1] == '--schema': print ',\n'.join('%s:%s' % (k, t.upper()) for k, t in columns) sys.exit(0) @@ -89,7 +97,10 @@ else: writer = csv.DictWriter(sys.stdout, [c for c,t in columns]) for row in bm_json.expand_json(js, js2): - if 'label' in row: - del row['label'] - del row['cpp_name'] - writer.writerow(row) + sane_row = {} + for name, sql_type in columns: + if name in row: + if row[name] == '': continue + sane_row[name] = SANITIZE[sql_type](row[name]) + writer.writerow(sane_row) + |