diff options
| author | Martin Czygan <martin.czygan@gmail.com> | 2021-07-15 03:33:19 +0200 |
|---|---|---|
| committer | Martin Czygan <martin.czygan@gmail.com> | 2021-07-15 03:33:19 +0200 |
| commit | 6e7baa12619e8b51218bcefb2c6dd75525eb7387 (patch) | |
| tree | c3c6c09256eda3eda5f756c3f1c5bab7094e906a /extra/cdx | |
| parent | 1e03f7f9b0dc4fe2bf856c542d011de72962ce10 (diff) | |
| download | refcat-6e7baa12619e8b51218bcefb2c6dd75525eb7387.tar.gz refcat-6e7baa12619e8b51218bcefb2c6dd75525eb7387.zip | |
cdx reshape: only include hits
Diffstat (limited to 'extra/cdx')
| -rw-r--r-- | extra/cdx/cdx_reshape.py | 3 |
1 files changed, 1 insertions, 2 deletions
diff --git a/extra/cdx/cdx_reshape.py b/extra/cdx/cdx_reshape.py index 6b3d6e5..28b6fc0 100644 --- a/extra/cdx/cdx_reshape.py +++ b/extra/cdx/cdx_reshape.py @@ -9,8 +9,7 @@ by_url = {} for line in fileinput.input(): line = line.strip() doc = json.loads(line) - last = doc.get("summary", {}).get("last", "") - if not last: + if doc.get("numRows") == 0: continue by_url[doc["line"]] = doc |