From e0aa71419f3a69c226bbdbda960267ac4dd04413 Mon Sep 17 00:00:00 2001 From: Martin Czygan Date: Thu, 15 Jul 2021 18:08:35 +0200 Subject: tasks: tweak CDXURL --- python/refcat/tasks.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'python') diff --git a/python/refcat/tasks.py b/python/refcat/tasks.py index 269c4f8..1dfaccb 100644 --- a/python/refcat/tasks.py +++ b/python/refcat/tasks.py @@ -1497,7 +1497,7 @@ class CDXURL(Refcat): Stub implementation of ad-hoc CDX. We only consider a subset of documents. """ cache = luigi.Parameter(default="/magna/.cache/skate/cdx", significant=False) - limit = luigi.IntParameter(default=10000, significant=False) + limit = luigi.IntParameter(default=50000, significant=False) def requires(self): return RefsURL() @@ -1509,7 +1509,7 @@ class CDXURL(Refcat): LC_ALL=C head -n {limit} | skate-cdx-lookup -q -s 50ms -c {cache} -j -B | skate-map -m cdxu | - LC_ALL=C sort -T {tmpdir} -k1,1 -S25% | + LC_ALL=C sort -u -T {tmpdir} -k1,1 -S25% | zstd -c -T0 > {output} """, tmpdir=self.tmpdir, -- cgit v1.2.3