From df2c003e6c7375cc2a30e9c05d68e74d2b604229 Mon Sep 17 00:00:00 2001
From: ale <ale@incal.net>
Date: Sun, 20 Nov 2011 14:53:08 +0000
Subject: [PATCH] query fingerprints in a more efficient way
---
server/djrandom/fingerprint/dedup.py | 15 ++++++++-------
1 file changed, 8 insertions(+), 7 deletions(-)
diff --git a/server/djrandom/fingerprint/dedup.py b/server/djrandom/fingerprint/dedup.py
index daf8b89..585a16a 100644
--- a/server/djrandom/fingerprint/dedup.py
+++ b/server/djrandom/fingerprint/dedup.py
@@ -64,15 +64,16 @@ class DeDuper(object):
log.debug('loading in-memory fingerprint database...')
start = time.time()
fp.erase_database(local=True, really_delete=True)
- mp3s = MP3.query.filter(
- (MP3.artist == u'bonobo')
- & (MP3.state == MP3.READY)
- & (MP3.has_fingerprint == True))
- for mp3 in mp3s:
- code = self._generate_code_json(mp3.echoprint_fp, mp3.sha1)
+ # Skip the ORM and directly query the SQL layer.
+ q = select([Fingerprint.sha1, Fingerprint.echoprint_fp],
+ (MP3.sha1 == Fingerprint.sha1)
+ & (MP3.state == MP3.READY)
+ & (MP3.has_fingerprint == True))
+ for row in engine.execute(q):
+ code = self._generate_code_json(row.echoprint_fp, row.sha1)
if not code:
continue
- self.codes[mp3.sha1] = code['fp']
+ self.codes[row.sha1] = code['fp']
fp.ingest(code, do_commit=False, local=True)
elapsed = time.time() - start
log.debug('loaded in-memory fingerprint database in %g seconds' % elapsed)
--
GitLab