Yurik has uploaded a new change for review.
https://gerrit.wikimedia.org/r/168986
Change subject: log2dfs fix pattern match picker
......................................................................
log2dfs fix pattern match picker
Change-Id: I698d2a5a0d2549dfda016ce712cebe2c88fbfcba
---
M scripts/log2dfs.py
M scripts/logprocessor.py
2 files changed, 9 insertions(+), 7 deletions(-)
git pull ssh://gerrit.wikimedia.org:29418/analytics/zero-sms
refs/changes/86/168986/1
diff --git a/scripts/log2dfs.py b/scripts/log2dfs.py
index 59aa394..a4d7b51 100644
--- a/scripts/log2dfs.py
+++ b/scripts/log2dfs.py
@@ -68,12 +68,12 @@
class LogConverter(LogProcessor):
- def __init__(self, settingsFile='settings/log2dfs.json',
logDatePattern=False):
+ def __init__(self, filePattern=False,
settingsFile='settings/log2dfs.json'):
super(LogConverter, self).__init__(settingsFile, 'w2h')
- if not logDatePattern:
- logDatePattern = r'\d\d\d\d\d\d\d\d'
- self.logFileRe = re.compile(unicode(logDatePattern), re.IGNORECASE)
+ if not filePattern:
+ filePattern = r'\d\d\d\d\d\d\d\d'
+ self.logFileRe = re.compile(unicode(filePattern), re.IGNORECASE)
self.dateRe = re.compile(r'(201\d-\d\d-\d\dT\d\d):\d\d:\d\d(\.\d+)?')
self.urlRe = re.compile(r'^(https?)://([^/]+)([^?#]*)(.*)',
re.IGNORECASE)
@@ -82,7 +82,7 @@
safePrint('Processing log files')
for f in os.listdir(self.pathLogs):
- if not self.logFileRe.match(f):
+ if not self.logFileRe.search(f):
continue
logFile = os.path.join(self.pathLogs, f)
statFile = os.path.join(self.pathCache, f)
@@ -235,5 +235,5 @@
if __name__ == '__main__':
- # LogConverter(logDatePattern=(sys.argv[1] if len(sys.argv) > 1 else
False)).manualRun()
- LogConverter(logDatePattern=(sys.argv[1] if len(sys.argv) > 1 else
False)).safeRun()
+ # LogConverter(filePattern=(sys.argv[1] if len(sys.argv) > 1 else
False)).manualRun()
+ LogConverter(filePattern=(sys.argv[1] if len(sys.argv) > 1 else
False)).safeRun()
diff --git a/scripts/logprocessor.py b/scripts/logprocessor.py
index 36c7548..0e425b1 100644
--- a/scripts/logprocessor.py
+++ b/scripts/logprocessor.py
@@ -125,6 +125,8 @@
self.onSettingsLoaded()
def normalizePath(self, path, relToSettings=True):
+ if not path:
+ return False
if not os.path.isabs(path) and relToSettings:
path = os.path.join(os.path.dirname(self.settingsFile), path)
path = os.path.abspath(os.path.normpath(path))
--
To view, visit https://gerrit.wikimedia.org/r/168986
To unsubscribe, visit https://gerrit.wikimedia.org/r/settings
Gerrit-MessageType: newchange
Gerrit-Change-Id: I698d2a5a0d2549dfda016ce712cebe2c88fbfcba
Gerrit-PatchSet: 1
Gerrit-Project: analytics/zero-sms
Gerrit-Branch: master
Gerrit-Owner: Yurik <yu...@wikimedia.org>
_______________________________________________
MediaWiki-commits mailing list
MediaWiki-commits@lists.wikimedia.org
https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits
