def datetime(value):
if not _convert:
_late_import()
if isinstance(value, (date, builtin_datetime)):
pass
elif value < 10000000000:
value = _convert.unix2datetime(value)
else:
value = _convert.milli2datetime(value)
return _convert.datetime2string(value, "%Y-%m-%d %H:%M:%S")
def unix(value):
if not _convert:
_late_import()
if isinstance(value, (date, builtin_datetime)):
pass
elif value < 10000000000:
value = _convert.unix2datetime(value)
else:
value = _convert.milli2datetime(value)
return str(_convert.datetime2unix(value))
def unix(value):
if not _convert:
_late_import()
if isinstance(value, (date, builtin_datetime)):
pass
elif value < 10000000000:
value = _convert.unix2datetime(value)
else:
value = _convert.milli2datetime(value)
return str(_convert.datetime2unix(value))
def datetime(value):
if not _convert:
_late_import()
if isinstance(value, (date, builtin_datetime)):
pass
elif value < 10000000000:
value = _convert.unix2datetime(value)
else:
value = _convert.milli2datetime(value)
return _convert.datetime2string(value, "%Y-%m-%d %H:%M:%S")
def get_last_updated(es):
try:
results = es.search({
"query": {"filtered": {
"query": {"match_all": {}},
"filter": {
"range": {
"modified_ts": {"gte": convert.datetime2milli(far_back)}
}}
}},
"from": 0,
"size": 0,
"sort": [],
"facets": {"modified_ts": {"statistical": {"field": "modified_ts"}}}
})
if results.facets.modified_ts.count == 0:
return convert.milli2datetime(0)
return convert.milli2datetime(results.facets.modified_ts.max)
except Exception, e:
return convert.milli2datetime(0)
def main(settings):
current_time = datetime.utcnow()
time_file = File(settings.param.last_replication_time)
# SYNCH WITH source ES INDEX
source = Index(settings.source)
destination = Cluster(settings.destination).get_or_create_index(settings.destination)
# GET LAST UPDATED
from_file = None
if time_file.exists:
from_file = convert.milli2datetime(convert.value2int(time_file.read()))
from_es = get_last_updated(destination) - timedelta(hours=1)
last_updated = MIN(coalesce(from_file, convert.milli2datetime(0)), from_es)
Log.note("updating records with modified_ts>={{last_updated}}", {"last_updated": last_updated})
pending = get_pending(source, last_updated)
with ThreadedQueue(destination, batch_size=1000) as data_sink:
replicate(source, data_sink, pending, last_updated)
# RECORD LAST UPDATED
time_file.write(unicode(convert.datetime2milli(current_time)))
def update_repo(repo, settings):
with MySQL(settings.database) as db:
try:
pull_repo(repo)
# GET LATEST DATE
existing_range = db.query("""
SELECT
max(`date`) `max`,
min(`date`) `min`,
min(revision) min_rev,
max(revision) max_rev
FROM
changesets
WHERE
repos={{repos}}
""", {"repos": repo.name})[0]
ranges = wrap([
{"min": coalesce(existing_range.max, convert.milli2datetime(0)) + timedelta(days=1)},
{"max": existing_range.min}
])
for r in ranges:
for g, docs in qb.groupby(get_changesets(date_range=r, repo=repo), size=100):
for doc in docs:
doc.file_changes = None
doc.file_adds = None
doc.file_dels = None
doc.description = doc.description[0:16000]
db.insert_list("changesets", docs)
db.flush()
missing_revisions = find_holes(db, "changesets", "revision", {"min": 0, "max": existing_range.max_rev + 1}, {"term": {"repos": repo.name}})
for _range in missing_revisions:
for g, docs in qb.groupby(get_changesets(revision_range=_range, repo=repo), size=100):
for doc in docs:
doc.file_changes = None
doc.file_adds = None
doc.file_dels = None
doc.description = doc.description[0:16000]
db.insert_list("changesets", docs)
db.flush()
except Exception, e:
Log.warning("Failure to pull from {{repos.name}}", {"repos": repo}, e)
def update_repo(repo, settings):
with MySQL(settings.database) as db:
try:
pull_repo(repo)
# GET LATEST DATE
existing_range = db.query(
"""
SELECT
max(`date`) `max`,
min(`date`) `min`,
min(revision) min_rev,
max(revision) max_rev
FROM
changesets
WHERE
repos={{repos}}
""", {"repos": repo.name})[0]
ranges = wrap([{
"min":
coalesce(existing_range.max, convert.milli2datetime(0)) +
timedelta(days=1)
}, {
"max": existing_range.min
}])
for r in ranges:
for g, docs in qb.groupby(get_changesets(date_range=r,
repo=repo),
size=100):
for doc in docs:
doc.file_changes = None
doc.file_adds = None
doc.file_dels = None
doc.description = doc.description[0:16000]
db.insert_list("changesets", docs)
db.flush()
missing_revisions = find_holes(db, "changesets", "revision", {
"min": 0,
"max": existing_range.max_rev + 1
}, {"term": {
"repos": repo.name
}})
for _range in missing_revisions:
for g, docs in qb.groupby(get_changesets(revision_range=_range,
repo=repo),
size=100):
for doc in docs:
doc.file_changes = None
doc.file_adds = None
doc.file_dels = None
doc.description = doc.description[0:16000]
db.insert_list("changesets", docs)
db.flush()
except Exception, e:
Log.warning("Failure to pull from {{repos.name}}", {"repos": repo},
e)
