def extract_from_datazilla_using_id(es, settings, transformer):
existing_ids = get_existing_ids(es, settings, transformer.pushlog.keys())
max_existing_id = nvl(MAX(existing_ids), settings.production.min)
holes = set(range(settings.production.min, max_existing_id)) - existing_ids
missing_ids = set(range(settings.production.min, max_existing_id+nvl(settings.production.step, NUM_PER_BATCH))) - existing_ids
Log.note("Number missing: {{num}}", {"num": len(missing_ids)})
Log.note("Number in holes: {{num}}", {"num": len(holes)})
#FASTER IF NO INDEXING IS ON
es.set_refresh_interval(-1)
#FILE IS FASTER THAN NETWORK
if (len(holes) > 10000 or settings.args.scan_file or settings.args.restart) and File(settings.param.output_file).exists:
#ASYNCH PUSH TO ES IN BLOCKS OF 1000
with Timer("Scan file for missing ids"):
with ThreadedQueue(es, size=nvl(es.settings.batch_size, 100)) as json_for_es:
num = 0
for line in File(settings.param.output_file):
try:
if len(line.strip()) == 0:
continue
col = line.split("\t")
id = int(col[0])
# if id==3003529:
# Log.debug()
if id < settings.production.min:
continue
if id in existing_ids:
continue
if num > settings.production.step:
return
num += 1
with Profiler("decode and transform"):
data = CNV.JSON2object(col[-1])
if data.test_run_id:
with Profiler("transform"):
data = transformer.transform(id, data)
json_for_es.extend({"value": d} for d in data)
Log.note("Added {{id}} from file", {"id": id})
existing_ids.add(id)
else:
Log.note("Skipped {{id}} from file (no test_run_id)", {"id": id})
num -= 1
except Exception, e:
Log.warning("Bad line id={{id}} ({{length}}bytes):\n\t{{prefix}}", {
"id": id,
"length": len(CNV.object2JSON(line)),
"prefix": CNV.object2JSON(line)[0:130]
}, e)
missing_ids = missing_ids - existing_ids
def arrays_add(id, path, r):
try:
if isinstance(r, dict):
for k, v in [(k, v) for k, v in r.items()]:
new_path = path + "[" + k + "]"
arrays_add(id, new_path, v)
elif isinstance(r, list):
try:
values = r.map(float)
arrays.append([id, path, len(values), 1])
except Exception, e:
for i, v in enumerate(r):
r[i] = arrays_add(id, path + "[" + str(i) + "]", v)
# return r
except Exception, e:
Log.warning("Can not summarize: {{json}}", {"json": CNV.object2JSON(r)})
added.add(id)
data = CNV.JSON2object(col[1])
records_for_db.add({
"id": nvl(data.test_run_id, id),
"branch": data.json_blob.test_build.branch,
"name": data.json_blob.test_build.name,
"version": data.json_blob.test_build.version,
"suite": data.json_blob.testrun.suite,
"revision": data.json_blob.test_build.revision,
"date": data.json_blob.testrun.date
})
Log.note("Added {{id}} from file", {"id": data.test_run_id})
except Exception, e:
Log.warning("Bad line ({{length}}bytes):\n\t{{prefix}}", {
"length": len(CNV.object2JSON(line)),
"prefix": CNV.object2JSON(line)[0:130]
}, e)
def main():
try:
settings = startup.read_settings(filename="file2db_settings.json")
Log.start(settings.debug)
with DB(settings.db) as db:
db.execute("""
DROP TABLE IF EXISTS b2g_tests
""")
with Timer("read {{id}} from DZ", {"id": id}):
content = requests.get(url, timeout=nvl(settings.production.timeout, 30)).content
except Exception, e:
Log.warning("Failure to read from {{url}}", {"url": url}, e)
return False
try:
if content.startswith("Id not found"):
Log.note("{{id}} not found {{url}}", {"id": id, "url": url})
if id < max_id:
return True
else:
return False
data = CNV.JSON2object(content.decode('utf-8'))
content = CNV.object2JSON(data) #ENSURE content HAS NO crlf
if data.test_run_id:
Log.println("Add {{id}} for revision {{revision}} ({{bytes}} bytes)", {
"id": id,
"revision": data.json_blob.test_build.revision,
"bytes": len(content)
})
with Profiler("transform"):
result = transformer.transform(id, data)
if result:
Log.println("{{num}} records to add", {
"num": len(result)
})
es_sink.extend({"value": d} for d in result)
