import psutil, time, datetime
from infi.clickhouse_orm import Database
from models import CPUStats
db = Database('demo')
db.create_table(CPUStats)
psutil.cpu_percent(percpu=True) # first sample should be discarded
while True:
time.sleep(1)
stats = psutil.cpu_percent(percpu=True)
timestamp = datetime.datetime.now()
print(timestamp)
db.insert([
CPUStats(timestamp=timestamp, cpu_id=cpu_id, cpu_percent=cpu_percent)
for cpu_id, cpu_percent in enumerate(stats)
])
def get_fragments(filename):
'''
Converts a text file at the given path to a generator
of Fragment instances.
'''
from os import path
document = path.splitext(path.basename(filename))[0]
idx = 0
for word, stem in parse_file(filename):
idx += 1
yield Fragment(document=document, idx=idx, word=word, stem=stem)
print('{} - {} words'.format(filename, idx))
if __name__ == '__main__':
# Load NLTK data if necessary
nltk.download('punkt')
nltk.download('wordnet')
# Initialize database
db = Database('default')
db.create_table(Fragment)
# Load files from the command line or everything under ebooks/
filenames = sys.argv[1:] or glob('ebooks/*.txt')
for filename in filenames:
db.insert(get_fragments(filename), batch_size=100000)
