p15679402
/
bigdata
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

.

This commit is contained in:
guange 2019-01-17 23:13:38 +08:00
parent 6392531b09
commit 4a8a5288f4
1 changed files with 20 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

21
chapter3/spark.py
View File

@ -25,6 +25,9 @@ def mysql_execute(sql):
finally: finally:
db.close() db.close()
def today():
return time.strftime('%Y-%m-%d')
if __name__ == "__main__": if __name__ == "__main__":
# $example on:spark_hive$ # $example on:spark_hive$
@ -38,6 +41,22 @@ if __name__ == "__main__":
.enableHiveSupport() \ .enableHiveSupport() \
.getOrCreate() .getOrCreate()
df = spark.sql("select * from jd limit 1") df = sqlContext.sql("select count(*) as N from jd_comment")
jd_comment_count = df.rdd.collect()[0]["N"]
df = sqlContext.sql("select count(*) as N from jd_comment where created_at like '"+today()+"%'")
jd_comment_today_count = df.rdd.collect()[0]["N"]
df = sqlContext.sql("select count(*) as N from jd")
jd_count = df.rdd.collect()[0]["N"]
df = sqlContext.sql("select count(*) as N from jd where created_at like '"+today()+"%'")
jd_today_count = df.rdd.collect()[0]["N"]
total_count = jd_comment_count + jd_count
today_total_count = jd_comment_today_count + jd_today_count
mysql_execute("insert into crawl_infos (total_count, today_total_count, product_count, today_product_count, comment_count, today_comment_count) values ({},{},{},{},{},{})".format(
total_count, today_total_count, jd_count,jd_today_count, jd_comment_count, jd_comment_today_count) )
spark.stop() spark.stop()