change print to logger

crawlab/app.py

@@ -8,7 +8,7 @@ from celery import Celery
 from flask import Flask
 from flask_cors import CORS
 from flask_restful import Api
-
+from utils.log import other
 from constants.node import NodeStatus
 from db.manager import db_manager
 from routes.schedules import ScheduleApi
@@ -78,7 +78,7 @@ def monitor_nodes_status(celery_app):
         })
 
     def update_nodes_status_online(event):
-        print(event)
+        other.info(f"{event}")
 
     with celery_app.connection() as connection:
         recv = celery_app.events.Receiver(connection, handlers={

crawlab/bin/run_flower.py

@@ -6,11 +6,11 @@ import subprocess
 file_dir = os.path.dirname(os.path.realpath(__file__))
 root_path = os.path.abspath(os.path.join(file_dir, '..'))
 sys.path.append(root_path)
-
+from utils.log import other
 from config import BROKER_URL
 
 if __name__ == '__main__':
     p = subprocess.Popen(['celery', 'flower', '-b', BROKER_URL], stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
     for line in iter(p.stdout.readline, 'b'):
         if line.decode('utf-8') != '':
-            print(line.decode('utf-8'))
+            other.info(line.decode('utf-8'))

crawlab/config.py

@@ -1,7 +1,8 @@
 # project variables
 # 爬虫源码路径
 PROJECT_SOURCE_FILE_FOLDER = '../spiders'
-
+# 配置python虚拟环境的路径
+PYTHON_ENV_PATH="/Users/chennan/Desktop/2019/env/bin/python"
 # 爬虫部署路径
 PROJECT_DEPLOY_FILE_FOLDER = '../deployfile'
 
@@ -29,4 +30,5 @@ MONGO_DB = 'crawlab_test'
 # flask variables
 DEBUG = True
 FLASK_HOST = '127.0.0.1'
-FLASK_PORT = 8000
+FLASK_PORT = 8000
+

crawlab/manage.py

@@ -24,7 +24,7 @@ from routes.spiders import SpiderApi, SpiderImportApi, SpiderManageApi
 from routes.stats import StatsApi
 from routes.tasks import TaskApi
 from tasks.celery import celery_app
-
+from utils.log import other
 # flask app instance
 app = Flask(__name__)
 app.config.from_object('config')
@@ -81,7 +81,7 @@ def run_flower():
     p = subprocess.Popen(['celery', 'flower', '-b', BROKER_URL], stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
     for line in iter(p.stdout.readline, 'b'):
         if line.decode('utf-8') != '':
-            print(line.decode('utf-8'))
+            other.info(line.decode('utf-8'))
 
 
 def run_worker():

crawlab/tasks/spider.py

@@ -2,15 +2,12 @@ import os
 from datetime import datetime
 
 from bson import ObjectId
-from celery.utils.log import get_logger
-
-from config import PROJECT_DEPLOY_FILE_FOLDER, PROJECT_LOGS_FOLDER
+from config import PROJECT_DEPLOY_FILE_FOLDER, PROJECT_LOGS_FOLDER,PYTHON_ENV_PATH
 from constants.task import TaskStatus
 from db.manager import db_manager
 from .celery import celery_app
 import subprocess
-
-logger = get_logger(__name__)
+from utils.log import other as logger
 
 
 @celery_app.task(bind=True)
@@ -19,6 +16,8 @@ def execute_spider(self, id: str):
     hostname = self.request.hostname
     spider = db_manager.get('spiders', id=id)
     command = spider.get('cmd')
+    if command.startswith("env"):
+        command = PYTHON_ENV_PATH + command.replace("env","")
 
     current_working_directory = os.path.join(PROJECT_DEPLOY_FILE_FOLDER, str(spider.get('_id')))
 

crawlab/utils/deploy.py

@@ -1,5 +1,5 @@
 import os, zipfile
-
+from utils.log import other
 
 # 打包目录为zip文件（未压缩）
 def zip_file(source_dir, output_filename):
@@ -20,4 +20,4 @@ def unzip_file(zip_src, dst_dir):
         for file in fz.namelist():
             fz.extract(file, dst_dir)
     else:
-        print('This is not zip')
+        other.info('This is not zip')

crawlab/utils/spider.py

@@ -3,7 +3,6 @@ import os
 from constants.spider import FILE_SUFFIX_LANG_MAPPING, LangType, SUFFIX_IGNORE, SpiderType
 from db.manager import db_manager
 
-
 def get_lang_by_stats(stats: dict) -> LangType:
     """
     :param stats: stats is generated by utils.file.get_file_suffix_stats
@@ -18,7 +17,7 @@ def get_lang_by_stats(stats: dict) -> LangType:
             return FILE_SUFFIX_LANG_MAPPING.get(top_suffix)
         return LangType.OTHER
     except IndexError as e:
-        print(e.args,stats)
+        pass