Commit f066fdea authored by liyang's avatar liyang

fix:爬取youtube

parent 04d6f615
...@@ -7,6 +7,7 @@ from utils.filse import save_json ...@@ -7,6 +7,7 @@ from utils.filse import save_json
from api.index import importJson, getReptileTask, importJsonPath from api.index import importJson, getReptileTask, importJsonPath
from utils.index import convert_to_traditional, yt_dlp_download,convert_string_to_time from utils.index import convert_to_traditional, yt_dlp_download,convert_string_to_time
# from pytube import YouTube # from pytube import YouTube
import os
from config.settings import get_base_file_url from config.settings import get_base_file_url
def reptile(browser=None, search_word=""): def reptile(browser=None, search_word=""):
""" """
...@@ -41,7 +42,7 @@ def reptile(browser=None, search_word=""): ...@@ -41,7 +42,7 @@ def reptile(browser=None, search_word=""):
base_urr = get_base_file_url() base_urr = get_base_file_url()
log.debug(url) log.debug(url)
state_download = yt_dlp_download(url, 'youtube') state_download = yt_dlp_download(url, 'youtube')
file_http_src = f'{base_urr}/youtube/{id}.mp4' file_http_src = f'{base_urr}youtube/{id}.mp4'
if state_download: if state_download:
# 组装数据 # 组装数据
obj = { obj = {
...@@ -52,7 +53,7 @@ def reptile(browser=None, search_word=""): ...@@ -52,7 +53,7 @@ def reptile(browser=None, search_word=""):
"reptileTime": str(int(time.time())), "reptileTime": str(int(time.time())),
"type": '视频', "type": '视频',
"author": element_author_list[index].text, "author": element_author_list[index].text,
"releaseTime": convert_string_to_time(element_time_list[index].text) "releaseTime": str(convert_string_to_time(element_time_list[index].text))
} }
data.append(obj) data.append(obj)
else: else:
...@@ -61,7 +62,7 @@ def reptile(browser=None, search_word=""): ...@@ -61,7 +62,7 @@ def reptile(browser=None, search_word=""):
if len(data) > 0: if len(data) > 0:
# 保存json文件到本地 # 保存json文件到本地
log.debug(os.path.abspath("../")) log.debug(os.path.abspath("../"))
file_dir = f'{os.path.join(os.path.abspath("../"), "network-assets-reptile", "reptile_data", "ptt", str(int(time.time())) + ".json")}' file_dir = f'{os.path.join(os.path.abspath("../"), "network-assets-reptile", "reptile_data", "youtube", str(int(time.time())) + ".json")}'
# file_dir = f'./reptile_data/ptt/{int(time.time())}.json' # file_dir = f'./reptile_data/ptt/{int(time.time())}.json'
state_save = save_json(file_dir, data) state_save = save_json(file_dir, data)
# log.debug("-----------------------------") # log.debug("-----------------------------")
......
...@@ -6,7 +6,7 @@ import ssl ...@@ -6,7 +6,7 @@ import ssl
import subprocess import subprocess
from utils.Logger import log from utils.Logger import log
from datetime import datetime, timedelta from datetime import datetime, timedelta
from dateutil.relativedelta import relativedelta # from dateutil.relativedelta import relativedelta
ssl._create_default_https_context = ssl._create_stdlib_context ssl._create_default_https_context = ssl._create_stdlib_context
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment