Skip to content

Commit daf7cec

Browse files
committed
add defaults for default headers, etc.
Signed-off-by: Zhiyuan Chen <this@zyc.ai>
1 parent 3a284d7 commit daf7cec

2 files changed

Lines changed: 9 additions & 8 deletions

File tree

icrawler/crawler.py

Lines changed: 2 additions & 8 deletions
Original file line numberDiff line numberDiff line change
@@ -11,6 +11,7 @@
1111
from .parser import Parser
1212
from .storage import BaseStorage
1313
from .utils import ProxyPool, Session, Signal
14+
from . import defaults
1415

1516

1617
class Crawler:
@@ -133,14 +134,7 @@ def set_session(self, headers=None):
133134
header to init the session)
134135
"""
135136
if headers is None:
136-
headers = {
137-
"Accept-Language": "zh-CN,zh;q=0.8,zh-TW;q=0.7,zh-HK;q=0.5,en-US;q=0.3,en;q=0.2",
138-
"User-Agent": (
139-
"Mozilla/5.0 (Windows NT 10.0; Win64; x64)"
140-
" AppleWebKit/537.36 (KHTML, like Gecko) "
141-
"Chrome/88.0.4324.104 Safari/537.36"
142-
),
143-
}
137+
headers = defaults.DEFAULT_HEADERS
144138
elif not isinstance(headers, dict):
145139
raise TypeError('"headers" must be a dict object')
146140

icrawler/defaults.py

Lines changed: 7 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,7 @@
1+
ACCEPT_LANGUAGES = "zh-CN,zh;q=0.8,zh-TW;q=0.7,zh-HK;q=0.5,en-US;q=0.3,en;q=0.2"
2+
USER_AGENT = "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/134.0.0.0 Safari/537.36"
3+
4+
DEFAULT_HEADERS = {
5+
"Accept-Language": ACCEPT_LANGUAGES,
6+
"User-Agent": USER_AGENT,
7+
}

0 commit comments

Comments
 (0)