From fb8ac287a9c871cfd4800570ce7d9c8693821d6b Mon Sep 17 00:00:00 2001
From: baoliang <j2eer@163.com>
Date: Tue, 26 Jan 2021 12:02:02 +0800
Subject: [PATCH 1/2] add gansudaily html parse

---
 .DS_Store      | Bin 10244 -> 10244 bytes
 polls/utils.py |  17 ++++++++++-------
 2 files changed, 10 insertions(+), 7 deletions(-)

diff --git a/.DS_Store b/.DS_Store
index afafaf25f0e3fcfd0a5f5bb8815286219469082b..564a53de6819fc754945c51b1b276a86c0231e37 100644
GIT binary patch
delta 339
zcmZn(XbIS0Aiz|9aB_@*#AGi4E;g5g9^TwnCr1gWz?dxps!T^1Ca)G)&eWSUxj|5B
za-JX;TjD-1kh(fSH3U;?^ASNEMs;DyQw$6Y><qaK`3!jsnG7XBx+pi_#U&{xKM5$q
z!7b_w(km{kzy@X73M(>+I806uHfORioSZ1EA%d*FdZr6d{UJw?#N@xiJZunw$^0T=
zE}t@hCbBXV0PV;D!eV57_I8`*6#(_2t4M*n1fpW{N)b86g3TvHZnCLW$${LSx*=+k
qe{kBD)D4YN={m?}HnInT+<shIfep%hE3L%%U}NET=FRL1f7t;YKW9Y%

delta 339
zcmZn(XbIS0Ai(4|dvc6`#AGi4F1E*ST7{PEog5{g0%Nuas4^YCFnP7Wa;CPX$qj;1
zlk)_**fz<xfz;Irsv($Cn~w<UFsk!C3}RqlU}wl>$Y;o7$Ydx1(nY!XE-pzq`AI+-
zj?#-S%pdKWEH13T24&g`D>4ZvOimCsXR>}cIZ;?c1X;cK^Ny181BV<z5|jT5^RPh#
zCi9Dgxg2{BG?A5|0BA=J5Ediri%pGvrL*&pBf5$dxJw`^Ca)BcV|3VjLgXf!8sArt
y+fz40P4W*;`;xk$Q7T;r*~~TEQ}6j7nS5MYfep%hE3L$+u(9wv^JaF1zw7{0N^|r8

diff --git a/polls/utils.py b/polls/utils.py
index 487dcbf..d92d3b0 100644
--- a/polls/utils.py
+++ b/polls/utils.py
@@ -97,8 +97,11 @@ def parse(url):
         "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8",
     }
     r = requests.get(url, headers=headers)
-    r.encoding = 'utf-8'
-    print(r.text)
+    if r.encoding=='ISO-8859-1':
+        r.encoding='GBK'
+    else:
+        r.encoding = 'utf-8'
+    # print(r.text)
     selector = Selector(text=r.text)
     if t == 'weixin':
         og_title = selector.xpath(
@@ -162,10 +165,10 @@ if __name__ == '__main__':
     #     'https://m.toutiao.com/i6883651337003729420/?tt_from=weixin&utm_campaign=client_share&app=news_article&utm_source=weixin&iid=1494959660475024&utm_medium=toutiao_android&wxshare_count=1')
     # og_title, og_description, og_url, og_image = parse(
     #     'http://www.gov.cn/xinwen/2020-10/13/content_5550906.htm')
-    # og_title, og_description, og_url, og_image = parse(
-    #     'http://xgs.gansudaily.com.cn/pages/h5/hot/b3297046a53e47f594ed19db90c1183c.html')
+    og_title, og_description, og_url, og_image = parse(
+        'http://gansu.gansudaily.com.cn/system/2021/01/25/030261998.shtml')
 
-    # print(og_title, og_description, og_url, og_image)
+    print(og_title, og_description, og_url, og_image)
     # print(send_voice_notify('13993199566'))
-    r = send_tnps(['13609346975'], '')
-    print(r.status_code, r.text)
+    # r = send_tnps(['13609346975'], '')
+    # print(r.status_code, r.text)

From e2f0507e68fda4f53606e7baf820b0132d1f2e4f Mon Sep 17 00:00:00 2001
From: baoliang <j2eer@163.com>
Date: Tue, 26 Jan 2021 12:23:57 +0800
Subject: [PATCH 2/2] fix gansudaily econding bug

---
 polls/utils.py | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/polls/utils.py b/polls/utils.py
index d92d3b0..2e2ab57 100644
--- a/polls/utils.py
+++ b/polls/utils.py
@@ -97,11 +97,13 @@ def parse(url):
         "Accept": "text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8",
     }
     r = requests.get(url, headers=headers)
-    if r.encoding=='ISO-8859-1':
+    print(r.encoding)
+
+    if t == 'xgs' and r.encoding=='ISO-8859-1':
         r.encoding='GBK'
     else:
         r.encoding = 'utf-8'
-    # print(r.text)
+    # way2@8797799
     selector = Selector(text=r.text)
     if t == 'weixin':
         og_title = selector.xpath(