Skip to content
Projects
Groups
Snippets
Help
Loading...
Sign in
Toggle navigation
C
crawler
Project
Project
Details
Activity
Cycle Analytics
Repository
Repository
Files
Commits
Branches
Tags
Contributors
Graph
Compare
Charts
Issues
0
Issues
0
List
Board
Labels
Milestones
Merge Requests
0
Merge Requests
0
CI / CD
CI / CD
Pipelines
Jobs
Schedules
Charts
Wiki
Wiki
Snippets
Snippets
Members
Members
Collapse sidebar
Close sidebar
Activity
Graph
Charts
Create a new issue
Jobs
Commits
Issue Boards
Open sidebar
backend
crawler
Commits
4309e098
Commit
4309e098
authored
Aug 20, 2021
by
李小芳
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
update
parent
4d1d6e83
Expand all
Hide whitespace changes
Inline
Side-by-side
Showing
9 changed files
with
111 additions
and
48 deletions
+111
-48
csv-plugin.xml
.idea/csv-plugin.xml
+3
-3
encodings.xml
.idea/encodings.xml
+7
-0
cika.py
dev/xinyang_ask_tag/cika.py
+51
-9
cika_to_csv.py
dev/xinyang_ask_tag/cika_to_csv.py
+0
-36
city_parse.py
dev/xinyang_ask_tag/city_parse.py
+0
-0
hospital.py
dev/xinyang_ask_tag/hospital.py
+0
-0
soyoung_service_cika.csv
dev/xinyang_ask_tag/soyoung_service_cika.csv
+0
-0
医院尾款劵数据.csv
dev/xinyang_ask_tag/医院尾款劵数据.csv
+50
-0
更美机构商务信息.csv
dev/xinyang_ask_tag/更美机构商务信息.csv
+0
-0
No files found.
.idea/csv-plugin.xml
View file @
4309e098
...
@@ -17,21 +17,21 @@
...
@@ -17,21 +17,21 @@
</Attribute>
</Attribute>
</value>
</value>
</entry>
</entry>
<entry
key=
"/dev/xinyang_ask_tag/soyoung_service_
1
.csv"
>
<entry
key=
"/dev/xinyang_ask_tag/soyoung_service_
cika
.csv"
>
<value>
<value>
<Attribute>
<Attribute>
<option
name=
"separator"
value=
","
/>
<option
name=
"separator"
value=
","
/>
</Attribute>
</Attribute>
</value>
</value>
</entry>
</entry>
<entry
key=
"/dev/xinyang_ask_tag/
soyoung_service_cika
.csv"
>
<entry
key=
"/dev/xinyang_ask_tag/
医院尾款劵数据
.csv"
>
<value>
<value>
<Attribute>
<Attribute>
<option
name=
"separator"
value=
","
/>
<option
name=
"separator"
value=
","
/>
</Attribute>
</Attribute>
</value>
</value>
</entry>
</entry>
<entry
key=
"/dev/xinyang_ask_tag/
soyoung_service_write_cika
.csv"
>
<entry
key=
"/dev/xinyang_ask_tag/
更美机构商务信息
.csv"
>
<value>
<value>
<Attribute>
<Attribute>
<option
name=
"separator"
value=
","
/>
<option
name=
"separator"
value=
","
/>
...
...
.idea/encodings.xml
0 → 100644
View file @
4309e098
<?xml version="1.0" encoding="UTF-8"?>
<project
version=
"4"
>
<component
name=
"Encoding"
>
<file
url=
"file://$PROJECT_DIR$/dev/xinyang_ask_tag/更美机构商务信息.csv"
charset=
"UTF-8"
/>
</component>
</project>
\ No newline at end of file
dev/xinyang_ask_tag/
test
.py
→
dev/xinyang_ask_tag/
cika
.py
View file @
4309e098
...
@@ -5,12 +5,13 @@ import re
...
@@ -5,12 +5,13 @@ import re
from
pypinyin
import
lazy_pinyin
from
pypinyin
import
lazy_pinyin
import
time
import
time
import
random
import
random
import
csv
monkey
.
patch_all
()
monkey
.
patch_all
()
def
get_service_info
(
spu_id
,
url
,
ids_bug
):
def
get_service_info
(
spu_id
,
url
,
ids_bug
):
time
.
sleep
(
random
.
uniform
(
1
,
2
))
time
.
sleep
(
random
.
uniform
(
2
,
3
))
url
=
url
+
"?tab=0&rj=0"
url
=
url
+
"?tab=0&rj=0"
headers
=
{
headers
=
{
...
@@ -56,7 +57,7 @@ def gevent_test():
...
@@ -56,7 +57,7 @@ def gevent_test():
gevent
.
joinall
(
tasks
)
gevent
.
joinall
(
tasks
)
def
get_cika_info_to_csv
():
def
get_cika_info_to_csv
(
get_data
=
False
):
read_file
=
open
(
"soyoung_service.csv"
,
"r"
,
encoding
=
"utf-8"
)
read_file
=
open
(
"soyoung_service.csv"
,
"r"
,
encoding
=
"utf-8"
)
write_file
=
open
(
"soyoung_service_cika.csv"
,
"a+"
,
encoding
=
"utf-8"
)
write_file
=
open
(
"soyoung_service_cika.csv"
,
"a+"
,
encoding
=
"utf-8"
)
all_cika_title
=
list
()
all_cika_title
=
list
()
...
@@ -66,7 +67,7 @@ def get_cika_info_to_csv():
...
@@ -66,7 +67,7 @@ def get_cika_info_to_csv():
try
:
try
:
cika_price_dict
=
dict
()
cika_price_dict
=
dict
()
print
(
"index:"
,
item
.
strip
()
.
split
(
","
)[
0
])
print
(
"index:"
,
item
.
strip
()
.
split
(
","
)[
0
])
if
int
(
item
.
strip
()
.
split
(
","
)[
0
])
>
34645
:
if
int
(
item
.
strip
()
.
split
(
","
)[
0
])
>
64526
:
service_id
=
item
.
strip
()
.
split
(
","
)[
2
]
service_id
=
item
.
strip
()
.
split
(
","
)[
2
]
url
=
item
.
strip
()
.
split
(
","
)[
-
1
]
url
=
item
.
strip
()
.
split
(
","
)[
-
1
]
if
service_id
in
have_read_service
:
if
service_id
in
have_read_service
:
...
@@ -74,12 +75,16 @@ def get_cika_info_to_csv():
...
@@ -74,12 +75,16 @@ def get_cika_info_to_csv():
else
:
else
:
have_read_service
.
append
(
service_id
)
have_read_service
.
append
(
service_id
)
sub_title_dict
=
get_service_info
(
str
(
service_id
),
url
,
ids_bug
)
sub_title_dict
=
get_service_info
(
str
(
service_id
),
url
,
ids_bug
)
if
sub_title_dict
:
if
get_data
==
False
:
all_cika_title
.
extend
(
sub_title_dict
.
keys
())
if
sub_title_dict
:
cika_price_dict
[
service_id
]
=
sub_title_dict
all_cika_title
.
extend
(
sub_title_dict
.
keys
())
print
(
cika_price_dict
)
cika_price_dict
[
service_id
]
=
sub_title_dict
write_file
.
write
(
str
(
cika_price_dict
))
print
(
cika_price_dict
)
write_file
.
write
(
"
\n
"
)
write_file
.
write
(
str
(
cika_price_dict
))
write_file
.
write
(
"
\n
"
)
else
:
return
sub_title_dict
except
Exception
as
e
:
except
Exception
as
e
:
print
(
e
)
print
(
e
)
pass
pass
...
@@ -89,5 +94,42 @@ def get_cika_info_to_csv():
...
@@ -89,5 +94,42 @@ def get_cika_info_to_csv():
write_file
.
close
()
write_file
.
close
()
def
read_cika_info
():
all_cika_title
=
{}
cika_file
=
open
(
"soyoung_service_cika.csv"
,
"r"
,
encoding
=
"utf-8"
)
for
item
in
cika_file
.
readlines
():
cika
=
eval
(
item
)
for
key
,
values
in
cika
.
items
():
all_cika_title
[
key
]
=
values
print
(
all_cika_title
)
cika_file
.
close
()
return
all_cika_title
def
np_write_csv_data
():
all_cika_title
=
read_cika_info
()
write_file
=
open
(
"soyoung_service_write_cika.csv"
,
"a+"
,
encoding
=
"utf-8"
)
# /Users/edz/Desktop/xinyang.csv
with
open
(
"soyoung_service.csv"
,
encoding
=
'utf-8'
)
as
f
:
reader
=
csv
.
reader
(
f
)
header_row
=
next
(
reader
)
print
(
header_row
)
for
row
in
reader
:
service_id
=
row
[
2
]
cika_info
=
all_cika_title
.
get
(
service_id
)
if
cika_info
:
print
(
'cika_info:'
,
cika_info
)
write_file
.
write
(
str
(
row
.
append
(
cika_info
)))
write_file
.
write
(
"
\n
"
)
else
:
sub_title_dict
=
get_cika_info_to_csv
(
get_data
=
True
)
print
(
'sub_title_dict:'
,
sub_title_dict
)
write_file
.
write
(
str
(
row
.
append
(
sub_title_dict
)))
write_file
.
write
(
"
\n
"
)
write_file
.
close
()
if
__name__
==
'__main__'
:
if
__name__
==
'__main__'
:
get_cika_info_to_csv
()
get_cika_info_to_csv
()
dev/xinyang_ask_tag/cika_to_csv.py
deleted
100644 → 0
View file @
4d1d6e83
import
csv
def
read_cika_info
():
all_cika_title
=
{}
cika_file
=
open
(
"soyoung_service_cika.csv"
,
"r"
,
encoding
=
"utf-8"
)
for
item
in
cika_file
.
readlines
():
cika
=
eval
(
item
)
for
key
,
values
in
cika
.
items
():
all_cika_title
[
key
]
=
values
print
(
all_cika_title
)
cika_file
.
close
()
return
all_cika_title
def
np_write_csv_data
():
all_cika_title
=
read_cika_info
()
write_file
=
open
(
"soyoung_service_write_cika.csv"
,
"a+"
,
encoding
=
"utf-8"
)
# /Users/edz/Desktop/xinyang.csv
with
open
(
"soyoung_service.csv"
,
encoding
=
'utf-8'
)
as
f
:
reader
=
csv
.
reader
(
f
)
header_row
=
next
(
reader
)
print
(
header_row
)
for
row
in
reader
:
service_id
=
row
[
2
]
cika_info
=
all_cika_title
.
get
(
service_id
)
if
cika_info
:
print
(
'cika_info:'
,
cika_info
)
write_file
.
write
(
str
(
row
.
append
(
cika_info
)))
write_file
.
write
(
"
\n
"
)
write_file
.
close
()
if
__name__
==
'__main__'
:
np_write_csv_data
()
dev/xinyang_ask_tag/city_parse.py
View file @
4309e098
This diff is collapsed.
Click to expand it.
dev/xinyang_ask_tag/hospital.py
0 → 100644
View file @
4309e098
This diff is collapsed.
Click to expand it.
dev/xinyang_ask_tag/soyoung_service_cika.csv
View file @
4309e098
This diff is collapsed.
Click to expand it.
dev/xinyang_ask_tag/医院尾款劵数据.csv
0 → 100644
View file @
4309e098
"大区", "城市", "商务", "机构名称", "满", "减", "劣势/缺失", "到期时间", "查到劣势日期", "最晚补券时间"
,北京,,北京好年华眼整形中心,10000,1000,,2027-02-28 到期,2021-08-20 17:36:56,
,北京,,北京好年华眼整形中心,100,5,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京好年华眼整形中心,500,30,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京清木医疗美容诊所,2000,100,,2023-03-31 到期,2021-08-20 17:36:56,
,北京,,北京清木医疗美容诊所,10000,300,,2022-06-30 到期,2021-08-20 17:36:56,
,北京,,北京清木医疗美容诊所,999,50,,2023-04-30 到期,2021-08-20 17:36:56,
,北京,,北京清木医疗美容诊所,2000,100,,2024-04-30 到期,2021-08-20 17:36:56,
,北京,,北京清木医疗美容诊所,300,30,,2024-04-30 到期,2021-08-20 17:36:56,
,北京,,美诗沁世茂医疗美容,20,20,,2022-08-31 到期,2021-08-20 17:36:56,
,北京,,美诗沁世茂医疗美容,500,50,,2022-08-31 到期,2021-08-20 17:36:56,
,北京,,美诗沁世茂医疗美容,5000,600,,2022-08-31 到期,2021-08-20 17:36:56,
,北京,,北京新星靓医疗美容医院,3000,500,,2026-03-31 到期,2021-08-20 17:36:56,
,北京,,北京新星靓医疗美容医院,223,222,,2024-04-26 到期,2021-08-20 17:36:56,
,北京,,北京新星靓医疗美容医院,9999,2000,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京新星靓医疗美容医院,9999,2000,,2025-05-31 到期,2021-08-20 17:36:56,
,北京,,北京新星靓医疗美容医院,9999,2000,,2025-05-31 到期,2021-08-20 17:36:56,
,北京,,北京新星靓医疗美容医院,2000,1000,,2025-05-31 到期,2021-08-20 17:36:56,
,北京,,北京新星靓医疗美容医院,9999,2000,,2028-05-31 到期,2021-08-20 17:36:56,
,北京,,北京新星靓医疗美容医院,1000,100,,2028-06-30 到期,2021-08-20 17:36:56,
,北京,,北京新星靓医疗美容医院,2000,200,,2024-06-28 到期,2021-08-20 17:36:56,
,北京,,北京润美玉之光医疗美容,1000,50,,2021-12-31 到期,2021-08-20 17:36:56,
,北京,,北京润美玉之光医疗美容,3000,150,,2021-12-31 到期,2021-08-20 17:36:56,
,北京,,北京润美玉之光医疗美容,5000,300,,2021-12-31 到期,2021-08-20 17:36:56,
,北京,,北京润美玉之光医疗美容,10000,500,,2021-12-31 到期,2021-08-20 17:36:56,
,北京,,北京叶子整形美容医院,1000,100,,2021-09-26 到期,2021-08-20 17:36:56,
,北京,,北京亚楠容悦医疗美容,100,8,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京亚楠容悦医疗美容,3000,100,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京画美医疗美容医院,10000,1000,,2022-07-31 到期,2021-08-20 17:36:56,
,北京,,北京画美医疗美容医院,5000,500,,2022-07-31 到期,2021-08-20 17:36:56,
,北京,,北京画美医疗美容医院,1500,200,,2022-07-31 到期,2021-08-20 17:36:56,
,北京,,北京画美医疗美容医院,500,50,,2022-07-31 到期,2021-08-20 17:36:56,
,北京,,北京爱悦丽格医疗美容,500,50,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京爱悦丽格医疗美容,1000,100,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,上上相整形·脂肪骨性移植,15000,350,,2021-10-31 到期,2021-08-20 17:36:56,
,北京,,上上相整形·脂肪骨性移植,30000,700,,2021-10-31 到期,2021-08-20 17:36:56,
,北京,,北京嘉禾医疗美容,500,50,,2022-06-30 到期,2021-08-20 17:36:56,
,北京,,北京嘉禾医疗美容,999,100,,2022-06-30 到期,2021-08-20 17:36:56,
,北京,,北京嘉禾医疗美容,1999,200,,2022-08-31 到期,2021-08-20 17:36:56,
,北京,,北京嘉禾医疗美容,4999,500,,2022-08-31 到期,2021-08-20 17:36:56,
,北京,,北京嘉禾医疗美容,19999,2000,,2022-08-31 到期,2021-08-20 17:36:56,
,北京,,北京雅靓医疗美容,1000,100,,2021-12-31 到期,2021-08-20 17:36:56,
,北京,,北京俊泰美好医疗美容门诊部(鼻修复及不明注射物取出中心),800,100,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京俊泰美好医疗美容门诊部(鼻修复及不明注射物取出中心),500,50,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京华悦府医疗美容,999,50,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京华悦府医疗美容,3980,2000,,2021-10-31 到期,2021-08-20 17:36:56,
,北京,,北京艺星医疗美容医院,10999,500,,2021-08-31 到期,2021-08-20 17:36:56,
,北京,,北京艺星医疗美容医院,1999,200,,2021-10-31 到期,2021-08-20 17:36:56,
,北京,,北京艺星医疗美容医院,4000,300,,2021-10-31 到期,2021-08-20 17:36:56,
,北京,,北京艺星医疗美容医院,8000,800,,2021-10-31 到期,2021-08-20 17:36:56,
dev/xinyang_ask_tag/更美机构商务信息.csv
0 → 100644
View file @
4309e098
This diff is collapsed.
Click to expand it.
Write
Preview
Markdown
is supported
0%
Try again
or
attach a new file
Attach a file
Cancel
You are about to add
0
people
to the discussion. Proceed with caution.
Finish editing this message first!
Cancel
Please
register
or
sign in
to comment