Skip to content
Projects
Groups
Snippets
Help
Loading...
Help
Submit feedback
Sign in / Register
Toggle navigation
A
aidso-data
Project
Project
Details
Activity
Releases
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
Yaowentong
aidso-data
Commits
1236b9ff
Commit
1236b9ff
authored
Jul 13, 2026
by
Yaowentong
Browse files
Options
Browse Files
Download
Email Patches
Plain Diff
豆包安卓修复
parent
66670d85
Changes
7
Expand all
Hide whitespace changes
Inline
Side-by-side
Showing
7 changed files
with
4065 additions
and
306 deletions
+4065
-306
doubao_android_data_process.py
aidso_geo/models/doubao_android_data_process.py
+5
-9
process.py
aidso_geo/models/process.py
+3
-3
qianwen_android_data_process.py
aidso_geo/models/qianwen_android_data_process.py
+1
-1
xiaohongshu_android_data_process.py
aidso_geo/models/xiaohongshu_android_data_process.py
+1
-2
yuanbao_android_data_process.py
aidso_geo/models/yuanbao_android_data_process.py
+2
-2
1.json
third_live/1.json
+68
-187
2.json
third_live/2.json
+3985
-102
No files found.
aidso_geo/models/doubao_android_data_process.py
View file @
1236b9ff
...
@@ -37,15 +37,11 @@ def doubao_mobile_process_original_data(task_data):
...
@@ -37,15 +37,11 @@ def doubao_mobile_process_original_data(task_data):
try
:
try
:
json_content
=
json
.
loads
(
payload
)
json_content
=
json
.
loads
(
payload
)
print
(
json_content
)
print
(
'-----'
)
print
(
'-----'
)
print
(
'-----'
)
print
(
'-----'
)
print
(
'-----'
)
except
(
IndexError
,
json
.
JSONDecodeError
):
except
(
IndexError
,
json
.
JSONDecodeError
):
continue
continue
if
not
isinstance
(
json_content
,
dict
):
continue
if
json_content
.
get
(
'content'
):
if
json_content
.
get
(
'content'
):
if
json_content
.
get
(
'content'
)
.
get
(
'content_block'
):
if
json_content
.
get
(
'content'
)
.
get
(
'content_block'
):
if
json_content
.
get
(
'content'
)
.
get
(
'content_block'
)[
0
]
.
get
(
'block_type'
)
==
10040
:
if
json_content
.
get
(
'content'
)
.
get
(
'content_block'
)[
0
]
.
get
(
'block_type'
)
==
10040
:
...
@@ -315,7 +311,7 @@ def doubao_mobile_process_original_data(task_data):
...
@@ -315,7 +311,7 @@ def doubao_mobile_process_original_data(task_data):
suggestions
,
rich_media_block
)
suggestions
,
rich_media_block
)
return
(
file_path
,
search_keyword
,
url_list
,
think_content
,
response_content
,
suggestions
)
return
(
file_path
,
search_keyword
,
url_list
,
think_content
,
response_content
,
suggestions
)
except
Exception
as
e
:
except
Exception
as
e
:
# traceback.print_exc()
parts
=
file_path
.
split
(
'/'
)
parts
=
file_path
.
split
(
'/'
)
platform
=
parts
[
2
]
platform
=
parts
[
2
]
task_id
=
parts
[
1
]
task_id
=
parts
[
1
]
...
@@ -348,7 +344,7 @@ if __name__ == '__main__':
...
@@ -348,7 +344,7 @@ if __name__ == '__main__':
# file_path = 'geo/c7eb465e-f385-4aa2-89c4-a7cf11897f45/KIMI/1(1).txt'
# file_path = 'geo/c7eb465e-f385-4aa2-89c4-a7cf11897f45/KIMI/1(1).txt'
data_list
=
bh_utils
.
query_data
(
data_list
=
bh_utils
.
query_data
(
f
"select * from geo_commit_task where reqId = '
0d408a82-c850-4796-9fa3-de5af109b2e4
'"
)
f
"select * from geo_commit_task where reqId = '
f6599e25c11544e69f911b53aff98ce7
'"
)
...
...
aidso_geo/models/process.py
View file @
1236b9ff
...
@@ -1995,7 +1995,7 @@ def run_data(PAGE_SIZE,MAX_WORKERS):
...
@@ -1995,7 +1995,7 @@ def run_data(PAGE_SIZE,MAX_WORKERS):
if
__name__
==
'__main__'
:
if
__name__
==
'__main__'
:
#
#
data_list
=
bh_utils
.
query_data
(
f
"select * from geo_commit_task where
reqId='df6b3f7850eb4a538843952f566f0b58
'"
)
data_list
=
bh_utils
.
query_data
(
f
"select * from geo_commit_task where
status='ING
'"
)
def
handle_item
(
i
):
def
handle_item
(
i
):
if
i
.
get
(
'comWordsMap'
):
if
i
.
get
(
'comWordsMap'
):
...
@@ -2011,8 +2011,8 @@ if __name__ == '__main__':
...
@@ -2011,8 +2011,8 @@ if __name__ == '__main__':
type_t
=
i
.
get
(
'type'
)
type_t
=
i
.
get
(
'type'
)
type_t
=
'batch'
type_t
=
'batch'
# commit_task(i,'ING')
# commit_task(i,'ING')
#
return task_send_queue(i,type_t)
return
task_send_queue
(
i
,
type_t
)
return
platform_process
(
i
)
#
return platform_process(i)
#
#
if
data_list
:
if
data_list
:
with
ThreadPoolExecutor
(
max_workers
=
30
)
as
executor
:
with
ThreadPoolExecutor
(
max_workers
=
30
)
as
executor
:
...
...
aidso_geo/models/qianwen_android_data_process.py
View file @
1236b9ff
...
@@ -324,7 +324,7 @@ def qianwen_android_process_original_data(task_data):
...
@@ -324,7 +324,7 @@ def qianwen_android_process_original_data(task_data):
response_content
+=
'render_ecom_card_widget_gaode_start:'
response_content
+=
'render_ecom_card_widget_gaode_start:'
gaode_str
=
json
.
dumps
(
gaode_list
,
ensure_ascii
=
False
)
gaode_str
=
json
.
dumps
(
gaode_list
,
ensure_ascii
=
False
)
response_content
+=
gaode_str
response_content
+=
gaode_str
response_content
+=
'render_ecom_card_widget_gaod_end:
\n
'
response_content
+=
'render_ecom_card_widget_gaod
e
_end:
\n
'
if
mime_type
==
'multi_load/iframe'
and
multi_load_type
==
'deep_think'
and
multi_load_status
==
'complete'
:
if
mime_type
==
'multi_load/iframe'
and
multi_load_type
==
'deep_think'
and
multi_load_status
==
'complete'
:
think_content
+=
multi_load_content
.
get
(
'think_content'
)
think_content
+=
multi_load_content
.
get
(
'think_content'
)
# 引用来源 回答 视频列表
# 引用来源 回答 视频列表
...
...
aidso_geo/models/xiaohongshu_android_data_process.py
View file @
1236b9ff
...
@@ -65,7 +65,7 @@ if __name__ == '__main__':
...
@@ -65,7 +65,7 @@ if __name__ == '__main__':
# xiaohongshu_android_process_original_data(file_path2)
# xiaohongshu_android_process_original_data(file_path2)
data_list
=
bh_utils
.
query_data
(
f
"select * from geo_commit_task where platform = 'XHSA' and taskId = '
aeb60818-5221-4062-820d-e4523f184d56
'"
)
data_list
=
bh_utils
.
query_data
(
f
"select * from geo_commit_task where platform = 'XHSA' and taskId = '
74866a04-8cc0-4c95-be49-836353f9f7bb
'"
)
# # #
# # #
# # #
# # #
...
@@ -81,7 +81,6 @@ if __name__ == '__main__':
...
@@ -81,7 +81,6 @@ if __name__ == '__main__':
i
[
'keywords'
]
=
json
.
loads
(
i
.
get
(
'keywords'
))
i
[
'keywords'
]
=
json
.
loads
(
i
.
get
(
'keywords'
))
if
i
.
get
(
'productWordsMap'
):
if
i
.
get
(
'productWordsMap'
):
i
[
'productWordsMap'
]
=
json
.
loads
(
i
.
get
(
'productWordsMap'
))
i
[
'productWordsMap'
]
=
json
.
loads
(
i
.
get
(
'productWordsMap'
))
print
(
i
.
get
(
'taskId'
))
return
xiaohongshu_android_process_original_data
(
i
)
return
xiaohongshu_android_process_original_data
(
i
)
# 喜欢非遗手作的看这里!周末到河北之涿州!
# 喜欢非遗手作的看这里!周末到河北之涿州!
...
...
aidso_geo/models/yuanbao_android_data_process.py
View file @
1236b9ff
...
@@ -47,6 +47,7 @@ def yuanbao_android_process_original_data(task_data):
...
@@ -47,6 +47,7 @@ def yuanbao_android_process_original_data(task_data):
# 提取并解析JSON数据
# 提取并解析JSON数据
data_str
=
i
.
split
(
"data: "
)[
1
]
data_str
=
i
.
split
(
"data: "
)[
1
]
json_data
=
json
.
loads
(
data_str
)
json_data
=
json
.
loads
(
data_str
)
except
(
IndexError
,
json
.
JSONDecodeError
):
except
(
IndexError
,
json
.
JSONDecodeError
):
continue
# 跳过格式错误的数据
continue
# 跳过格式错误的数据
if
json_data
.
get
(
'type'
)
==
'searchGuid'
:
if
json_data
.
get
(
'type'
)
==
'searchGuid'
:
...
@@ -146,7 +147,6 @@ def yuanbao_android_process_original_data(task_data):
...
@@ -146,7 +147,6 @@ def yuanbao_android_process_original_data(task_data):
suggestions
,
rich_media_block
)
suggestions
,
rich_media_block
)
return
(
file_path
,
search_keyword
,
url_list
,
think_content
,
response_content
,
suggestions
)
return
(
file_path
,
search_keyword
,
url_list
,
think_content
,
response_content
,
suggestions
)
except
Exception
as
e
:
except
Exception
as
e
:
parts
=
file_path
.
split
(
'/'
)
parts
=
file_path
.
split
(
'/'
)
platform
=
parts
[
2
]
platform
=
parts
[
2
]
task_id
=
parts
[
1
]
task_id
=
parts
[
1
]
...
@@ -175,7 +175,7 @@ if __name__ == '__main__':
...
@@ -175,7 +175,7 @@ if __name__ == '__main__':
# yuanbao_android_process_original_data(file_path2)
# yuanbao_android_process_original_data(file_path2)
data_list
=
bh_utils
.
query_data
(
data_list
=
bh_utils
.
query_data
(
f
"select * from geo_commit_task where taskId = '2
e35cfd9-6c36-417c-9384-9140d9a7dcd1
' and platform = 'TXYBA'"
)
f
"select * from geo_commit_task where taskId = '2
a7fca28-5389-4506-ad2a-76c222fe424e
' and platform = 'TXYBA'"
)
# # #
# # #
...
...
third_live/1.json
View file @
1236b9ff
This diff is collapsed.
Click to expand it.
third_live/2.json
View file @
1236b9ff
This source diff could not be displayed because it is too large. You can
view the blob
instead.
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