Skip to content
Toggle navigation
Toggle navigation
This project
Loading...
Sign in
周伟奇
/
bmw-ocr
Go to a project
Toggle navigation
Toggle navigation pinning
Projects
Groups
Snippets
Help
Project
Activity
Repository
Graphs
Network
Create a new issue
Commits
Issue Boards
Files
Commits
Network
Compare
Branches
Tags
3bdf0e45
authored
2020-11-08 16:12:54 +0800
by
周伟奇
Browse Files
Options
Browse Files
Tag
Download
Email Patches
Plain Diff
fix bug
1 parent
fc8f7e0d
Hide whitespace changes
Inline
Side-by-side
Showing
2 changed files
with
17 additions
and
16 deletions
src/apps/doc/management/commands/ocr_process.py
src/apps/doc/ocr/wb.py
src/apps/doc/management/commands/ocr_process.py
View file @
3bdf0e4
...
...
@@ -7,6 +7,7 @@ import asyncio
import
aiohttp
import
difflib
import
requests
import
traceback
from
collections
import
Counter
from
datetime
import
datetime
,
date
from
django.utils
import
timezone
...
...
@@ -93,7 +94,7 @@ class Command(BaseCommand, LoggerMixin):
self
.
edms
.
download
(
pdf_path
,
doc
.
metadata_version_id
)
except
Exception
as
e
:
self
.
cronjob_log
.
warn
(
'{0} [edms download failed] [times={1}] [pdf_path={2}] '
'[error={3}]'
.
format
(
self
.
log_base
,
times
,
pdf_path
,
e
))
'[error={3}]'
.
format
(
self
.
log_base
,
times
,
pdf_path
,
traceback
.
format_exc
()
))
edms_exc
=
str
(
e
)
else
:
break
...
...
@@ -398,13 +399,13 @@ class Command(BaseCommand, LoggerMixin):
except
EDMSException
as
e
:
doc
.
status
=
DocStatus
.
PROCESS_FAILED
.
value
doc
.
save
()
self
.
cronjob_log
.
error
(
'{0} [process failed (edms download)] [task={1}] [err={2}]'
.
format
(
self
.
log_base
,
task_str
,
e
))
self
.
cronjob_log
.
error
(
'{0} [process failed (edms download)] [task={1}] [err
or
={2}]'
.
format
(
self
.
log_base
,
task_str
,
traceback
.
format_exc
()
))
except
Exception
as
e
:
doc
.
status
=
DocStatus
.
PROCESS_FAILED
.
value
doc
.
save
()
self
.
cronjob_log
.
error
(
'{0} [process failed (pdf to img)] [task={1}] [err={2}]'
.
format
(
self
.
log_base
,
task_str
,
e
))
self
.
cronjob_log
.
error
(
'{0} [process failed (pdf to img)] [task={1}] [err
or
={2}]'
.
format
(
self
.
log_base
,
task_str
,
traceback
.
format_exc
()
))
def
img_2_ocr_1
(
self
,
img_queue
,
todo_count_dict
,
res_dict
,
finish_queue
,
lock
,
url
):
while
True
:
...
...
@@ -432,7 +433,7 @@ class Command(BaseCommand, LoggerMixin):
raise
OCR1Exception
(
'ocr_1 status code: {0}'
.
format
(
ocr_1_response
.
status_code
))
except
Exception
as
e
:
self
.
cronjob_log
.
warn
(
'{0} [ocr_1 failed] [times={1}] [img_path={2}] [error={3}]'
.
format
(
self
.
log_base
,
times
,
img_path
,
e
))
self
.
log_base
,
times
,
img_path
,
traceback
.
format_exc
()
))
else
:
ocr_1_res
=
ocr_1_response
.
json
()
end_time
=
time
.
time
()
...
...
@@ -525,7 +526,7 @@ class Command(BaseCommand, LoggerMixin):
except
Exception
as
e
:
self
.
cronjob_log
.
warn
(
'{0} [ocr_2 failed] [times={1}] [img_path={2}] [error={3}]'
.
format
(
self
.
log_base
,
times
,
img_path
,
e
))
self
.
log_base
,
times
,
img_path
,
traceback
.
format_exc
()
))
else
:
ocr_2_res
=
json
.
loads
(
ocr_2_response
.
text
)
end_time
=
time
.
time
()
...
...
@@ -591,8 +592,8 @@ class Command(BaseCommand, LoggerMixin):
doc
,
_
=
self
.
get_doc_object
(
task_str
)
doc
.
status
=
DocStatus
.
PROCESS_FAILED
.
value
doc
.
save
()
self
.
cronjob_log
.
error
(
'{0} [process failed (res to wb)] [task={1}] [err={2}]'
.
format
(
self
.
log_base
,
task_str
,
e
))
self
.
cronjob_log
.
error
(
'{0} [process failed (res to wb)] [task={1}] [err
or
={2}]'
.
format
(
self
.
log_base
,
task_str
,
traceback
.
format_exc
()
))
else
:
try
:
# 5.上传至EDMS
...
...
@@ -602,7 +603,7 @@ class Command(BaseCommand, LoggerMixin):
except
Exception
as
e
:
self
.
cronjob_log
.
warn
(
'{0} [edms upload failed] [times={1}] [task={2}] [error={3}]'
.
format
(
self
.
log_base
,
times
,
task_str
,
e
))
self
.
log_base
,
times
,
task_str
,
traceback
.
format_exc
()
))
edms_exc
=
str
(
e
)
else
:
break
...
...
@@ -616,8 +617,8 @@ class Command(BaseCommand, LoggerMixin):
if
hasattr
(
doc
,
field
):
setattr
(
doc
,
field
,
count
)
doc
.
save
()
self
.
cronjob_log
.
error
(
'{0} [process failed (edms upload)] [task={1}] [err={2}]'
.
format
(
self
.
log_base
,
task_str
,
e
))
self
.
cronjob_log
.
error
(
'{0} [process failed (edms upload)] [task={1}] [err
or
={2}]'
.
format
(
self
.
log_base
,
task_str
,
traceback
.
format_exc
()
))
write_zip_file
(
img_save_path
,
os
.
path
.
join
(
doc_data_path
,
'{0}_img.zip'
.
format
(
doc
.
id
)))
else
:
...
...
src/apps/doc/ocr/wb.py
View file @
3bdf0e4
...
...
@@ -161,13 +161,13 @@ class BSWorkbook(Workbook):
if
max_find_count
==
0
:
return
round
(
random
.
uniform
(
75
,
80
),
2
)
elif
max_find_count
==
1
:
return
round
(
random
.
uniform
(
80
,
85
))
return
round
(
random
.
uniform
(
80
,
85
)
,
2
)
elif
max_find_count
==
2
:
return
round
(
random
.
uniform
(
85
,
90
))
return
round
(
random
.
uniform
(
85
,
90
)
,
2
)
elif
max_find_count
==
3
:
return
round
(
random
.
uniform
(
90
,
95
))
return
round
(
random
.
uniform
(
90
,
95
)
,
2
)
else
:
return
round
(
random
.
uniform
(
95
,
100
))
return
round
(
random
.
uniform
(
95
,
100
)
,
2
)
@staticmethod
def
month_split
(
dti
,
date_list
,
date_statistics
):
...
...
Write
Preview
Styling with
Markdown
is supported
Attach a file
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 post a comment