#297 still has issues with python2
venix1 opened this issue · 5 comments
A recent upgrade to 1.1.20 produced the following
[WARNING] must be unicode, not str
The warning was tracked down to this line.
with io.open(definitions_file, "w", encoding="utf-8") as sqlf:
sqlf.write(invalid_views)`
Added here . A solution was to convert invalid_views to utf-8 before writing.
with io.open(definitions_file, "w", encoding="utf-8") as sqlf:
sqlf.write(invalid_views.decode('utf-8', 'ignore'))
Thanks for reporting this. I've been unable to reproduce this issue on Centos 7 with multiple different versions of Mysql/Mariadb. Can you please let me know what locale the system is set to?
[root@e317edbbd18d /]# python --version
Python 2.7.5
[root@e317edbbd18d /]# holland --version
1.1.20
[root@e317edbbd18d /]# cat /etc/redhat-release
CentOS Linux release 7.7.1908 (Core)
[root@e317edbbd18d /]# mysql --version
mysql Ver 14.14 Distrib 5.7.28, for Linux (x86_64) using EditLine wrapper
I'm able to duplicate it with the following test case.
import io
invalid_views = "--\n-- DDL of Invalid Views\n-- Created automatically by Holland\n--\n"
with io.open('test.txt', "w", encoding="utf-8") as sqlf:
sqlf.write(invalid_views)
As for everything else, it matches yours.
# python --version
Python 2.7.5
# holland --version
1.1.20
# cat /etc/redhat-release
CentOS Linux release 7.7.1908 (Core)
# mysql --version
mysql Ver 14.14 Distrib 5.7.28, for Linux (x86_64) using EditLine wrapper
# locale
LANG=en_US.UTF-8
LC_CTYPE="en_US.UTF-8"
LC_NUMERIC="en_US.UTF-8"
LC_TIME="en_US.UTF-8"
LC_COLLATE="en_US.UTF-8"
LC_MONETARY="en_US.UTF-8"
LC_MESSAGES=C
LC_PAPER="en_US.UTF-8"
LC_NAME="en_US.UTF-8"
LC_ADDRESS="en_US.UTF-8"
LC_TELEPHONE="en_US.UTF-8"
LC_MEASUREMENT="en_US.UTF-8"
LC_IDENTIFICATION="en_US.UTF-8"
LC_ALL=
# localectl status
System Locale: LANG=en_US.UTF-8
LC_MESSAGES=C
VC Keymap: us
X11 Layout: us
Something strange is happening. I get the error as you describe when running your code snippet.
[root@07a488274925 /]# cat test.py
#!/usr/bin/env python
import io
invalid_views = "--\n-- DDL of Invalid Views\n-- Created automatically by Holland\n--\n"
with io.open('test.txt', "w", encoding="utf-8") as sqlf:
sqlf.write(invalid_views)
[root@07a488274925 /]# ./test.py
Traceback (most recent call last):
File "./test.py", line 7, in <module>
sqlf.write(invalid_views)
TypeError: must be unicode, not str
But holland still prints out the invalid_views.sql file without an error.
[root@07a488274925 /]# cat /var/spool/holland/mysqldump/20191230_161731/invalid_views.sql
--
-- DDL of Invalid Views
-- Created automatically by Holland
--
--
-- Current View: `test`.`v1`
--
CREATE ALGORITHM=UNDEFINED DEFINER=`root`@`localhost` SQL SECURITY DEFINER VIEW `test`.`v1` AS select `test`.`t1`.`a` AS `a` from `test`.`t1`;
Using codecs.open has the desired effect without having to decode invalid_views. It's actually being used for the same purpose elsewhere in the file and I didn't notice before.
[root@07a488274925 /]# cat test.py
#!/usr/bin/env python
import codecs
invalid_views = "--\n-- DDL of Invalid Views\n-- Created automatically by Holland\n--\n"
with codecs.open('test.txt', "wt", encoding="utf-8") as sqlf:
sqlf.write(invalid_views)
I'm hesitant to release this without understanding what's going on. I'd like to figure out why the code works in holland in my test environment and not on the system you reported this for.