Bug #51815 | SQL Editor crashes on syntax checking: Sql_editor::check_sql() | ||
---|---|---|---|
Submitted: | 8 Mar 2010 0:40 | Modified: | 26 Mar 2010 13:32 |
Reporter: | Bob Lamberson | Email Updates: | |
Status: | Closed | Impact on me: | |
Category: | MySQL Workbench: SQL Editor | Severity: | S2 (Serious) |
Version: | 5.2.16 beta | OS: | Any |
Assigned to: | Sergei Tkachenko | CPU Architecture: | Any |
[8 Mar 2010 0:40]
Bob Lamberson
[8 Mar 2010 0:56]
Bob Lamberson
After closing the reported error dialog, I typed in use test on the first line of the sql window and got another error message show in the jpg screen snip added in the files section.
[8 Mar 2010 0:57]
Bob Lamberson
2nd error dialog
Attachment: addedUseTest db then this error.JPG (image/pjpeg, text), 69.50 KiB.
[8 Mar 2010 5:49]
Peter Laursen
As this happens after first linebreak my guess is that WB expects Unix linebreaks (\n) and the file created by Notepad has Windows linebreaks (\r\n). I think there is a similar or maybe even identical report already. Also a utf8-file saved by Notepad has BOM's (Byte Order marks) what I do not know if WB handles properly. Peter (not a MySQL person)
[8 Mar 2010 6:08]
Peter Laursen
Besides WB should handle ANSI-encoded files smoothly (if it does not). Most .sql/.txt files generated by (Microsoft or non-Microsoft) Windows programs will be ANSI encoded.
[8 Mar 2010 6:18]
Bob Lamberson
Bug #51308 does deal with the issue of line endings, if that is the cause of the error messages described here, as Peter Laursen suggests.
[8 Mar 2010 13:33]
Sergei Tkachenko
I have strong reasons to blame a threading issue related to missing mutex quarder. Fix for that was pushed to repository. Let me know if this bug reoccur in next version.
[23 Mar 2010 13:30]
Johannes Taxacher
fix confirmed in repository
[26 Mar 2010 13:32]
Tony Bedford
An entry has been added to the 5.2.17 changelog: MySQL Workbench crashed while attempting to edit code in the SQL Editor.
[3 Dec 2011 20:36]
Gordon MacKay
I have at least one crash a day where the SQL parser looses it. Good fix would be the ability to turn the parser off and avoid the libwpublic.be.dylib Sql_edit::check_sql crashes.