QueueMetrics > Running QueueMetrics

[major bug] unicode issues

(1/2) > >>

qmax:
Installation:

mysql server is configured to use utf8 by all defaults:
character-set-client-handshake    TRUE
character-set-filesystem          binary
character-set-server              utf8
default-character-set             utf8

all database tables were created or converted to:
CREATE TABLE ... ( ... ) DEFAULT CHARSET=utf8

mysql jdbcdriver is set up like following:
jdbc:mysql://localhost/queuemetrics?...&useUnicode=true&characterEncoding=utf8&characterSetResults=utf8

file agents.conf are like following:
agent=>001,,"Василий Пупкин
(here is utf-8 string after quote)

Content-Type of all pages is: text/html; charset=UTF-8

User selected Language is "Russian"

Issues:

1. Setup Wizard
when loading agents from file agents.conf, after specifying filenames,
after page "The specified sources have been succesfully read."
on stage "Source > Agents" it displays unicode strings incorrectly.
(a string of invalid chars 2 times longer then initial)

if to proceed further, on last stage it produces error:

--- Quote ---Fail: Problema update DB: java.sql.SQLException: You have an error in your SQL syntax; check the manual that corresponds to your MySQL server version for the right syntax to use near 'INVALID CHARS' at line 1
--- End quote ---
the "INVALID CHARS" is string invalid chars (4 times longer then initial string.)

when agent name is put into database tables directly, using mysql-tools, then quemetrics displays it correctly everywhere.

2. Track QA
when track QA window is opened, and there comment put, it saves in database incorrectly,
and then redisplayed in the wrong way.

when comment is put into database tables directly, using mysql-tools, then quemetrics displays it correctly

3. Exporting
Exporting just any table into XLS or CSV replaces all unicode characters into "?"
Only XML export works.

qmax:
Some details:

in the list of loaded agents to select for inmorting,
each byte of original utf8 string is displayed as U+FFDD "replacement char"

when comments are placed in QA form,
every byte of string is saved as U+00xx, where xx is value of that byte.
e.g. character U+0410 in utf presended as bytes D0 90 and thansaved as U+00D0 U+0090

P.S.
headers and questions in QA forms are all saved and displayed properly.

marcos:
Hi,

Thank you for your detailed explanations.
I've submitted this as bug #775 in our internal bug tracker and it will be planned to be fixed on next coming release 1.5.4.

QueueMetrics:
That's fixed now.

marcos:
The bug is fixed since the version 1.5.4.
To have it working, please be sure to set the key:

default.charsetEncoding=UTF8

in the configuration.properties file.

Remember to restart the queuemetrics service and to log-off and log-on again to QM session.

Navigation

[0] Message Index

[#] Next page