xStarter 1.9.3.55 Beta - Unicode test

General questions about xStarter, how to create tasks, etc.
Janusz
Posts: 30
Joined: Sun Mar 01, 2009 1:32 pm
Location: Poland, Zabrze
Contact:

xStarter 1.9.3.55 Beta - Unicode test

Postby Janusz » Wed Jun 10, 2009 7:18 pm

Hi Alex,

I’ve performed first Unicode support test for 1.9.3.55 beta. Below are details:
I made new task with WriteToLog action. To log message I typed following Polish umlauts: “ąęłóśżź”.

Unchecked “Unicode” option

- On Polish Windows XP (Windows-1250 codepage) I’ve got correctly saved polish umlauts in output txt file. As testing tools I used Notepad and PSPad text editor:

Image

- On English Windows XP I’ve got incorrectly saved Polish umlauts in output txt file, but this is normal behavior, because Windows-1252 doesn’t support Polish umlauts by default. As testing tools I used Notepad and PSPad text editor
http://www.sisulizer.com - Three simple steps to localize

Janusz
Posts: 30
Joined: Sun Mar 01, 2009 1:32 pm
Location: Poland, Zabrze
Contact:

Re: xStarter 1.9.3.55 Beta - Unicode test

Postby Janusz » Wed Jun 10, 2009 7:36 pm

Checked “Unicode” option (UTF-8)

- On Polish Windows XP and Polish Vista I’ve got following result after opening with Notepad:

Image

- Below is this same file opened in PSPad (Polish Windows XP and Polish Vista):

Image

PSPad recognized this created by xStarter file as HEX file, not TXT file (BOM issue?)!

- On English Windows XP I’ve got with PSPad similar result as with Notepad. That is, content of test.txt file is displayed as text format, not binary!

Image

- Next I trimmed log message only to 3 Polish umlauts (ąęł) and PSPad began to recognize this file as TXT file on Polish Win XP and Vista:

Image
http://www.sisulizer.com - Three simple steps to localize

Janusz
Posts: 30
Joined: Sun Mar 01, 2009 1:32 pm
Location: Poland, Zabrze
Contact:

Re: xStarter 1.9.3.55 Beta - Unicode test

Postby Janusz » Wed Jun 10, 2009 7:45 pm

Checked “Unicode” option (UTF-8) with added Timestamp

After checking option “Write date and time before message” for shorter version of polish umlauts sequence PSPad and Notepad correctly displayed date and time, but for longer version PSPad still display file as binary file.

Image

Seem to me that xStarter can’t correctly convert Polish umlauts from ANSI string typed to xStarter dialog to output UTF-8 format, and Polish umlauts are incorrectly displayed. In some cases (long sequence of Polish umlauts on Polish Windows OS) could even completely to break file format.

Below is link to my sample task with included Polish umlauts. I hope it allows you (at least partially) to reproduce issue:
http://grzybek.anet.pl/xStarter/Unicode_Test.rar

Best regards,
Janusz
http://www.sisulizer.com - Three simple steps to localize

User avatar
Alex
Site Admin
Posts: 341
Joined: Mon May 21, 2007 4:09 pm
Contact:

Re: xStarter 1.9.3.55 Beta - Unicode test

Postby Alex » Thu Jun 11, 2009 4:55 am

Hi Janusz,

I was really didn't thought about BOM. How you think, it's better to add BOM checkbox in text related actions or add in codepage combobox UTF-8 BOM, UTF-16 BOM items?

I know it's better to always use BOM, but know issues with PHP files what uses resources only without BOM (in my case it was language files for phpbb3).

Additionally I was found codepage of files with BOM can be autodetected. So, in the ReadFromLog action it's possible to add item "Autodetect by BOM".

Janusz
Posts: 30
Joined: Sun Mar 01, 2009 1:32 pm
Location: Poland, Zabrze
Contact:

Re: xStarter 1.9.3.55 Beta - Unicode test

Postby Janusz » Thu Jun 11, 2009 8:50 am

Hi Alex,

I know it's better to always use BOM, but know issues with PHP files what uses resources only without BOM (in my case it was language files for phpbb3).

You are right, BOM could to generate problems in some web solutions based on LAMP, but it is required for desktop (Windows) applications, so implementing both options, that is, possibility of saving Unicode file with BOM header and without it will be very flexible solution and allows on using files generated by xStarter both on Windows systems and e.g. Apache, PHP, MySQL systems.

How you think, it's better to add BOM checkbox in text related actions or add in codepage combobox UTF-8 BOM, UTF-16 BOM items?

Second idea looks more simple (you don’t need to add new checkbox to UI), but first idea is more logical, because UTF-8 with BOM or without BOM is still this same UTF-8 format with different settings, while dropdown menu currently contains really different Unicode formats. So first idea (adding new option as checkbox) seem to me better.

Additionally I was found codepage of files with BOM can be autodetected. So, in the ReadFromLog action it's possible to add item "Autodetect by BOM".

Exactly. First I thought about implementing it by adding new item to ReadFromLog dropdown menu (Autodetect, UTF, UTF-16), but xStarter with selected Autodetect option couldn’t to set up correct Unicode format if source log file doesn’t contains BOM header, so seem to me that “autodetction” feature should to be implemented in another way.

Janusz
http://www.sisulizer.com - Three simple steps to localize

User avatar
Alex
Site Admin
Posts: 341
Joined: Mon May 21, 2007 4:09 pm
Contact:

Re: xStarter 1.9.3.55 Beta - Unicode test

Postby Alex » Fri Jun 12, 2009 1:06 am

New beta ready with BOM support:
viewtopic.php?f=5&t=199&p=584#p573

Janusz
Posts: 30
Joined: Sun Mar 01, 2009 1:32 pm
Location: Poland, Zabrze
Contact:

Re: xStarter 1.9.3.55 Beta - Unicode test

Postby Janusz » Fri Jun 12, 2009 7:49 am

Thanks for info. I tested new build and adding BOM setting resolved issue.

Janusz
http://www.sisulizer.com - Three simple steps to localize


Return to “Discussion”

Who is online

Users browsing this forum: No registered users and 8 guests