|
|||
|
|||
| Moderated by: Renate.Reinartz, Markus.Kreisel, Jaakko.Salmenius, Ilkka.Salmenius |
|
|||||||||||||
| XSL and more - Wish list for software localization tool - Technical Support (You need to be registered at the forum to write) - Localization Tool for VB, Delphi, .NET, C#, VB.NET, XML, Online Help, HTML ... | ||||||||||||||
| Author | Post | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
||||||||||||||
|
stephan Member
|
Hi, We are evaluating Sisulizer for our future localization projects. We like the price and features, but there is one blocking issue: XSL localization. Is there a chance that you add better support in the near future? 1. XSL: Look at Passolo and how they allow to define the xml nodes to localize. E.g. we would want to instruct Sisulizer to extract ALL nodes, WHATEVER their location in the xml document, e.g. <td>..</td>, <span>..</span>, <p>..</p>. Currently with Sisulizer, xsl localization is almost mission (or configuration) impossible. Consider that in web development xsl is most oftenly used to render html... Sisulizer requires to configure each xsl file individually (not always working). Please try by yourself. This should be easy to add and would make Sisulizer really web ready and our product choice. Other wishes in order of priority: 2. Resource files: Allow marking of string lists as being of data type HTML. In .Net 2.0 local resources are extracted from asp controls and typically contain html formatted text. It would be great if Sisulizer would - on demand - identify and protect the html tags embedded in such strings. You already do that when localizing web pages. 3. Text files: Currently Sisulizer can extract quoted strings only (or a custom delimiting character). We would require a configuration such as: Extract strings preceeded by [character sequence] and terminated by [character sequence]. 4. Search tool to look up texts in TMs. 5. We had tested the database localization features. Fantastic! Efficient and easy to use. My only wish: The wizard scans the entire database to look for localizable tables. This process takes some 15 minutes with our database (it's huge). I would actually prefer to get just the raw list of all tables (one db query) and then manually select the tables I want to add to a project. Thank you, Stephan
|
|||||||||||||
| ||||||||||||||
| ||||||||||||||
|
Jaakko.Salmenius Administrator 
|
Hi, 2: Do you mean that you have a custom resource in .resx that contains HTML data. You want to scan thet data as HTML data instead of plain text. 3: Can you send me a sample text file or reply a sample line where I can figure out your needs. 4: Sisulizer 2.0 will contains Translation Memory Editor that lets you maintain and edit the contents of translation memory. 5. Good idea. I will implement it. Berst regards, Jaakko
____________________ http://www.sisulizer.com - Three simple steps to localize |
|||||||||||||
| ||||||||||||||
|
||||||||||||||
|
stephan Member
|
Hello, Thanks for the replies. 2: A string in a resource file may be "Read more on <b>our product</b>". Tag protection would be useful but cannot currently be activated. I also think of html content stored in databases (typical case with content mngt systems). It would be nice to be able to mark strings as being HTML, individually per string and globally for an entire stringlist. .Net 2.0 automatically adds such a string to the resource file for code: <asp:Localize ID=Localize1 runat=server>Read more on <b>our product</b></asp:Localize> 3: We have JS file that contains lines such as: aI("text=At a glance;url=" + _dbbd + "/Products/Default.aspx;"); 4: Will you also add a concordance search tool that allows translators to find fuzzy matches while translating? This is a feature that exists in all the classic translation memory tools (Trados, SDL, etc.). Regards, Stephan
|
|||||||||||||
| ||||||||||||||
| ||||||||||||||
|
Jaakko.Salmenius Administrator
|
2: We will implement resource string tagging as HTML data. What do you mean by tag protection? 3: This looks like source code line. If the file is source code you should use source code localizaiton not text file localization. What is the file extension? 4: Sisulizer translation memory does not have fuzzy matching. It is more like translation repository. If you want to use fuzzy matching you can use Trados and use Sisulizer's trados interface. Jaakko
____________________ http://www.sisulizer.com - Three simple steps to localize |
|||||||||||||
| ||||||||||||||
|
||||||||||||||
|
stephan Member
|
2: We will implement resource string tagging as HTML data. What do you mean by tag protection? I mean: A translator cannot overwrite or remove protected tags. The target string must contain the same tags in order for the tool to accept the translation. Example: Hello <a>world</a> => Der Welt einen guten Tag</a> : ERROR Example: Hello <a>world</a> => <a>Der Welt</a> einen guten Tag : OK 3: This looks like source code line. If the file is source code you should use source code localizaiton not text file localization. What is the file extension? Extension: js (Javascript). Can you please explain how to configure Sisulizer to extract all strings that appear between "test=" and ";url"?
|
|||||||||||||
| ||||||||||||||
| ||||||||||||||
|
Jaakko.Salmenius Administrator
|
2: I got it. Sisulizer will check this after we have implemented HTML editing in resouce string 3: If you have a JavaScript file use the source code localization. Select the .js file. Sisulizer will extract string from the file and created a localized copy of the file. aI("text=At a glance;url=" + _dbbd + "/Products/Default.aspx;"); Here Sisulizer will find two strings: text=At a glance;url= and /Products/Default.aspx; Do you mean that you just want to At a glance to be extracted? Jaakko
____________________ http://www.sisulizer.com - Three simple steps to localize |
|||||||||||||
| ||||||||||||||
|
||||||||||||||
|
stephan Member
|
> Do you mean that you just want to At a glance to be extracted? Exactly.
|
|||||||||||||
| ||||||||||||||
| ||||||||||||||
|
Jaakko.Salmenius Administrator
|
We will implement feature where you can set the format of a resource string to be onw of the following: 1) Plain text (this is the current implementation) 2) HTML. Sisulizer treats string as HTML data and localizes it as HTML 3) XML. Sisulizer treats string as XML data and localizes it as XML 4) Combined data. The user can give a regular expression that specifies the format of the string. Sisulizer extracts and localizes the variables. For example if we have string: One;Two;Three=[Four] And we want to have One, Two, Three and Four as separate strings we would have to give a regular expression #;#;#=\[#\] Where # specify strings to be extracted and localized. We already use regular expressions in SRX segmentation so we have classes to implement this quickly. Most likely it will take a week or two. The string format can be assigned as per row (affecting only to that node) or per node (affecting all rows of that node). This feature will be available to all string resources such as: - WIN32 STRINGTABLE - .NET string resource - Java's resource bundle - Source code string - Ini file string - Database string - XML string (regular expression only) If you have comments or other wisher please let me know. Best regards, Jaakko
____________________ http://www.sisulizer.com - Three simple steps to localize |
|||||||||||||
| ||||||||||||||
|
||||||||||||||
|
stephan Member
|
Sounds good. Thank you.
|
|||||||||||||
| ||||||||||||||
| ||||||||||||||
|
tscherf Member
|
I wanted to ask one more question about the fuzzy matching. I don't quite know what your overal goal is of the tool, but I can see that it may very well be able to replace technologies like SDL Trados in the long run. One major usage factor for translators is in fact the ability to fuzzy match. This can happen in a fuzzy match suggestion box, which shows the top 3 or so strings that match the current string to be translated. This is also important, because it will improve the overall quality of work due to consistency of translation, especially if you have teams translating. Thanks Toby
|
|||||||||||||
| ||||||||||||||
|
||||||||||||||
|
Markus.Kreisel Administrator 
|
Hi Toby, that is exactly what we have in mind. Our developers already have developed and excellent fuzzy matching routine. But fuzzy matching is a slow process so it can not be used in the automatic translation memory process as we have now. In the moment TM is used as a batch and it would be not nice if Sisulizer would ask you x times in a row in that batch to choose from multiple options. So we decided to have a more advanced editor like you described with options from the TM sorted by fuzzy rate for the current row only. It is already under development and we are very happy to see that the request for such is huge. Best Markus
____________________ http://www.sisulizer.com - Three simple steps to localize |
|||||||||||||
| ||||||||||||||
Sisulizer software localization tool - Three simple steps to localize
