Catbo is computer-assisted translation web service, CAT; it is designed to help a human translator to translate documentation and other texts. Automatic machine translation systems available today are not able to produce high-quality translations.
Catbo has full support for texts in the formats: HTML PDF EPUB SRT. Our next goal is to support wikipedia/mediawiki format, odt/docx and so on (xml/docbook/rst/...).
Projects are used to setup a language pair for all documents to translate in them. For example, the pair English => Russian is to translate an English text to Russian. Translation Memory and Term Base are local to a project, too.
A registered user is to open the Add New Project link:
Let's say you have a project Test-enru which is accessible at http://catbo.net/project/Test-enru/. The next step is to load a HTML document with "Add New File" link. Before you may start to translate it is to be processed with Indexer. You may watch the status of the corresponding job by opening the file's dialog, click its name at Files table. The maximum file size is 10 Mb now.
Every new added translatable document is placed in the queue, and eventually will be parsed into segments. A segment is a least translatable element in Catbo, and usually is a single sentence. If Indexer was able to define it as a translatable document then the total number of segments is shown. To start translating, open file's dialog and follow the Translate now link.
You may upload/update not only a single HTML file, but a zip arhive with many documents at once. Thus you are able to start translation a multipage documentation easily.
Every document is translated in Translation Editor. Translation Editor is the crucial element of the translation process. It is a grid each row of which represents a segment. A segment row consists of the following fields:
The translation process is rather simple:
Enter).
Ctrl+1..9
to insert a found translation from Translation Memory or append a found term from Term Base.
Ctrl+Enter.
The translation grid is auto-expanding as you select last visible segments.
Once a segment is translated the generation of new translation of the document is planned automatically after a while (4 hours). You may see the last translation following the link Show the last translation in the file's dialog.
| Keyboard Shortcut | Description |
|---|---|
Down
| Select next segment |
Up
| Select previous segment |
Enter
| Open current segment |
Esc
| Close current opened segment without saving |
Ctrl+Enter
| Save current segment and go to next one |
Ctrl+1, Ctrl+2, ... Ctrl+9
| Insert the nth hit from Translation Memory or Term Base into the current segment target |
Ctrl+Shift+S
| Copy the source text into the target one |
Ctrl+Alt+S
| Select the source text |
Ctrl+Alt+W
| Insert markup tags from the source text |
Ctrl+G
| Go to Segment Dialog |
Ctrl+E
| Add a term to Term Base |
Ctrl+P
| Search for a term |
Ctrl+Alt+G
| Toogle Google Translate support |
Ctrl+Alt+D
| Search dictionaries from dict.catbo.net |
Ctrl+Alt+H
| Show shortcut's list |
Ctrl+Alt+M
| Merge existing translation |
Ctrl+Shift+Up/Down
| Select previous/next merging segment |
| Dialog Shortcuts | |
Esc
| Cancel the current dialog |
Enter
| Ok the current dialog |
There is a percent value just after every TM hit, i.e. 90%.
The higher the value, the more both segments look alike. Here are some predefined values and their meanings:
98% means that both segments are the same except their formatting; formatting changes includes
different punctuation and tags like bold and italic text emphasis.
100% means that both segments are the same.
101% means that both segments are the same and their contexts are the same, too. It means
that their previous and next segments are the same.
Every Catbo user may help to translate documents (until the project owner limits the access to experienced translators with a test question). Moreover, we the Catbo developers encourage users to participate in making translations in a wiki manner.
Special rights of a project owner (for now, may be improved or changed by request):
Modern documentation is not written at a time but updated at times; so the translation should be updated, too. Once a document is updated via Update/upload new version function Indexer not only parses it into segments, but also fill in 101% match cases from the past translation.
Note: just after update the generated document becomes untranslated. Use Force translation output to regenerate the translation.
Once a file named README.markdown is uploaded it is treated as an amplification of the project; see the project HTML5-enru for illustrative purposes. It should be written in Markdown formatting syntax "which allows you to write using an easy-to-read, easy-to-write plain text format", and which is to be converted to HTML. Here is an Markdown example:
A First Level Header ==================== This is just a regular paragraph. ### Header 3 > This is a blockquote.
You may test the HTML generation of Markdown. Catbo uses the Github flavoured Markdown.
If you have a certain translation or part of translation you may use the feature "Merge existing translation" to merge the translation into your project to accelerate the translation process. At first you are to add that already translated document into your project. After Catbo parsed it, go to Translation Editor, select an appropriate segment and press Ctrl+Alt+M. Then search needed translated segment and select it. After it you get translated segments as tips. Use Ctrl+Shift+Up/Down to calibrate current merging segment. To cancel merging, press Ctrl+Alt+M again and click "Stop merging".
To parse an already translated document according to the target language, not the source one, (re)name an document with the prefix "translated-".
Use the contact form to ask questions about Catbo.