Okapi Framework

• Creation of translation packages (e.g. XLIFF, OmegaT, RTF).
• Conversion of character encoding or line-breaks.
• Conversion to various file formats.
• Terminology extraction.
• Pseudo-translation and various modification of the source text for testing.
• Batch translation using various MT and TM sources.
• Apply XSLT template in batch mode.
• Quality check report.
• Comparison between different translations.
• Search and replace on filtered text.
• Source and translation alignment.
• And more...

Like other Okapi applications, Rainbow uses the components called "filters" that take input files in a given native format and extract its translatable content into common resources that the Okapi tools can manipulate and modify. There are filters for amny formats, for example: OpenOffice, XML, HTML, Properties, DTD, MS Office, tables, etc. Many translation-related formats are also supported: XLIFF, PO, TS, TMX, TTX, etc.

Rainbow also uses the Okpai components called "connectors": They allow you to access different type of local or remote translation resources. For example: Google-MT, Microsoft-MT, OpenTran, MyMemory, Translate Toolkit TM, Apertium, TDA-Search, and more.

Note that, in addition to third-party TM access, Rainbow can also use Okapi's own TM egine: the Pensieve TM.

Another feature of Rainbow is to allow the user to build "pipelines" of tasks using simple components called "steps" that perform well-defined limited actions. You can chain those steps to execute many tasks in a single process.

Some of the verifications performed are:

• Repeated words
• Corrupted characters
• Patterns in the source text that should correspond to a given pattern in the translation.
• Inline codes differences.
• Translation suspiciously longer or shorter than the source.
• Missing translation.
• Leading and trailing spaces.
• And more...

You can disabled warnings that happen to be incorrect, so the next time you re-run the verification those warnings remain disabled. You can also save your session and come back to it later.

In addition to the interactive list of warnings, the tool can generate complete reports of the issues found.

CheckMate allows you to integrate directly the verifications done by the open-source LanguageTool checker, which offers a wide range of simple and complex checks for many languages.

You can work with any bilingual or multilingual file format supported by the Okapi filters, for example: XLIFF, TMX, PO, TS, TTX, etc.

Its features includes for example:

• Immediate previous of the segmentation as you edit the rules.
• Can enabled/disable rules for easier debugging.
• Apply the rules on input files for testing.
• Copy/Paste the SRX document to Clipboard for easy exchange with other tools.
• Can test with different language codes
• And more...

Segmentations rules are used in several of Okapi's components to break text units into small parts such as sentences. Using SRX rules allows to be more interoperable with other sets of tools. For example, when sharing a translation memory, SRX allows two different tools to also exchange the rules used to break down the paragraphs into the sentences that are in the TM.

• Extracting and merging XLIFF documents
• Querying different MT and TM systems
• Performing format conversions
• Translate directly files in various formats using TM or MT ssytems.
• And more...

Like other Okapi applications, Tikal uses the components provided by the Okapi libraries, for example the connectors to differents translation resource engines such as Apertium, OpenTran, Googl-MT, Microsoft-MT, MyMemory, etc. It also uses the Okapi filters, allowing you to process files in many different formats: HTML, ODT, DOCX, PO, XLIFF, TMX, XML, and many more.