Services

lenses on document

Maxprograms provides custom software development and integration services focused on XML standards. Localization, through formats such as TMX, XLIFF, and DITA, is where that focus is most visible, but the underlying work, parsing, validating, and transforming structured data, applies wherever XML or similar standards are involved.

XML is our specialty. We maintain custom libraries for it in both Java and TypeScript, and combine and extend them to support specific workflows, integrate with existing systems, and automate processes that would otherwise be done by hand.

That work spans more than twenty years, and the libraries and tools built along the way, covering XML parsing, translation memory, and language data, are the clearest evidence of it.

What we build

When nothing existing solves the problem, we build custom software that does. That work covers:

Workflow automation and batch processing pipelines
Custom import and export filters for proprietary or specialized formats
Integration with content management systems, CAT tools, and repositories

Translation memory processing, normalization, and optimization
Validation and quality control tools for TMX and XLIFF data
DITA to XLIFF localization pipelines and content reintegration

Built on our own libraries

We built custom libraries instead of relying only on general-purpose ones, giving us direct control over how XML, translation memory, and language data are processed. We also extend the platforms we build on, Java and TypeScript, with libraries those languages lack. These libraries are the foundation of our work.

TypeScript has no native XML parser, so we wrote our own: TypesXML, a full XML parser and processing library, with DTD and XML Schema support and XML Catalog resolution. TypesXLIFF builds on it for XLIFF-specific processing, while HybridTM handles semantic translation memory, combining traditional matching with vector search. Two smaller libraries round out the set: TypesBCP47 for parsing and validating BCP 47 language tags, and TypesTerms for terminology extraction.

On the Java side, XMLJava builds on Java's own XML parsers, which are solid but not enough on their own, adding what's missing for our processing needs, and OpenXLIFF Filters converts dozens of file formats to and from XLIFF for translation. The same pattern repeats here: BCP47J for language tags, Terms for terminology extraction.

A third group adds AI to the mix. MTEngines provides a common interface to multiple machine translation engines, XliffAI applies AI models directly to XLIFF translation, and XliffDeepl integrates DeepL into XLIFF translation workflows.

We combined these libraries to build our tools portfolio, all public in our GitHub organization, github.com/maxprograms-com, which holds 31 repositories. See the Developer Platform for the full catalog, with installation instructions for each library.

Need a custom workflow or integration?

Get in touch to discuss your requirements.