Size limit of Semantic XML in TBCME

[Jack Hodges]

I was looking in the archives to see if there is a limit to the size of files using Semantic XML and found only 1 thread from way back (2009). I have to import large datasets into a knowledge graph (offline for now) and, like the original poster, I have no problem with very small files but even moderately-sized files are producing errors. I am using TBCME version 6.3 on a MacBook running macOS Catalina, 10.15.7 with 32 GB of memory. I have allocated (in the Contents/Eclipse/TopBraid Composer.ini file) to 12 GB of memory with no improvement.

[Holger Knublauch]

On 2021-01-13 2:53 am, Jack Hodges wrote:
> I was looking in the archives to see if there is a limit to the size
> of files using Semantic XML and found only 1 thread from way back
> (2009). I have to import large datasets into a knowledge graph
> (offline for now) and, like the original poster, I have no problem
> with very small files but even moderately-sized files are producing
> errors.

Are these out-of-memory errors, or what?

> I am using TBCME version 6.3 on a MacBook running macOS Catalina,
> 10.15.7 with 32 GB of memory. I have allocated (in the
> Contents/Eclipse/TopBraid Composer.ini file) to 12 GB of memory with

> no improvement. --

Semantic XML is quite chatty and produces lots of triples for each XML
element, so some overhead is to be expected. We didn't encode some kind
of limit though (that I could think of). If you get stuck there are all
kinds of 3rd party approaches that would allow you to pre-process the
XML, for example using XSL-T to produce RDF/XML, or any API in an
imperative language. Not every problem can be solved by TopBraid alone.

Holger