Experiences converting DBpedia's bz2 files

[kurzum]

Hi all,
I just tried to convert DBpedia to hdt and would like to share my experience. Some things did not owrk for me, although I am not sure, whether it is my fault or just missing features.

Java

1. I saw that bz2 works for Java. Good!
2. I tried to read in several bz2. files in Java as input and this didn't work (several input files not allowed)
3. I tried to read one file via stdin in java which worked
4. I tried to read in several files in Java via stdin with process substitution and mvn exec <(lbzip2 file1.bz2) <(lbzip2 file2.bz2) but it didn't work
5. I tried to create a big Java Jar with all dependencies with mvn assembly:single, but it just created the bin files (I was too lazy to adjust the descriptors in pom.xnml)

cpp

1. Although I did apt-get install serdi serd-dbg make failed, I disabled it in the Makefile then it compiled
2. bz2 didn't work. actually the unzipping seemed to work, but the parser threw a lot of errors
3. reading via stdin with process substitution failed as cpp doesn't accept stdin
4. reading from several input files worked. Well, it said Sorting triples and I lost patience as the percentage didn't go up

Overall, I solved it now by sort -um <(lbzip file1.bz2) .... | gzip2 > core.gz

My question is just whether all of the above is intended and expected behaviour or whether I should try some of the methods again as I might have done something wrong.
All the best,
Sebastian