Report does not accurately reflect repeated and parallel stages

[GoogleCodeExporter]

I have a pipeline that runs steps in parallel. The pipeline works very nicely, 
however the report is not consistent with what has been run.
Here is the pipeline with comments after the # where there is an inconsistency.

I have attached the report generated.

In addition only one output is ever reported - ( you can see this where a stage 
looks like it has one output, but the next stage shows 2 inputs )

Bpipe.run {
    "%" * [ \
            base_init \                                     
            + base_get_illumina_reads \           
            + base_check_inputs \                   
            + "%" * [ qc_fastqc ] \                     # only first instance reported
            + "%" * [ base_md5sum ] \              # not reported at all
            + base_check_inputs \
            + qc_filter_raw_reads \
            + base_check_inputs \
            + "%" * [ base_md5sum ] \              # not reported
            + "%" * [ qc_fastqc ] \                     # not reported
            + base_check_inputs \
    ]
}

Original issue reported on code.google.com by ssade...@gmail.com on 16 Jul 2014 at 4:57