Rename pairs in a "list of pairs" dataset collection for input to pipeline

[pjbriggs]

The pipeline tool can accept a "list of pairs" dataset collection as input, but currently the names of each pair (taken from the names of the fastqs which comprise the pair) are ignored when preparing input for the pipeline.

Instead the names are taken in order from the first column of the supplied Metatable.txt file; however this procedure is not transparent and unsafe.

A preferable approach would be to use the names directly from the collection, and provide a way to easily rename the pairs e.g. by supplying a tool which could create a new collection using a file mapping pair names to new names.

[pjbriggs]

The relabelling tool implemented in PR #20 isn't really fit for purpose (see comments in the PR).

A vastly superior implementation is available here galaxyproject/galaxy#3603.

Galaxy code can be patched to enable the tool using e.g.

cd /PATH/TO/galaxy
wget https://github.com/galaxyproject/galaxy/pull/3603.patch
patch [--dry-run] -i 3603.patch -p 1

Also if the Galaxy instance has a config/tool_conf.xml then the new tool must be added there too:

<tool file="${model_tools_path}/relabel_from_file.xml" />

[pjbriggs]

Closing this issue as Galaxy now has a Relabel List Identifiers... tool (under Collection Operations) which can be used for doing this.