GitLab

fix keyword error

This function can be called when one needs to check disk space. If the
disk space is below a specified percentage threshhold, the function
notifies the system administrators.
A boolean is returned to indicate the disk space situation.

Fused files with many samples are large.
By default, we limit the number of samples taken into account.

In case of success we can remove the file in the temporary directory as it is
stored elsewhere through Web2py mechanism. Storing it twice was space
consuming.

We modifie the log message correspondingly so that it can display the real
stored file.

If a file from the germlines.data doesn't exist in the current directory,
complain about it but provide a possible solution.

Reraising an exception was done following martineau's advices (see
http://stackoverflow.com/a/6062799).

Code was not working without providing the output filename.
Just make it mandatory.

The file will provide us with the list of germline files
that will be used by the browser (eg. IGHV.fa, TRAJ.fa...).

The need to provide them was error prone as we had to indicate
those files in several places (in germlines.data, in the
Makefile, ...)

PDF output has been replaced by HTML output since more than 18 months.

We don't expect make data to install a webserver.
Moreover we have other ways to do so, more reliably.

Use the -m option of tar so that the dates of the
germline files are the extraction date.
This will prevent re-extracting files if they are not newer

Standardize'' the Makefile with all rule and add a .PHONY to avoid ambiguities.

It recovers the files from the *saved* germlines, it does not retrieve the new ones.
This rule has no reason to create the germline.js. That file is created by the
corresponding rule.

It relies on other rules to do so.

DATA is used as a prerequisite in some rules. We cannot use wildcards as
it assumes that the files have already have been retrieved, which is not necessarily the case.

The OTHER_DATA is only used in the make clean, and can make use of the wildcard.

Previously, the informations coming from the best top were kept.
However we could have a sequence in the top 1, with few reads, in a poorly segmented dataset that
was preferred over a top 2 sequence with lots of reads in a highly segmented dataset.

Now we just take into account the number of reads, by assuming with the higher number
of reads we should have the higher reliability. This should be a better criterion than the previous
one (wishful thinking).

For more informations on those variables, see http://distill.pub/2016/misread-tsne/

Sort function doesn't want None values. We avoid that by
defaulting to the minimal date.

Utility function to prevent from having a None value.
This can be useful when giving values from the DB to
functions that only expect defined values (eg. when sorting)

In previous package releases it was unclear to which program the data
storage folders belonged. By default the upload and result folders were
created in /mnt.
They are now created in /mnt/vidjil.

This prevented one of the two tests (test_loading_analysis.rb or test_analysis-example1.rb)
to be analysed, and taken into account, by Jenkins.

It is part of the ECMAScript 2016 standard while indexOf is part of the
ECMAScript 5.1 standard (2011).

Check that the segmented sequence outputs the span at the right places.
For the test we needed to change the sequence (hence its GC content) to
reliably test the positions of interest.

This failed since at least f682e77f (2016/03). This should now be corrected, and tested.

this.seg was always defined (to an empty object) therefore testing
whether its type is undefined was useless.
Instead we rely on a method in Clone to know whether the seg field is defined.

The window (id field) should be found has a potential field