Bycon

* streamlining of internal data retrieval; now the storage objects for
the matched ids are generated at the end of the rewritten query execution pipeline
* `PGXseg class`
- for now only export handling
- to be expanded fro file reading
- needs template based parameter mapping & re-use of `__table_line_from_pgxdoc`
method from `ByconDatatableExporter` etc.
- additional fixes in the ByconBundler class for handling pgzseg imports
(some of it will be refactored to PGXseg later on)
* `PGXfreq` class, used to export CNV frequency files (list or matrix)
* fixed gene position retrieval and `geneId` based queries
- now able to handle multi-gene queries though additional constraints (size,
CNV class) are the same for all genes
* (beacon/biosamples/?geneId=CDKN2A,TP53&variantMinLength=10000&variantMaxLength=20000000&variantType=DEL&filters=NCIT:C3510)[http://progenetix.org/beacon/biosamples/?geneId=CDKN2A,TP53&variantMinLength=10000&variantMaxLength=20000000&variantType=DEL&filters=NCIT:C3510]

* fixed data retrieval for some services where the handover key wasn't defined
for the specific service (e.g. samplemap) and the previous complex remapping
had be removed

This is a test run for a major query module change:

* treating all queries individually in an AND context and aggregating by intyersection
of id values
- e.g. 2 filtering terms with value collision in the same property (e.g. `EFO:0010942,EFO:0010943`
in `biosamples.biosample_status.id`) wwill run as separate internal db queries
and the intersection of values will lead to an empty list for `biosample_id`
but for a list of `individual_id` values for individuals which have both
samples for EFO:0010942 ("primary tumor sample") and EFO:0010943 ("recurrent tumor sample")
- this also applies to variant queries; _i.e._ the so far experimental mullti
variant queries will result in matches for `analysis_id` values for such
with both/multiple variants; or even only at a higher level if multiple
samples ... (e.g. hit on germline and different on the tumor sample in same
individual)
* TODO
- extensive testing (check special queries especially, e.g. geo queries)
- code cleanup
- documentation
- re-introduction of Boolean option ...

* fixed wrong aggregation of `individuals.sex` data (notes in the Progenetix news
after updating ...)
* corrected filter logic where so far terms for the same field were treated as `OR`
(which allows some nice multi-target queries but is in conflict w/ the Beacon behaviour)
* added `type` -> `VRStype` to variants, to be used for query disambiguation
- currently `Adjacency`, `Allele` and `CopyNumberChange` are being used
- query integration to follow
* modified the format for export of annotated variants (e.g. with `MolecularEffects`)
to follow the Beacon schema (some nesting level mismatch before)
* some code cleanup, e.g. adding some initialization functions which aren't used
anywhere else to the root `__init__.py`.
* moved input parameter processing to new `ByconParameters` class
* fixed sample strip plots (some default nesting error)
* rewrite/finishing of Cytobands class and move to `genome_utils`
* schema parsing now in new `ByconSchemas` class

* fixed the all datasets info return for the `/datasets` endpoint

* some refactoring of file and table handling
* changed the URL generation for handovers to always use the protocol sequrity
from the request (http or https)