[ENH] allow fMRIprep input for GLM

[Remi-Gau]

• uses new bids matlab for parsing and querying BIDS files
• adapt bidsCopy to use new bids-matlab
• incoporates rsHRF to work on fmriprep processed resting state data
• renaming of all preprocessed files to be bids compliant
• hence drop the use of SPM prefixes to 'identify' files: almost always done through bids.query
• change how directories have to be specified in the option structure
• add a dry run option
• make fwhm part of the option structure
• adds jupyter book and binder ; fixes Add jupyter notebooks and binder #147
• add quick way to use fieldmap ; fixes Add fieldmap analysis (fast way) #277
• removes parfor loop from certain workflows until a better solution is found ; fixes making for loop the default in spatialPreProcessing script #389
• fix bug with parallel toolbox missing ; fixes check if matlab parallel computing toolbox is present before trying to open a parallel pool #372 Parallel Computing Toolbox bug #379

---

Fixes / Closes Issue

---

TODO

• make sure that a functional mask is created
• test for the bidsRename workflow: standardize how files are renamed:
  • create a cpp_spm specific mapping for spm_2_bids
  • create a test for that
  • then refactor bidsRename
  • will possibly require to refactor spm_2_bids in an object oriented manner
• test bidsWholeBrainFuncMask and some other workflows
• needs ability to filter what confounds are used for the GLM based on what is specified in the model ; fixes make sure that inclusion of the realignement parameters in the GLM follows the specification of the BIDS model json file #125

---

Tests to run

• on an fmriprep dataset
• classic openneuro demos