multichannel_datasource.py: add options for lvshm data source for customization, needed for O2 replay

lal_transferfunction:  Perform checks on FFTs before computing transfer functions, to minimize loss of calculated stuff.

gstlal-burst: add required deps (fftw, orc, gobject-intro) in configure.ac, fix indent issues in gstlal-burst.spec.in

Added documents dependencies that weren't in the gstlal-dev:stretch-mkl container. Added support for burst. 

add gstlal_feature_synchronizer + gstlal_feature_hdf5_sink to gstlal-burst for saving and synchronization of features from extractor

gstlal_compute_strain:  Simplified adders in state vector calculation. Edited --help output. Corrected ESD line removal.

Makefile.gstlal_idq_icc: fix linking so that gstlal install is not skipped, fixed target for kafka-python install, updated version for kafka python bindings to use seek features, fixed issues with kafka install when Makefile is run multiple times, switched order of pulling of repos so that user is prompted immediately for credentials for idq git pull

gstlal-calibration:  New pipeline gstlal_compute_kappas which only computes the kappas (much cheaper than computing h(t)). It can also apply a filter to DARM_ERR to simulate the effect of time-dependent correction factors (\kappa_tst, \kappa_pu, \kappa_c, f_cc, f_s, Q), so that tests can be done quickly.

inspiral_lr.py: Add population model to ranking statistic. Population model information in inspiral_intrinsics.py.

- pass the SVD bank's PSD to mkwhitened_multirate_src() for use as a phase reference

- allow the zero-latency whitening kernel's phase to be tweaked to approximate the phase response of the kernel corresponding to some reference PSD

- add option to reduce the phase difference between a the FIR filter for the current PSD and the FIR filter for some other reference PSD.  the use case is in reducing the phase discrepancy between a signal and a previously whitened template.

- remove gratuitous [:] slices from expressions
- use .real, .imag array properties instead of numpy.real() numpy.imag()
- move gain & phase calc for zero-latency whitener up a few lines

- can rely on gracedb module import to work

- unused and won't work because services are already advertised

- template bank files are .xml.gz files

- was using "rates" variable with undefined initialization

- the template IDs are stored in the Gamma0 column, which is a float.  the ID values must be ints or the ranking statistic barfs.

gstlal_compute_strain:  added options use either kappa_tst or kappa_pu to compute actuation time delay, instead of handling the timing of the actuation stages individually.