Add support for etc detailed index files (core and synthetic BGC)

[gmaze]

This PR should close #555 providing support for 2 new index files from the etc folder:

• etc/argo-index/argo_profile_detailled_index.txt with columns: file,date,latitude,longitude,ocean,profiler_type,institution,date_update,profile_temp_qc,profile_psal_qc,profile_doxy_qc,ad_psal_adjustment_mean,ad_psal_adjustment_deviation,gdac_date_creation,gdac_date_update,n_levels.
• etc/argo-index/argo_synthetic-profile_detailled_index.txt, with columns: file,date,latitude,longitude,ocean,profiler_type,institution,parameters,parameter_data_mode,parameter_quality,date_update,gdac_date_creation,gdac_date_update,n_levels.

New API

Search profile qc:

idx = ArgoIndex(index_file='core+')
# idx = ArgoIndex(index_file='etc/argo-index/argo_profile_detailled_index.txt')

data = [] 
for qc in ['', '1', 'A', 'B', 'C', 'D', 'E', 'F']:
    data.append({'qc': qc, 'n': idx.query.profile_qc({'TEMP': qc}).N_MATCH})
data

[{'qc': '', 'n': 3298},
 {'qc': '1', 'n': 48},
 {'qc': 'A', 'n': 2601276},
 {'qc': 'B', 'n': 520228},
 {'qc': 'C', 'n': 8250},
 {'qc': 'D', 'n': 3782},
 {'qc': 'E', 'n': 1005},
 {'qc': 'F', 'n': 110495}]

(and yes, '1' should not be there, it's from float 4900566)

Search psal adjustment by bounds:

idx = ArgoIndex(index_file='core+')
idx.query.psal_adj(le=0.1) # Upper bound on mean psal adjustment
idx.query.psal_adj(ge=0.1) # Lower bound on mean psal adjustment
idx.query.psal_adj(ge=.1, le=0.2)

# Also search the adjusment std:
idx.query.psal_adj(le=0.1, where='std') 

Search maximum number of pressure levels contained in a profile:

idx = ArgoIndex(index_file='core+')
idx.query.n_levels(ge=2000) # Lower bound 
idx.query.n_levels(le=100) # Upper bound
idx.query.n_levels(ge=1, le=1000)