bufr2ncCommon.py

Go to the documentation of this file.

#!/usr/bin/env python3
 
from __future__ import print_function
 
 
###########################################################################
# SUBROUTINES
###########################################################################
 
###########################################################################
# CONSTANTS
###########################################################################
 
# Some handy constants. These become global variables in this script. Using
# naming convention of all caps to remind us that these are not to be changed.
 
# MAX_STRING_LEN is good with 10 characters. This is the length of the long
# format for date and time. Most of the id labels are 6 or 8 characters.
MAX_STRING_LEN = 20
 
# MAX_EVENTS will usually be limited to 255 due to an array size in the Fortran
# interface. In practice, there are typically a handful of events (4 or 5)
# since the events are related to the steps that are gone through to convert a
# raw BUFR file to a prepBUFR file (at NCEP). It is generally accepted that 20
# is a safe limit (instead of 255) for the max number of events, so set
# MAX_EVENTS to 20 to help conserve file space.
MAX_EVENTS = 20
 
# MAX_LEVELS should be limited to 255 by a Fortran array size. This may need to
# change in the future since this number corresponds to the number of
# atmospheric levels in an observation.
MAX_LEVELS = 255
 
# BUFR file types
BFILE_UNDEF = 0
BFILE_BUFR = 1
BFILE_PREPBUFR = 2
 
# BUFR types
BTYPE_UNDEF = 0
BTYPE_HEADER = 1
BTYPE_DATA = 2
BTYPE_EVENT = 3
BTYPE_REP = 4
 
# Data types
DTYPE_UNDEF = 0
DTYPE_STRING = 1   # for CCITT IA5 units in the BUFR table
DTYPE_INTEGER = 2   # for CODE TABLE, FLAG TABLE units in the BUFR table
DTYPE_FLOAT = 3   # for all other units in the BUFR table
DTYPE_UINT = 4   # for dimension coordinates
DTYPE_DOUBLE = 5   # temporary: for strings that are expected to be double
#            in downstream flows (GSI)