awips2/pythonPackages/scientific/Scientific/IO/FortranFormat.py
root 9f19e3f712 Initial revision of AWIPS2 11.9.0-7p5
Former-commit-id: 64fa9254b946eae7e61bbc3f513b7c3696c4f54f
2012-01-06 08:55:05 -06:00

323 lines
10 KiB
Python
Executable file

# This module defines a class that handles I/O using
# Fortran-compatible format specifications.
#
#
# Warning: Fortran formatting is a complex business and I don't
# claim that this module works for anything complicated. It knows
# only the most frequent formatting options. Known limitations:
#
# 1) Only A, D, E, F, G, I, and X formats are supported (plus string constants
# for output).
# 2) No direct support for complex numbers. You have to split them into
# real and imaginary parts before output, and for input you get
# two float numbers anyway.
# 3) No overflow check. If an output field gets too large, it will
# take more space, instead of being replaced by stars.
#
#
# Written by Konrad Hinsen <hinsen@cnrs-orleans.fr>
# With contributions from Andreas Prlic <andreas@came.sbg.ac.at>
# last revision: 2008-8-18
#
"""
Fortran-style formatted input/output
This module provides two classes that aid in reading and writing
Fortran-formatted text files.
Examples::
Input::
>>>s = ' 59999'
>>>format = FortranFormat('2I4')
>>>line = FortranLine(s, format)
>>>print line[0]
>>>print line[1]
prints::
>>>5
>>>9999
Output::
>>>format = FortranFormat('2D15.5')
>>>line = FortranLine([3.1415926, 2.71828], format)
>>>print str(line)
prints::
'3.14159D+00 2.71828D+00'
"""
import string
#
# The class FortranLine represents a single line of input/output,
# which can be accessed as text or as a list of items.
#
class FortranLine:
"""Fortran-style record in formatted files
FortranLine objects represent the content of one record of a
Fortran-style formatted file. Indexing yields the contents as
Python objects, whereas transformation to a string (using the
built-in function 'str') yields the text representation.
Restrictions:
1. Only A, D, E, F, G, I, and X formats are supported (plus string
constants for output).
2. No direct support for complex numbers; they must be split into
real and imaginary parts before output.
3. No overflow check. If an output field gets too large, it will
take more space, instead of being replaced by stars according
to Fortran conventions.
"""
def __init__(self, line, format, length = 80):
"""
@param line: either a sequence of Python objects, or a string
formatted according to Fortran rules
@param format: either a Fortran-style format string, or a
L{FortranFormat} object. A FortranFormat should
be used when the same format string is used repeatedly,
because then the rather slow parsing of the string
is performed only once.
@param length: the length of the Fortran record. This is relevant
only when data is a string; this string is then
extended by spaces to have the indicated length.
The default value of 80 is almost always correct.
"""
if type(line) == type(''):
self.text = line
self.data = None
else:
self.text = None
self.data = line
if type(format) == type(''):
self.format = FortranFormat(format)
else:
self.format = format
self.length = length
if self.text is None:
self._output()
if self.data is None:
self._input()
def __len__(self):
"""
@returns: the number of data elements in the record
@rtype: C{int}
"""
return len(self.data)
def __getitem__(self, i):
"""
@param i: index
@type i: C{int}
@returns: the ith data element
"""
return self.data[i]
def __getslice__(self, i, j):
"""
@param i: start index
@type i: C{int}
@param j: end index
@type j: C{int}
@returns: a list containing the ith to jth data elements
"""
return self.data[i:j]
def __str__(self):
"""
@returns: a Fortran-formatted text representation of the data record
@rtype: C{str}
"""
return self.text
def isBlank(self):
"""
@returns: C{True} if the line contains only whitespace
@rtype: C{bool}
"""
return len(string.strip(self.text)) == 0
def _input(self):
text = self.text
if len(text) < self.length: text = text + (self.length-len(text))*' '
self.data = []
for field in self.format:
l = field[1]
s = text[:l]
text = text[l:]
type = field[0]
value = None
if type == 'A':
value = s
elif type == 'I':
s = string.strip(s)
if len(s) == 0:
value = 0
else:
# by AP
# sometimes a line does not match to expected format,
# e.g.: pdb2myd.ent.Z chain: - model: 0 : CONECT*****
# catch this and set value to None
try:
value = string.atoi(s)
except:
value = None
elif type == 'D' or type == 'E' or type == 'F' or type == 'G':
s = string.lower(string.strip(s))
n = string.find(s, 'd')
if n >= 0:
s = s[:n] + 'e' + s[n+1:]
if len(s) == 0:
value = 0.
else:
try:
value = string.atof(s)
except:
value = None
if value is not None:
self.data.append(value)
def _output(self):
data = self.data
self.text = ''
for field in self.format:
type = field[0]
if type == "'":
self.text = self.text + field[1]
elif type == 'X':
self.text = self.text + field[1]*' '
else: # fields that use input data
length = field[1]
if len(field) > 2: fraction = field[2]
value = data[0]
data = data[1:]
if type == 'A':
self.text = self.text + (value+length*' ')[:length]
else: # numeric fields
if value is None:
s = ''
elif type == 'I':
s = `value`
elif type == 'D':
s = ('%'+`length`+'.'+`fraction`+'e') % value
n = string.find(s, 'e')
s = s[:n] + 'D' + s[n+1:]
elif type == 'E':
s = ('%'+`length`+'.'+`fraction`+'e') % value
elif type == 'F':
s = ('%'+`length`+'.'+`fraction`+'f') % value
elif type == 'G':
s = ('%'+`length`+'.'+`fraction`+'g') % value
else:
raise ValueError('Not yet implemented')
s = string.upper(s)
self.text = self.text + ((length*' ')+s)[-length:]
self.text = string.rstrip(self.text)
#
# The class FortranFormat represents a format specification.
# It ought to work for correct specifications, but there is
# little error checking.
#
class FortranFormat:
"""
Parsed Fortran-style format string
FortranFormat objects can be used as arguments when constructing
FortranLine objects instead of the plain format string. If a
format string is used more than once, embedding it into a FortranFormat
object has the advantage that the format string is parsed only once.
"""
def __init__(self, format, nested = False):
"""
@param format: a Fortran format specification
@type format: C{str}
@param nested: I{for internal use}
"""
fields = []
format = string.strip(format)
while format and format[0] != ')':
n = 0
while format[0] in string.digits:
n = 10*n + string.atoi(format[0])
format = format[1:]
if n == 0: n = 1
type = string.upper(format[0])
if type == "'":
eof = string.find(format, "'", 1)
text = format[1:eof]
format = format[eof+1:]
else:
format = string.strip(format[1:])
if type == '(':
subformat = FortranFormat(format, 1)
fields = fields + n*subformat.fields
format = subformat.rest
eof = string.find(format, ',')
if eof >= 0:
format = format[eof+1:]
else:
eof = string.find(format, ',')
if eof >= 0:
field = format[:eof]
format = format[eof+1:]
else:
eof = string.find(format, ')')
if eof >= 0:
field = format[:eof]
format = format[eof+1:]
else:
field = format
format = ''
if type == "'":
field = (type, text)
else:
dot = string.find(field, '.')
if dot > 0:
length = string.atoi(field[:dot])
fraction = string.atoi(field[dot+1:])
field = (type, length, fraction)
else:
if field:
length = string.atoi(field)
else:
length = 1
field = (type, length)
fields = fields + n*[field]
self.fields = fields
if nested:
self.rest = format
def __len__(self):
return len(self.fields)
def __getitem__(self, i):
return self.fields[i]
# Test code
if __name__ == '__main__':
f = FortranFormat("'!!',D10.3,F10.3,G10.3,'!!'")
l = FortranLine([1.5707963, 3.14159265358, 2.71828], f)
print str(l)
f = FortranFormat("F12.0")
l = FortranLine('2.1D2', f)
print l[0]