Source code for openmdao.util.namelist_util

Utilities for reading and writing Fortran namelists.

# pylint: disable-msg=E0611,F0401
import ordereddict

from numpy import ndarray, array, append, vstack, zeros, \
                  int32, int64, float32, float64
from enthought.traits.trait_base import not_none
from enthought.traits.trait_handlers import TraitListObject 

from pyparsing import CaselessLiteral, Combine, ZeroOrMore, Literal, \
                      Optional, QuotedString, Suppress, Word, alphanums, \
                      oneOf, nums, TokenConverter, Group

from openmdao.util.filewrap import ToFloat, ToInteger

def _floatfmt(val):
    """ Returns the output format for a floating point number.
    The general format is used with 16 places of accuracy, except for when
    the floating point value is an integer, in which case a decimal point
    followed by a single zero is used."""
    if int(val) == val:
        return "%.1f"
        return "%.16g"

def _intfmt(val):
    """ Returns the output format for an integer """
    return "%d"

def _strfmt(val):
    """ Returns the output format for a string """
    return "'%s'"
def _boolfmt(val):
    """ Returns the output format for a boolean """
    if val == True:
        return 'T%.0s'
        return 'F%.0s'
def _process_card_info(card):
    """ Function to extract info from a card as returned from PyParsing a
    namelist file. """
    name =

    # Sometimes we have a 1D array declared by element
    if card.index:
        index = card.index[0]-1
        val = array(card[2:])
        value = zeros(index+len(val))
        value[index:] = val
    # Alternate array specification
    elif card.dimension:
        dim = card.dimension
        value = zeros(dim)
    # Comma-delimited arrays
    elif len(card) > 2:
        value = array(card[1:])
        value = card.value
    return name, value
[docs]class Card(object): """ Data object that stores the value of a single card for a namelist.""" def __init__(self, name, value, is_comment=0): = name self.value = value self.is_comment = is_comment
[docs]class ToBool(TokenConverter): """Converter for PyParsing that is used to turn a token into a Boolean."""
[docs] def postParse( self, instring, loc, tokenlist ): """Converter to make token into a bool.""" if tokenlist[0] in ['T', 'True', 'TRUE', 'true', '.TRUE.']: return True elif tokenlist[0] in ['F', 'False', 'FALSE', 'false', '.FALSE.']: return False else: raise RuntimeError('Unexpected error while trying to identify a' ' Boolean value in the namelist.')
[docs]class Namelist(object): """Utility to ease the task of constructing a formatted output file.""" def __init__(self, comp): self.filename = None self.delimiter = ", " self.terminator = "/" self.title = "" self.comp = comp self.groups = [] = [] self.currentgroup = 0
[docs] def set_filename(self, filename): """Set the name of the file that will be generated or parsed. filename: string Name of the file to be written.""" self.filename = filename
[docs] def set_title(self, title): """Sets the title for the namelist Note that a title is not required. title: string The title card in the namelist - generally optional.""" self.title = title
[docs] def add_group(self, name): """Add a new group to the namelist. Any variables added after this are added to this new group. name: string Group name to be added.""" self.groups.append(name) self.currentgroup = len(self.groups)-1[])
[docs] def add_var(self, varpath): """Add an openmdao variable to the namelist. varpath: string varpath is the dotted path (e.g., comp1.container1.var1).""" paths = varpath.split('.') name = paths[-1] value = self.comp.get(varpath)[self.currentgroup].append(Card(name, value))
[docs] def add_newvar(self, name, value): """Add a new variable to the namelist. name: string Name of the variable to be added. value: int, float, string, ndarray, list, bool Value of the variable to be added."""[self.currentgroup].append(Card(name, value))
[docs] def add_container(self, varpath=''): """Add every variable in an OpenMDAO container to the namelist. This can be used it your component has containers of variables. varpath: string dotted path of container in the data hierarchy""" target_container = self.comp.get(varpath) for name, val in target_container.items(iotype=not_none): self.add_var("%s.%s" % (varpath, name))
[docs] def add_comment(self, comment): """Add a comment in the namelist. comment: string Comment text to be added. Text should include comment character if one is desired. (Note that a comment character isn't always needed in a Namelist. It seems to figure out whether something is a comment without it.)"""[self.currentgroup].append(Card("C", comment, 1))
[docs] def generate(self): """Generates the input file. This should be called after all cards and groups are added to the namelist.""" data = [] data.append("%s\n" % self.title) for i, group_name in enumerate(self.groups): # Groups get a '&', freeform cards don't. if[i]: data.append("&%s\n" % group_name) else: data.append("%s\n" % group_name) for card in[i]: if card.is_comment: line = " %s\n" % (card.value) elif isinstance(card.value, bool): fstring = " %s = " + _boolfmt(card.value) + "\n" line = fstring % (, card.value) elif isinstance(card.value, int): fstring = " %s = " + _intfmt(card.value) + "\n" line = fstring % (, card.value) elif isinstance(card.value, float): fstring = " %s = " + _floatfmt(card.value) + "\n" line = fstring % (, card.value) elif isinstance(card.value, str): fstring = " %s = " + _strfmt(card.value) + "\n" line = fstring % (, card.value) # Lists are mainly supported for the Enum Array elif isinstance(card.value, list): line = " %s = " % ( sep = "" for val in card.value: # We can have integer, real, or string lists if isinstance(val, bool): fmt = _boolfmt elif isinstance(val, (int, int32, int64)): fmt = _intfmt elif isinstance(val, (float, float32, float64)): fmt = _floatfmt else: fmt = _strfmt fstring = sep + fmt(val) line += fstring % val sep = self.delimiter line += "\n" elif isinstance(card.value, (ndarray)): # We can have integer, real, or string arrays if card.value.dtype == bool: fmt = _boolfmt elif card.value.dtype in (int, int32, int64): fmt = _intfmt elif card.value.dtype in (float, float32, float64): fmt = _floatfmt else: fmt = _strfmt # We don't need to output 0D arrays if len(card.value) == 0: continue elif len(card.value.shape) == 1: line = " %s = " % ( sep = "" for val in card.value: fstring = "%s" + fmt(val) line += fstring % (sep, val) sep = self.delimiter line += "\n" elif len(card.value.shape) == 2: line = " " for row in range(0, card.value.shape[0]): line += + "(1," + str(row+1) + ") =" for col in range(0, card.value.shape[1]): val = card.value[row, col] fstring = " " + fmt(val) + "%s" line += fstring % (val, self.delimiter) line += "\n" else: raise RuntimeError("Don't know how to handle array" + \ " of %s dimensions" \ % len(card.value.shape)) else: raise RuntimeError("Error generating input file. Don't" + \ "know how to handle data in variable" + \ "%s in group %s." % (, \ group_name)) data.append(line) # A group with no cards is treated like a free-form entity. if len([i])>0: data.append("%s\n" % self.terminator) outfile = open(self.filename, 'w') outfile.writelines(data) outfile.close()
[docs] def parse_file(self): """Parses an existing namelist file and creates a deck of cards to hold the data. After this is executed, you need to call the load_model() method to extract the variables from this data structure.""" infile = open(self.filename, 'r') data = infile.readlines() infile.close() # Lots of numerical tokens for recognizing various kinds of numbers digits = Word(nums) dot = "." sign = oneOf("+ -") ee = CaselessLiteral('E') | CaselessLiteral('D') num_int = ToInteger(Combine( Optional(sign) + digits )) num_float = ToFloat(Combine( Optional(sign) + ((digits + dot + Optional(digits)) | (dot + digits)) + Optional(ee + Optional(sign) + digits) )) # special case for a float written like "3e5" mixed_exp = ToFloat(Combine( digits + ee + Optional(sign) + digits )) # I don't suppose we need these, but just in case (plus it's easy) nan = ToFloat(oneOf("NaN Inf -Inf")) numval = num_float | mixed_exp | num_int | nan strval = QuotedString(quoteChar='"') | QuotedString(quoteChar="'") b_list = "T TRUE True true F FALSE False false .TRUE. .FALSE." boolval = ToBool(oneOf(b_list)) fieldval = Word(alphanums) # Tokens for parsing a line of data numstr_token = numval + ZeroOrMore(Suppress(',') + numval) \ | strval data_token = numstr_token | boolval index_token = Suppress('(') + num_int + Suppress(')') card_token = Group(fieldval("name") + \ Optional(index_token("index")) + \ Suppress('=') + \ data_token("value") + Optional(Suppress('*') + num_int("dimension"))) multi_card_token = (card_token + ZeroOrMore(Suppress(',') + card_token)) array_continuation_token = numstr_token.setResultsName("value") array2D_token = fieldval("name") + Suppress("(") + \ Suppress(num_int) + Suppress(',') + \ num_int("index") + Suppress(')') + \ Suppress('=') + numval + \ ZeroOrMore(Suppress(',') + numval) # Tokens for parsing the group head and tail group_name_token = (Literal("$") | Literal("&")) + \ Word(alphanums).setResultsName("name") + \ Optional(multi_card_token) group_end_token = Literal("/") | Literal("$END") | Literal("$end") # Comment Token comment_token = Literal("!") # Loop through each line and parse. current_group = None for line in data: line_base = line line = line.strip() # blank line: do nothing if not line: continue if current_group: # Skip comment cards if comment_token.searchString(line): pass # Process orindary cards elif multi_card_token.searchString(line): cards = multi_card_token.parseString(line) for card in cards: name, value = _process_card_info(card)[-1].append(Card(name, value)) # Catch 2D arrays like -> X(1,1) = 3,4,5 elif array2D_token.searchString(line): card = array2D_token.parseString(line) name = card[0] index = card[1] value = array(card[2:]) if index > 1: old_value =[-1][-1].value new_value = vstack((old_value, value))[-1][-1].value = new_value else:[-1].append(Card(name, value)) # Arrays can be continued on subsequent lines # The value of the most recent card must be turned into an # array and appended elif array_continuation_token.searchString(line): card = array_continuation_token.parseString(line) if len(card) > 1: element = array(card[0:]) else: element = card.value if isinstance([-1][-1].value, ndarray): new_value = append([-1][-1].value, element) else: new_value = array([[-1][-1].value, element])[-1][-1].value = new_value # Lastly, look for the group footer elif group_end_token.searchString(line): current_group = None # Everything else must be a pure comment else: print "Comment ignored: %s" % line.rstrip('\n') # Group ending '/' can also conclude a data line. if line[-1] == '/': current_group = None #print[-1][-1].name,[-1][-1].value else: group_name = group_name_token.searchString(line) # Group Header if group_name: group_name = group_name_token.parseString(line) current_group = self.add_group(current_group) # Sometimes, variable definitions are included on the # same line as the namelist header if len(group_name) > 2: cards = group_name[2:] for card in cards: name, value = _process_card_info(card)[-1].append(Card(name, value)) # If there is an ungrouped card at the start, take it as the # title for the analysis elif len( == 0 and self.title == '': self.title = line # All other ungrouped cards are saved as free-form (card-less) # groups. # Note that we can't lstrip because column spacing might be # important. else: self.add_group(line_base.rstrip())
[docs] def load_model(self, rules=None, ignore=None, single_group=-1): """Loads the current deck into an OpenMDAO component. rules: dict of lists of strings (optional) An optional dictionary of rules can be passed if the component has a hierarchy of containers for its input variables. If no rules dictionary is passed, load_model will attempt to find each namelist variable in the top level of the model hierarchy. ignore: list of strings (optional) List of variable names that can safely be ignored. single_group: integer (optional) Group id number to use for processing one single namelist group. Useful if extra processing is needed, or if multiple groups have the same name. Returns a tuple containing the following values: (empty_groups, unlisted_groups, unlinked_vars). These need to be examined after calling load_model to make sure you loaded every variable into your model. empty_groups: ordereddict( integer : string ) Names and ID number of groups that don't have cards. This includes strings found at the top level that aren't comments; these need to be processed by your wrapper to determine how the information fits into your component's variable hierarchy. unlisted_groups: ordereddict( integer : string ) This dictionary includes the names and ID number of groups that have variables that couldn't be loaded because the group wasn't mentioned in the rules dictionary. unlinked_vars: list containing all variable names that weren't found in the component. """ # See Pylint W0102 for why we do this if not ignore: ignore = [] if not self.groups: msg = "Input file must be read with parse_file before " \ "load_model can be executed." raise RuntimeError(msg) if single_group > -1: use_group = {single_group : self.groups[single_group]}.iteritems() else: use_group = enumerate(self.groups) empty_groups = ordereddict.OrderedDict() unlisted_groups = ordereddict.OrderedDict() unlinked_vars = [] used_groups = [] for i, group_name in use_group: # Report all groups with no cards if len([i]) == 0: empty_groups[i] = group_name continue # If a group_name appears twice, we really don't know where to # stick the variables, and there are potential data overwrite # issues. Those cases have to be handled individually. if group_name in used_groups: unlisted_groups[i] = group_name continue else: used_groups.append(group_name) # Process the cards in this group for card in[i]: name = value = card.value found = False if rules: # If the group isn't in the rules dict, we can't handle # it now. A separate function call will be required. try: containers = rules[group_name] except KeyError: unlisted_groups[i] = group_name #print "Group not found: %s" % group_name break for container in containers: # Note: FORTRAN is case-insensitive, OpenMDAO is not varpath1 = "%s.%s" % (container, name) varpath2 = "%s.%s" % (container, name.lower()) if self.comp.contains(varpath1): found = True varpath = varpath1 break if self.comp.contains(varpath2): found = True varpath = varpath2 break else: if self.comp.contains(name): found = True varpath = name elif self.comp.contains(name.lower()): found = True varpath = name.lower() if not found: if name not in ignore and name.lower() not in ignore: unlinked_vars.append(name) print "Variable not found: " + \ "%s in group %s." % (name, group_name) else: # 1D arrays must become ndarrays target = self.comp.get(varpath) if isinstance(target, ndarray) and \ isinstance(value, (float, int)): value = array([value]) # Enum ndarrays must become lists elif isinstance(target, TraitListObject): if isinstance(value, ndarray): value = list(value) else: value = [value] self.comp.set(varpath, value) #print varpath, value return empty_groups, unlisted_groups, unlinked_vars
OpenMDAO Home