This module converts a molecule object to a string with the formula and back. It registers the 'formula'
format with Chemistry::Mol. Besides its obvious use, it is included in the Chemistry::Mol distribution
because it is a very simple example of a Chemistry::File derived I/O module.
Writingformulas
The format can be specified as a printf-like string with the following control sequences, which are
specified with the formula_format parameter to $mol->print or $mol->write.
%s symbol
%D number of atoms
%d number of atoms, included only when it is greater than one
%d{substr} substr is only included when number of atoms is greater than one
%j{substr} substr is inserted between the formatted string for each element. (The 'j' stands for
'joiner'.) The format should have only one joiner, but its location in the format string doesn't matter.
%% a percent sign
If no format is specified, the default is "%s%d". Some examples follow. Let's assume that the formula is
C2H6O, as it would be formatted by default.
"%s%D"
Like the default, but include explicit indices for all atoms. The formula would be formatted as
"C2H6O1"
"%s%d{<sub>%d</sub>}"
HTML format. The output would be "C<sub>2</sub>H<sub>6</sub>O".
"%D %s%j{, }"
Use a comma followed by a space as a joiner. The output would be "2 C, 6 H, 1 O".
SymbolSortOrder
The elements in the formula are sorted by default in the "Hill order", which means that:
1) if the formula contains carbon, C goes first, followed by H, and the rest of the symbols in
alphabetical order. For example, "CH2BrF".
2) if there is no carbon, all the symbols (including H) are listed alphabetically. For example, "BrH".
It is possible to supply a custom sorting subroutine with the 'formula_sort' option. It expects a
subroutine reference that takes a hash reference describing the formula (similar to what is returned by
parse_formula, discussed below), and that returns a list of symbols in the desired order.
For example, this will sort the symbols in reverse asciibetical order:
my $formula = $mol->print(
format => 'formula',
formula_sort => sub {
my $formula_hash = shift;
return reverse sort keys %$formula_hash;
}
);
ParsingFormulas
Formulas can also be parsed back into Chemistry::Mol objects. The formula may have parentheses and
square or triangular brackets, and it may have the following abbreviations:
Me => '(CH3)',
Et => '(CH3CH2)',
Bu => '(C4H9)',
Bn => '(C6H5CH2)',
Cp => '(C5H5)',
Ph => '(C6H5)',
Bz => '(C6H5CO)',
The formula may also be preceded by a number, which multiplies the whole formula. Some examples of valid
formulas:
Formula Equivalent to
--------------------------------------------------------------
CH3(CH2)3CH3 C5H12
C6H3Me3 C9H12
2Cu[NH3]4(NO3)2 Cu2H24N12O12
2C(C[C<C>5]4)3 C152
2C(C(C(C)5)4)3 C152
C 1 0 H 2 2 C10H22 (whitespace is completely ignored)
When a formula is parsed, a molecule object is created which consists of the set of the atoms in the
formula (no bonds or coordinates, of course). The atoms are created in alphabetical order, so the
molecule object for C2H5Br would have the atoms in the following sequence: Br, C, C, H, H, H, H, H.
If you don't want to create a molecule object, but would rather have a simple hash with the number of
atoms for each element, use the "parse_formula" method:
my %formula = Chemistry::File::Formula->parse_formula("C2H6O");
use Data::Dumper;
print Dumper \%formula;
which prints something like
$VAR1 = {
'H' => 6,
'O' => 1,
'C' => 2
};
The "parse_formula" method is called internally by the "parse_string" method.
Non-integernumbersinformulas
The "parse_formula" method can also accept formulas that contain floating-point numbers, such as
H1.5N0.5. The numbers must be positive, and numbers smaller than one should include a leading zero (e.g.,
0.9, not .9).
When formulas with non-integer numbers of atoms are turned into molecule objects as described in the
previous section, the number of atoms is always roundedup. For example, H1.5N0.5 will produce a molecule
object with two hydrogen atoms and one nitrogen atom.
There is currently no way of producing formulas with non-integer numbers; perhaps a future version will
include an "occupancy" property for atoms that will result in non-integer formulas.