Configuration file parser for a somewhat structured properties file format. (The syntax borrows a lot from matjam/StructuredProperties)
The parser is designed to be as small and simple as possible.
The syntax of the config files should resemble standard properties (prop = value) with the additional functionality for structuring the data. A "structprop" file is an UTF-8 file.
Comments start with #
and continue until the end of the line.
Line feeds, spaces and tabs are all treated as white space. There are
three special characters; =
, {
and }
. =
is used to assign a
value to a property. {
and }
are used to encapsulate structured
data.
Arrays are made up out of simple values enclosed inside a {}
-pair.
For example (array with values 1, 2 and "abc").
key = { 1 2 abc }
Since linefeed is like any other whitespace character, the value can be broken up onto several lines:
key = {
1
2
abc
}
The data can be structued into objects (think of it as a hash-map) by
adding a {}
-pair after the key. The content of the object is key-value
pairs or other objects.
name {
key = value
what = foo
}
Keys and values are regular strings that can contain any characters except the ones mentioned earlier (linefeed, space, tab, #, {, }, =) provided that the string is not double-quoted.
key = "a value with a space in it"
The python module contains two functions: loads
and dumps
. loads
parses a string into a python dictionary. dumps
writes a structprop
representation of a python dictionary as a string.
loads
will try to parse values as regular JSON values (the string
"1"
will become an integer with value 1
). If it fails, it will leave
the value as a string.
Example:
# This is a simple example config file
database {
hostname = localhost
username = dbuser
password = secret
port = 12361
database = TheDatabase
}
tables = { Table1 Table2 }