Configuration file
When running retmap
, the only required argument is the name of the configuration file,
usually called config.in
. This contains all the information needed by the application
to run, such as the location of the needed files and the expected output.
The configuration file is a plain text file structured as:
KEY_WORD argument
where KEY_WORD
represents the option needed to be edited and argument
the respective
option. The distance between the two entries should be at least one empty space. In case
multiple entries are needed (e.g. when multiple data files should be processed at the
same time), the arguments should be added on the same line with at least one space in
between them.
KEY_WORD argument1 argument2 ... argumentN
In case multiple data files are processed at the same time, the user can choose to use either the same number of arguments for all settings or only one argument. In the later case it is assumed that all the arguments should take the same value.
KEY_WORD1 argument1 argument2 ... argumentN
KEY_WORD2 argument1 (argument1) ... (argument1)
There are however a few special keys that do not abide by this rule (e.g. those that denote ranges, prediction parameters, etc.), which will be discussed in their respective parts.
To allow for easy organization of the config.in
file, empty lines can be used to
separate different settings blocks. Also, if the first word of the line is not in the
list of recognized key-words, the line will be ignored. This behaviour allows for
comments to be used inside the configuration file, with little limitations. Still, it
is recommended that comments start with some sequence of symbols (e.g. //
) to ensure
that the first word does not match an implemented key-word. This is also good for
future proofing, as no key-words starting with symbols are planned to be
implemented.
// This is the first block of settings
KEY_WORD1 argument1 argument2 ... argumentN
KEY_WORD2 argument1
// This is the second block of settings
KEY_WORD3 argument1
Both key-words and arguments are case-sensitive, with the former written in uppercase
and the latter in lower case. For the boolean clauses (a.i. those that evaluate for
either True
or False
), yes
represents True
while any other string represents
False
. However, it is recommended that no
is used for false to future-proof the
file.