Input Lists and Recall Files

Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.atnf.csiro.au/computing/software/gipsy/hermes/inputlists.html
Дата изменения: Unknown
Дата индексирования: Sat Jan 17 00:58:20 2009
Кодировка:

Prespecifying inputs

If a task requests a keyword repeatedly, the user can pre-specify inputs in two ways. The first method is separating the inputs with semicolons, e.g.

POS= 10 20; 15 25; 30 30

Note that a trailing semicolon designates an empty input for which a default is to be taken.

Recall files

The second method consists of specifying a file as input to the keyword. This must be a text file with a name like ``name.rcl''. Every line in this file is a separate input. Lines must not be longer than 500 characters. Either the whole file or a part of it can be specified:

<name        use the whole file;
<name n:m    use line numbers n to m;
<name :n     use line numbers 1 to n;
<name n:     use line numbers n to the end of the file;
<name n      use only line number n.

Here the n:m range specifier should not be confused with a list of numbers as described in the section about number input.

It is possible to specify a recall file in a semicolon-separated list, but within a recall file semicolons cannot be used. Recall files cannot be used recursively, i.e. a recall file can not contain a reference to an other recall file.

Task behaviour

When the task has ``consumed'' all pre-specified input on a keyword, the next request for that keyword will cause the user to be prompted in the normal way. Inputs given on repeatedly requested keywords are stored in a semicolon-separated list, so if a task is re-run with the previous set of parameters, all inputs are available again.

If unprompted input is given on a keyword of which all pre-specified input has not been consumed yet, the new input will supersede the complete previous set of inputs.