LLOOP Standard Symbols, Tokens and Pre-processors

bool

This defines the following boolean types:

- 2-states boolean: reads 'yes' or 'no'

- 3-states boolean (phi-boolean): 'yes', 'no', 'unknown'

char

Reads a char. Preceding whitespaces are eaten as usual.

double

Reads a "double" C type (including overflow check).

id

Reads a sequence of alphanumeric letters, underscores, colons, dots.

int

Reads a "long" C type (including overflow check).

nil

NO-OP symbol. Does nothing.

quote

Reads any sequence of chars enclosed by double or simple quote chars. Escape chars are supported.

quoteorstring

Reads either a quote token or a string token.

string

Reads any sequence of printable chars (between 0x20 (' ') and 0x7E ('~')), except space chars.

text

Reads any sequence of printable chars (between 0x20 (' ') and 0x7E ('~')).

word

Reads a sequence of letters, underscores and numeric chars starting with a letter. This token is implemented using the standard String class. The chars read from the input are appended to the final string using the appropriate methods of the inherited String class.

file

This defines several tokens related to files:

- file: reads a file name or path, either UNIX or Windows

- existingfile: same as file, except that the file must exist

- directory: same as file, except that the file must be a valid directory if it exists

- existingdir: same as directory, except the directory must exist and must be a directory

- forcedir: same as directory, except the directory is created if it does not exist. An error is raised if it already exists and it is not a directory.

fileunix

Reads a unix file/directory name or path.

Names are separated with slash chars '/'.

Any ASCII char can be used for the names, but the first encountered white space char marks the end of the path.

Whitespaces inside file paths must be banalized using backslash chars '\', or alternatively the full path can be enclosed by quotes. Escape chars are allowed.

filewindows

Reads a windows file/directory name or path. This is quite different from reading a unix path.

Names are separated with backslash chars '\' and no escape sequence is allowed.

Any printable char (0x20 to 0x7E in ASCII) except '/', ',', ':', '*', '?', '"', '<', '>' and '|' can be used for the names, but the first encountered white space char marks the end of the path. File names containing white space chars must be enclosed by double quotes.

image

Reads the name or the path of an existing image.

It also checks whether the file has a valid image file extension, ie. png, jpg, jpeg, gif, ppm, bmp, xpm, xbm, pbm, pcx, pgm, eps.

Raises an error when the file does not exist or when the image extension is not recognised.

urlfile

Reads an URL file, ie. a Universal Resource Locator starting with file://. Platform-specific paths are accepted following the URL prefix.

Example:

 file://readme.txt
 file:///home/myaccount/myfile.txt

ctrl

Control symbols. Control symbols do not parse anything but rather act like parse control functions. There are intended to add some control over parent symbols, like whitespace enabling.

any

Reads any multi-line sequence of chars with an end delimiter (any constant string).

The Any token is a specialization of the BlockToken, whereby the left delimiter is forced to an empty string value.

byte

Any byte, inclusive those standing for whitespace chars.

block

Reads any multi-line sequence of chars surrounded by start and end delimiters (i.e. any constant string).

line

All chars until next new line.

double

Reads a "double" C type (including overflow check).

int

Reads a "long" C type (including overflow check).

quote

Reads any sequence of chars enclosed by double or simple quote chars. Escape chars are supported.

comment_c

Reads a C comment.

comment_cpp

Reads a C++ comment, ie. either a pure C comment or a one-line C++ comment.

ISO_639

This defines two symbols:

- ISO639_2LetterCode: ISO 639 Country 2-Letter Codes

- ISO639_3LetterCode: ISO 639 Country 3-Letter Codes

backslashnewline

Concatenates lines separated with a banalized new line, regardless of used platform line separator (\n, \n\r...).

c_comment

Pre-processes C comments. Strips the input of any C comment, but regardless of whether the comment is within quotes.

cplusplus_comment

Pre-processes C++ comments (//) Strips the input of any C++ // comment, but regardless of whether the comment is within quotes.

dos_comment

Pre-processes MS-DOS comments (') Strips the input data of all ' comments regardless of whether the comment is within quotes. The REM notation is not supported.

define

Pre-processes C #define. Fulfils the same function as a C #define statement.

The static member function "add" allows to add a new define, given the define name and the replacement string

Example:

 #define NAME "James" => define::add("NAME", "\"James\"");

ini_comment

Pre-processes Windows Ini File comments.

Strips the input of all ini comments, i.e. all chars following ; till end of line.

shell_comment

Pre-processes shell script comments.

Strips the input of all shell comments, i.e. all chars following # till end of line.