Free arbitrary software for linux

acoc is a regular-expression based colour formatter for programs that display output on the command-line. It works as a wrapper around the target program, executing it and capturing the stdout stream. Optionally, stderr can be redirected to stdout, so that it, too, can be manipulated. acoc then applies matching rules to patterns in the output and applies colours to those matches.

Ever wondered why the output of your favourite UNIX/Linux commands is still displayed in black-and-white after all these years?

Ever had to search back through your scroll-buffer in search of gcc errors and salient information to tell you what went wrong with your programs execution?

acoc is a regular expression based colour formatter for programs that display output on the command-line. It works as a wrapper around the target program, executing it and capturing the stdout stream. Optionally, stderr can be redirected to stdout, so that it, too, can be manipulated.

acoc then applies matching rules to patterns in the output and applies colour sets to those matches. A picture is worth a thousand words, so look at the sample screenshots in the next section.

Configuration

The configuration files used by the program are /usr/local/etc/acoc.conf, /etc/acoc.conf and ~/acoc.conf. One or more of these must exist. A sample /etc/acoc.conf is supplied with some example matching rules.

Blank lines and those that begin with a # are ignored.

A program configuration stanza is introduced as follows:

[program_spec]

The square brackets are mandatory literal characters. Alternatively, the @ symbol may be used, to allow [ and ] to retain their usual semantics in program specs comprising a regular expression:

@program_spec@

program_spec is defined as one or more instances of the following component, separated by a comma:

invocation[/flags]

where invocation consists of the programs name (not including its directory path component) plus any initial arguments.

Alternatively, invocation may be a regular expression, which can be used to match multiple programs and/or command-line arguments in arbitrary order. Regular expressions are automatically anchored to the beginning of the command line.

flags, if present, is separated from invocation by a slash and consists of one or more of the following characters:

a
continue to attempt to find matching patterns after the first match has been found. By default, acoc will stop processing a line and display it after the first match has been found.
e
redirect the target programs stderr to stdout, allowing it, too, to be matched by rules
p
allocate a pseudo-terminal in which to run the target program

Some programs, such as ls(1), behave differently when their stdout is not connected to a tty. Use of this option will fool the target program into believing it is outputting to a tty, rather than a pipe to acoc.

Use of this flag requires Masahiro Tomitas Ruby/TPty library to be installed. Otherwise, the flag is silently ignored.

Note that the pseudo-terminal communication enabled by this flag is one-way only, from the target program to acoc. It is thus not possible to use acoc in combination with interactive programs, such as the interactive Ruby interpreter (irb).
t
apply colour formatting even if stdout is not a tty. By default, formatting is not applied if the output stream is not attached to a terminal.

Heres an example of a line that introduces a configuration stanza:

[rpm/ae,rpmbuild/ae]

which says to apply the following rules to the rpm and rpmbuild commands, attempt to apply all matching rules, and also apply those rules to the programs stderr stream.

Another example:

[ls/p]

This says to allocate a pseudo-terminal to ls(1), fooling it into believing that its output is being sent to a regular terminal instead of a pipe to acoc.

With this flag, the effect will be this:

$ ls
file1 file2 file3 file4 file5 file6

Without it, ls will detect that its stdout is connected to a pipe and behave accordingly:

$ ls
file1
file2
file3
file4
file5
file6

A third example:

[diff/t,rcsdiff/t,cvs diff/t,p4 diff/t]

This says that the rules that follow should be applied to all invocations of diff(1) and rcsdiff(1), as well as those invocations of cvs(1) and p4 that are followed by the argument diff.

Additionally, colouring should be applied even when stdout is not connected to a tty, so that the colours still show up when the output is displayed in a pager such as more(1) or less(1).

Yet another example:

/ps -.*(e.*f|f.*e)/

In this example, the ps(1) command will be matched, as long as the e and f options are both passed in either order.

An alternative way to write the above spec is:

@ps -.*[ef].*[ef]@

There are two things to note in this alternative:

1. @ has been used to delimit the spec, because [ and ] are required for the character lists in the regular expression.
2. While this form is less specific (in that it allows matches against duplicated command line options), it makes for considerably shorter specs if one wishes to test for the inclusion of a set of more than 2 or 3 command line flags. In the original form, one must manually list all of the possible permutations, which is equal to x! (factorial). For 3 command line flags, this is 6 permutations; for 4, it is 24, etc.

Heres one more example:

[tcpdump/r]

If this were placed in ~/.acoc.conf, it would remove any matching rules that had been installed for the diff command by either /etc/acoc.conf or /usr/local/etc/acoc.conf.

After defining the program name and operational flags, matching rules can be defined. These take the following form:

/regex/[flags] colour_spec

where regex is a Ruby-compatible regular expression. The delimiting / characters can be any character, as long as that character is not present in the regular expression itself. flags, if present, consists of one or more characters from the following list:

g
find every match on the line, not just the first. When using this flag, regex should not include parentheses.

colour_spec is defined as a comma-separated list of one or more colour_groups, which are defined as a plus-separated (+) list of one or more of the following:

* black
* blink
* blue
* bold
* clear
* concealed
* cyan
* dark
* green
* italic
* magenta
* negative
* on_black
* on_blue
* on_cyan
* on_green
* on_magenta
* on_red
* on_white
* on_yellow
* rapid_blink
* red
* reset
* strikethrough
* underline
* underscore
* white
* yellow

Examples of a colour_group are white+bold, black+on_white, etc. A complete colour_spec might look like this:

red+bold,white,yellow+bold,black+on_green

Except when using the g flag, each component of the regex that you wish to colour should be placed in parentheses. Text outside parentheses will be used for matching, but will not be coloured.

For example, examine the following:

/^(d+)foos*(w+)/

This will match a line that starts with more or one digits, followed by the string foo and any amount of white space, followed by one or more word characters. However, only the initial group of digits and the group of word characters will be coloured. The string foo and the white space that follows it will be used for matching, but will not be coloured.

Separated from the regex by white space is the colour_spec. Usually, you will include in this as many colours (separated by commas) as you have parenthesised expressions in the regex. However, its also permissible to have fewer. If, for example, you have three parenthesised expressions in the regex, but only two colours listed in the colour_spec, then the second colour will be used for colouring both the second and third matches.

If you have more colours listed in the colour_spec than there are parenthesised expressions in the regex, the surplus colours are ignored.

When using the g flag to perform a global match on the line, you may list as many colours as you want. The same rules apply here. If there are more matches than colours, the remaining matches will be coloured using the last colour listed. Surplus colours are ignored.

Java DIME Library project implements Direct Internet Message Encapsulation (DIME), which is a lightweight binary encapsulation format that can be used to encapsulate multiple payloads of arbitrary types and sizes into a single message.
DIME is ideal for incorporating binary attachments into any XML-based protocol. The library fully supports the latest DIME specifications.
Enhancements:
- Corrected DimeRecord.isFirst() thanks to John Russell

Generic Hash Table library is a hash table implementation in C for storing arbitrary types of data. Generic Hash Table library is meant to be small, easily extensible (in terms of hash functions etc), and easy to understand codewise.
Enhancements:
- config.guess and config.sub were updated to work with current systems.
- The API was extended with two new functions to allow returning the keysize on iteration.

Math::Roman contains arbitrary sized Roman numbers and conversion from and to Arabic.

SYNOPSIS

use Math::Roman qw(roman);

$a = new Math::Roman MCMLXXIII; # 1973
$b = roman(MCMLXI); # 1961
print $a - $b,"n"; # prints XII

$d = Math::Roman->bzero(); #
$d++; # I
$d += 1998; # MCMXCIX
$d -= MCM; # XCIX

print "$dn"; # string "MCMIC"
print $d->as_number(),"n"; # Math::BigInt "+1999"

Math::Fraction is a Perl module to manipulate exact fractions.
SYNOPSIS
use Math::Fraction;
$a = frac(1,2); $b = frac(6,7);
print "$a + $b = ", $a + $b, "$a * $b = ", $a * $b;
print $a->num;
Main features:
- Being able to add, subtract, multiply, and divide, among other things just like you would normal numbers thats to the overload module.
- Being able to convert a decimal, including repeating ones, into a fraction. For example, 1.142857142857 would become 8/7.
- Being able to control how the fraction is displayed. For example 8/7 verses 1 1/7
- Being able to use arbitrary size numbers in the numerator and the denominator.
- Being able to covert between SMALL (using normal floats/integers) and BIG (using arbitrary size floats/integers) as needed so you do not have to worry about it. (New as of ver .4a)
- Being able to have multiple default sets so that a function can modify the defaults with out effecting other functions (New as of ver .4a)

Data::PropertyList is a Perl module that can convert arbitrary objects to/from strings.

SYNOPSIS

use Data::PropertyList qw(astext fromtext);

$hash_ref = { items => [ 7 .. 11 ], key => value };
$string = astext($hash_ref);
# ...
$hash_ref = fromtext($string);
print $hash_ref->{items}[0];

$array_ref = [ 1, { key => value }, Omega ];
$string = astext($array_ref);
# ...
$array_ref = fromtext($string, -array=>1 );
print $array_ref->[1]{key};

Data::Propertylist provides functions that turn data structures with nested references into NeXTs Property List text format and back again.

You may find this useful for saving and loading application information in text files, or perhaps for generating error messages while debugging.

astext( $reference ) : $propertylist_string;

Writes out a nested Perl data structure in NeXT property list format.

fromtext( $propertylist_string ) : $hash_ref

fromtext( $propertylist_string, -array=>1 ) : $array_ref

Reconstructs a Perl data structure of nested references and scalars from a NeXT property list. Use the -array flag if the string encodes an array rather than a hash.

Mar project is a simple binary file format for storing arbitrary meta and user data.
Both the mar library and the accompanying utility are written in C with the intention of porting to many platforms. A set of C++ wrapper classes (along with a test file that utilises these classes) are included.
These classes provide familar stream and iterator abstractions over the mar library. The package has currently been tested on PCs running FreeBSD, Linux and Windows ME, 2000 & XP. Some testing has also been conducted on architectures running HPUX & Solaris
Enhancements:
- The general portability of the package has been improved.
- Alignment issues have been resolved on non-i386 architectures.
- A Makefile has now been added to the source directory.
- Additional debug output has been added.
- A number of API calls now return MAR_NOMATCH.

fid-graphics project is a simple toolkit for the Frigand Imperial Desktop.
It is probably the easiest toolkit to write a display engine for, but it does not support any widgets.
Enhancements:
- src/Desktop/Fid/Window.lhs (displayWaitRedisplay): New function.
- (requestRedisplay): Tidy up.
- src/Desktop/Fid/Main.lhs (data WinCapLine): New type.
- (data WindowType): Use it.
- (&&*, ||*, isnt, haveEnvVar, haveOpt): Deleted.
- (lookupFeature, lookupProtocol, lookupKeywords): New functions.
- (windowTypes): Use WindowCapLine.
- (forkProcess): Unused; deleted.
- (data FidOpt): Unused; deleted.
- (fidMain): Dont consider forking.
- Only call getOpt to sort out verbose options.
- src/Desktop/Fid/Control.lhs (PrimKey): New type.
- (instance Eq PrimKey, instance Ord PrimKey)
- (instance Show PrimKey, instance Arbitrary PrimKey): Instances for t.
- (data Key): Use it.
- (charKey, control, keyToChar): Do the right thing with PrimKey.
- (symbolKey): New function.
- (instance Arbitrary Key): Use instance Arbitrary PrimKey.
- (instance Arbitrary Char): New Instance.