man omake-pervasives (Commandes) - omake is a flexible build system designed for building a wide variety of projects. This document describes the built-in objects available to all programs. For an overview of omake, see the omake(1) man page.

NAME

omake is a flexible build system designed for building a wide variety of projects. This document describes the built-in objects available to all programs. For an overview of omake, see the omake(1) man page.

PERVASIVES

Pervasives defines the objects that are defined in all programs. The following objects are defined.

OBJECT

Parent objects: none.

The Object object is the root object. Every class is a subclass of Object.

It provides the following fields:

*
$(o.object-length): the number of fields and methods in the object.
*
$(o.object-mem <var>): returns true iff the <var> is a field or method of the object.
*
$(o.object-add <var>, <value>): adds the field to the object, returning a new object.
*
$(o.object-find <var>): fetches the field or method from the object; it is equivalent to $(o.<var>), but the variable can be non-constant.
*
$(o.object-map <fun>): maps a function over the object. The function should take two arguments; the first is a field name, the second is the value of that field. The result is a new object constructed from the values returned by the function.
*
o.object-foreach: the foreach form is equivalent to map, but with altered syntax. 

   o.foreach(<var1>, <var2>)
      <body>

For example, the following function prints all the fields of an object o. 

   PrintObject(o) =
      o.foreach(v, x)
         println($(v) = $(x))

The export form is valid in a foreach body. The following function collects just the field names of an object. 

   FieldNames(o) =
      names =
      o.foreach(v, x)
         names += $(v)
         export
      return $(names)

MAP

Parent objects: Object.

A Map object is a dictionary from values to values. The <key> values are restricted to simple values: integers, floating-point numbers, strings, files, directories, and arrays of simple values.

The Map object provides the following methods.

*
$(o.mem <key>): returns true iff the <key> is defined in the map.
*
$(o.add <key>, <value>): adds the field to the map, returning a new map.
*
$(o.find <key>): fetches the field from the map.
*
$(o.map <fun>): maps a function over the map. The function should take two arguments; the first is a field name, the second is the value of that field. The result is a new object constructed from the values returned by the function.
*
o.foreach: the foreach form is equivalent to map, but with altered syntax. 

   o.foreach(<var1>, <var2>)
      <body>

For example, the following function prints all the fields of an object o. 

   PrintObject(o) =
      o.foreach(v, x)
         println($(v) = $(x))

The export form is valid in a foreach body. The following function collects just the field names of the map. 

   FieldNames(o) =
      names =
      o.foreach(v, x)
         names += $(v)
         export
      return $(names)

There is also simpler syntax when the key is a string. The table can be defined using definitions with the form $|key| (the number of pipe symbols | is allowed to vary). 

    $|key 1| = value1
    $||key1|key2|| = value2    # The key is key1|key2
    X = $|key 1|               # Define X to be the value of field $|key 1|

The usual modifiers are also allowed. The expression $`|key| represents lazy evaluation of the key, and $,|key| is normal evaluation.

NUMBER

Parent objects: Object.

The Number object is the parent object for integers and floating-point numbers.

INT

Parent objects: Number.

The Int object represents integer values.

FLOAT

Parent objects: Number.

The Float object represents floating-point numbers.

SEQUENCE

Parent objects: Object.

The Sequence object represents a generic object containing sequential elements. It provides the following methods.

*
$(s.length): the number of elements in the sequence.
*
$(s.map <fun>): maps a function over the fields in the sequence. The function should take two arguments; the first is a field name, the second is the value of that field. The result is a new sequence constructed from the values returned by the function.
*
s.foreach: the foreach form is equivalent to map, but with altered syntax. 

   s.foreach(<var>)
      <body>

For example, the following function prints all the elements of the sequence. 

   PrintSequence(s) =
      s.foreach(x)
         println(Elem = $(x))

The export form is valid in a foreach body. The following function counts the number of zeros in the sequence. 

   Zeros(s) =
      count = $(int 0)
      s.foreach(v)
         if $(equal $(v), 0)
            count = $(add $(count), 1)
            export
         export
      return $(count)

ARRAY

Parent objects: Sequence.

The Array is a random-access sequence. It provides the following additional methods.

*
$(s.nth <i>): returns element i of the sequence.
*
$(s.rev <i>): returns the reversed sequence.

STRING

Parent objects: Array.

FUN

Parent objects: Object.

The Fun object provides the following methods.

*
$(f.arity): the arity if the function.

RULE

Parent objects: Object.

The Rule object represents a build rule. It does not currently have any methods.

TARGET

Parent object: Object.

The Target object contains information collected for a specific target file.

*
target: the target file.
*
effects: the files that may be modified by a side-effect when this target is built.
*
scanner_deps: static dependencies that must be built before this target can be scanned.
*
static-deps: statically-defined build dependencies of this target.
*
build-deps: all the build dependencies for the target, including static and scanned dependencies.
*
build-values: all the value dependencies associated with the build.
*
build-commands: the commands to build the target.

The object supports the following methods.

*
find(file): returns a Target object for the given file. Raises a RuntimeException if the specified target is not part of the project.
*
find-optional(file): returns a Target object for the given file, or false if the file is not part of the project.

NOTE: the information for a target is constructed dynamically, so it is possible that the Target object for a node will contain different values in different contexts. The easiest way to make sure that the Target information is complete is to compute it within a rule body, where the rule depends on the target file, or the dependencies of the target file.

NODE

Parent objects: Object.

The Node object is the parent object for files and directories. It supports the following operations.

*
$(node.stat): returns a stat object for the file. If the file is a symbolic link, the stat information is for the destination of the link, not the link itself.

*
$(node.lstat): returns a stat object for the file or symbolic link.
*
$(node.unlink): removes the file.
*
$(node.rename <file>): renames the file.
*
$(node.link <file>): creates a hard link <dst> to this file.
*
$(node.symlink <file>): create a symbolic link <dst> to this file.
*
$(node.chmod <perm>): change the permission of this file.
*
$(node.chown <uid>, <gid>): change the owner and group id of this file.

FILE

Parent objects: Node.

The file object represents the name of a file.

DIR

Parent objects: Node.

The Dir object represents the name of a directory.

CHANNEL

Parent objects: Object.

A Channel is a generic IO channel. It provides the following methods.

*
$(o.close): close the channel.

INCHANNEL

Parent objects: Channel.

A InChannel is an input channel. The variable stdin is the standard input channel.

It provides the following methods.

*
$(InChannel.fopen <file>): open a new input channel.

OUTCHANNEL

Parent object: Channel.

A OutChannel is an output channel. The variables stdout and stderr are the standard output and error channels.

It provides the following methods.

*
$(OutChannel.fopen <file>): open a new output channel.
*
$(OutChannel.append <file>): opens a new output channel, appending to the file.
*
$(c.flush): flush the output channel.
*
$(c.print <string>): print a string to the channel.
*
$(c.println <string>): print a string to the channel, followed by a line terminator.

LOCATION

Parent objects: Location.

The Location object represents a location in a file.

POSITION

Parent objects: Position.

The Position object represents a stack trace.

EXCEPTION

Parent objects: Object.

The Exception object is used as the base object for exceptions. It has no fields.

RUNTIMEEXCEPTION

Parent objects: Exception.

The RuntimeException object represents an exception from the runtime system. It has the following fields.

*
position: a string representing the location where the exception was raised.
*
message: a string containing the exception message.

SHELL

Parent objects: Object.

The Shell object contains the collection of builtin functions available as shell commands.

You can define aliases by extending this object with additional methods. All methods in this class are called with one argument: a single array containing an argument list.

*
echo

The echo function prints its arguments to the standard output channel.

*
jobs

The jobs method prints the status of currently running commands.

*
cd

The cd function changes the current directory. Note that the current directory follows the usual scoping rules. For example, the following program lists the files in the foo directory, but the current directory is not changed. 

   section
      echo Listing files in the foo directory...
      cd foo
      ls

echo Listing files in the current directory... ls

*
bg

The bg method places a job in the background. The job is resumed if it has been suspended.

*
fg

The fg method brings a job to the foreground. The job is resumed if it has been suspended.

*
stop

The stop method suspends a running job.

*
wait

The wait function waits for a running job to terminate. It is not possible to wait for a suspended job.

The job is not brought to the foreground. If the wait is interrupted, the job continues to run in the background.

*
kill

The kill function signal a job.

kill [signal] <pid...>.

The signals are either numeric, or symbolic. The symbolic signals are named as follows.

ABRT, ALRM, HUP, ILL, KILL, QUIT, SEGV, TERM, USR1, USR2, CHLD, STOP, TSTP, TTIN, TTOU, VTALRM, PROF.

*
exit

The exit function terminates the current session.

*
which, where

See the documentation for the corresponding functions.

*
rehash

Reset the search path.

*
history

Print the current command-line history.

*
Win32 functions.

Win32 doesn't provide very many programs for scripting, except for the functions that are builtin to the DOS cmd.exe. The following functions are defined on Win32 and only on Win32. On other systems, it is expected that these programs already exist.

*
grep 

   grep [-q] [-n] pattern files...

The grep function calls the omake grep function.

By default, omake uses internal versions of the following commands: cp, rm, mkdir, chmod, test, find. If you really want to use the standard system versions of these commands, set the USE_SYSTEM_COMMANDS as one of the first definitions in your OMakeroot file.

*
mkdir 

    mkdir [-m <mode>] [-p] files

The mkdir function is used to create directories. The -verb+-m+ option can be used to specify the permission mode of the created directory. If the -p option is specified, the full path is created.

*
cp
*
mv 

    cp [-f] [-i] [-v] src dst
    cp [-f] [-i] [-v] files dst
    mv [-f] [-i] [-v] src dst
    mv [-f] [-i] [-v] files dst

The cp function copies a src file to a dst file, overwriting it if it already exists. If more than one source file is specified, the final file must be a directory, and the source files are copied into the directory.

-f
Copy files forcibly, do not prompt.
-i
Prompt before removing destination files.
-v
Explain what is happening.

*
rm 

   rm [-f] [-i] [-v] [-r] files
   rmdir [-f] [-i] [-v] [-r] dirs

The rm function removes a set of files. No warnings are issued if the files do not exist, or if they cannot be removed.

Options:

-f
Forcibly remove files, do not prompt.
-i
Prompt before removal.
-v
Explain what is happening.
-r
Remove contents of directories recursively.

*
chmod 

    chmod [-r] [-v] [-f] mode files

The chmod function changes the permissions on a set of files or directories. This function does nothing on Win32. The mode may be specified as an octal number, or in symbolic form [ugoa]*[-=][rwxXstugo]+. See the man page for chmod for details.

Options:

-r
Change permissions of all files in a directory recursively.
-v
Explain what is happening.
-f
Continue on errors.

*
test 

   test \emph{expression}
   \verb+[+ \emph{expression} +]+
   \verb+[ --help+
   \verb+[ --version+

See the documentation for the test function.

*
find 

   find \emph{expression}

See the documentation for the find function.

REFERENCES

SEE ALSO

omake(1), omake-quickstart(1), omake-options(1), omake-root(1), omake-language(1), omake-shell(1), omake-rules(1), omake-base(1), omake-system(1), omake-pervasives(1), osh(1), make(1)

VERSION

Version: 0.9.6.7 of December 28, 2005.

LICENSE AND COPYRIGHT

(C)2003-2005, Jason Hickey, Caltech 256-80, Pasadena, CA 91125, USA

This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.

AUTHOR

Jason Hickey

Caltech 256-80

Pasadena, CA 91125, USA

Email: jyh@cs.caltech.edu

WWW: http://www.cs.caltech.edu/~jyh