=head1	NAME

ODF::lpOD::Common - Common utilities provided by the Perl lpOD library

=head1  DESCRIPTION

This manual page describes miscellaneous functions and auxiliary features of
the lpOD Perl implementation that are not directly linked to the lpOD functional
specification.

=head1  Data conversion and type checking utilities

The utilities introduced in this section are implemented as exported functions.
They may be used without package or object reference.

=head3  is_true(value),

Returns C<TRUE> if the given value may be regarded as true (in the Perl lpOD
implementation). The usual Perl true values are interpreted as C<TRUE>, with a
few exceptions. The strings C<'false'>, C<'no'>, and C<'off'>, like 0,
C<undef>, and the empty string, are regarded as C<FALSE>.

Note that the standard ODF false value is the string C<'false'>; so it's
correctly interpreted as false by C<is_true()>.

The common constants C<TRUE> and C<FALSE> (corresponding to 1 and 0) may be
used by the applications.

=head3  is_false(value)

Returns C<TRUE> if the given value is either a regular Perl false value or one
of the strings C<'false'>, C<'no'>, C<'off'>. Returns C<FALSE> otherwise.

=head3  odf_boolean(value)

Translates into an ODF-compliant boolean value (i.e. C<'true'> or C<'false'>)
an arbitrary value. The result may be used as the value of any ODF
boolean attribute. Examples:

        say odf_boolean(0);             # "false"
        say odf_boolean(undef);         # "false"
        say odf_boolean("All right");   # "true"
        say odf_boolean('no');          # "false"
        say odf_boolean(123);           # "true"

=head3  is_odf_datatype(type)

Returns C<TRUE> if the given argument is the name of a valid ODF data type for
table cells or variable fields, C<FALSE> otherwise. For example, the first
instruction below returns C<TRUE> and the second one returns C<FALSE>:

        $check1 = is_odf_datatype('float');
        $check2 = is_odf_datatype('complexType');

=head3  odf_value(value, type)

Converts the given value according to the given type (which must be a regular
ODF data type), and checks it if the value is already in ODF format.

The following example formats the current system date so the result may be
used as the value of a date field in a document:

        $odf_date = odf_value(time(), 'date');

The following instruction returns C<"2011-09-30T19:55:03">, i.e. it translates
a numeric time into a ISO-8601, ODF-compliant alphanumeric date:

        $d = odf_value(1317405303, 'date');

This function returns C<undef> if the given value is not compatible with the
given type.

I<Note: this function doesn't work for any type in the present version.>

=head3  iso_date(time)

Translates a numeric C<time> into an ISO-8601, ODF-compliant date.

Without argument, returns the current date in ODF-compliant format.

=head3  numeric_date(odf_date)

Translates an ODF-formatted date (ISO-8601) into a Perl computable C<time>
value.

=head3  translate_coordinates(alpha_spreadsheet_coordinates)

Returns the zero-based row number and column number according to a string
representing spreadsheet-like coordinates. For example "A1" gives (0, 0),
"Z10" gives (9, 25), and so on. Note that the row number comes first in the
result, while the column letter comes first in the argument.

=head3  translate_range(alpha_spreadsheet_range)

Returns the zero-based numeric coordinates of the two positions that define
a range in a spreadsheet, according to an alphanumeric range expression. As
an example, "B4:F12" gives (3, 1, 11, 5). Note that the given string must be
separated in two part by a colon; each part is translated according to the
same logic as C<translate_coordinates()>, producing a list of 4 values.

=head1  External file control

These functions are provided for specific physical operations of ODF
document parts.

=head3  file_parse(file_path)

This utility is a variant of the standard C<File::Basename::fileparse>
function, mainly for internal use by lpOD.

In scalar context, extracts the full file name (i.e. including the suffix, if
any) from the given string, that is supposed to be a full path (URL or local).

In array context, returns the path, then the full name (including the suffix,
if any), then the suffix itself (without the dot). Beware that such a behavior
is not exactly the same as C<File::Basename::fileparse>.

The following example will return C<("logo.png", "/usr/share/images/", "png")>:

        ($base, $path, $suffix) = file_parse("/usr/share/images/logo.png");

This one will return C<("mydoc.odt", "http://mydomain.com/", "odt")>:

        ($base, $path, $suffix) = file_parse("http://mydomain.com/mydoc.odt");

This last one will just return C<"mydoc.odt">:

        $base = file_parse("http://mydomain.com/mydoc.odt");

=head3  file_type(file_path)

Returns the MIME type of the resource corresponding to the given file path, or
C<undef> (without error) if the resource is not available, if the C<File::Type>
module is not installed, or if the resource is not supported by C<File::Type>.

I<Beware: This function uses the C<File::Type> logic and doesn't determine the
type according to the file name suffix.>

=head3  image_size(image [, document => $doc])

Returns the size, expressed in I<points> (C<pt>), of the image corresponding to
the given resource, or C<undef> (without error) if the resource is not
available, or if the image type is not supported by C<Image::Size>.

The argument may be a local file path, a regular URL, a file handle, or the
reference of a in-memory buffer (scalar reference). If a C<document> optional
parameter is provided, its value must be a C<odf_document> object and the first
argument is interpreted as a I<document part identifier>; in such a situation,
the image resource is looked for in the given document, allowing the user to
get the orginal size of an image file previously loaded in the ODF file by
another application.

In scalar context, the return value (if defined) is an array ref of 2 strings
(the width and the height), each one containing a numeric value and terminated
by "pt". This array ref may be directly used in order to set the size of a
rectangular shape through the lpOD API (typically, in order to adjust the size
of the object that will contain the image). In array context, C<image_size()>
returns a list of 2 numeric values, namely the width and the height (both in
points, but without the "pt" suffix).

=head1  General configuration

Some methods are provided by the C<lpod> pseudo-object in order to get or set
some configuration parameters.

=head3  Installation information

The C<info> method returns some information about the current lpOD
installation, as a string in scalar context, or as a hash in array context.

For example, the instruction 'say scalar lpod->info' produces an information
string containing "ODF::lpOD", the CPAN package build date, and the current
installation path, like in the example below:

 ODF::lpOD 1.106 2011-02-15T15:07:39 /usr/lib/perl5/site_perl/5.10.0/ODF/lpOD

There is a C<signature> method that produces the same result as the
scalar value of C<info>.

The C<installation_path> method returns the path of the C<ODF::lpOD> module
installation in the user's file system.

=head3  Color name translation

Knowing that lpOD allows the user to specify color codes for various objects,
some codes may be replaced by symbolic names. A few hundreds of symbolic names
and the corresponding values are defined by default, according to a standard
Xorg/RGB vocabulary. The user may add custom color names thanks to
C<load_color_map()>, whose argument is the full path of a RGB text file whose
format complies with the typical Xorg "rgb.txt".

The C<unload_color_map()> removes every color name for the current process,
while C<load_color_map()> without argument restores the default name/code
mapping. C<load_color_map()> may be used repeatedly in order to cumulate
several RGB files.

Two explicit color translation functions are provided:

=over

=item

C<color_code(name)> : returns the color code corresponding to the given color
name, if known; ex: C<color_code('antique white')> produces '#faebd7'; returns
C<undef> if the name is unknown;

=item

C<color_name(code)> : returns a symbolic name, if any, corresponding to the
given color code; ex: C<color_name('#faebd7')> returns 'antique white'; returns
C<undef> if no name is known for this code.

=back

Beware that different names may correspond to the same code, so at the end of
the sequence below, C<$a> may differ from C<$b>:

        $c = color_code($a);
        $b = color_name($c);

=head3  Character sets handling

lpOD is (as soon as possible) locale- and platform- neutral, so its default
input and output character set is always C<utf8> unless the user makes an
explicit alternative selection.

All the text/attribute oriented methods of the C<odf_element> may automatically
convert the processed content from or to the local character set of the user.
The default character set is C<utf8>. If the user provides non-utf8 content,
the input character set must be declared using the C<set_input_charset()>, that
is a lpOD installation method (not a document method). Example:

        lpod->set_input_charset('iso-8859-1');

If the user wants to get non-utf8 outputs from any content-extraction method,
the output character set must be declared in a similar way through
C<set_output_charset()>:

        lpod->set_output_charset('iso-8859-1');

Input and output charsets may be changed at any time, so the user may, for
example, successively insert texts using various encodings (that could prove
useful for document generation from heterogeneous sources).

Beware that the input and output charsets are not always the same. The user
could want, for example, populate a document from non-utf8 web pages, and
in the same session export some content from the same to document to local
log files or the console, that may require C<utf8> or a another character set
(depending of the user's locale).

The currently active character sets may be checked using C<get_input_charset()>
and C<get_output_charset()> (as methods of the C<lpod> pseudo-object).

For the list of supported character sets, see the documentation of the Perl
C<Encode> module.

=head3  Warning information

The "lpod->debug()" method, when called with C<TRUE> of C<FALSE> as argument,
switches on or off the debug flag. If this flag is I<on>, the call stack is
displayed with every error message of the lpOD API.

A so-called C<alert()> function may be used by the applications with one or
more strings as arguments; if the debug flag is C<FALSE>, it just produces a
C<warn> (with a line break) for each argument; if the debug flag is C<TRUE>,
it's a wrapper for C<Carp::cluck()> (so it outputs the call stack trace).

=head1	AUTHOR/COPYRIGHT

Developer/Maintainer: Jean-Marie Gouarne L<http://jean.marie.gouarne.online.fr>
Contact: jmgdoc@cpan.org

Copyright (c) 2010 Ars Aperta, Itaapy, Pierlis, Talend.
Copyright (c) 2014 Jean-Marie Gouarne.

This work was sponsored by the Agence Nationale de la Recherche
(L<http://www.agence-nationale-recherche.fr>).

License: GPL v3, Apache v2.0 (see LICENSE).

=cut