Pod::Simple::Search - find POD documents in directory trees |
$search->survey( @directories )
$search->simplify_name( $str )
$search->find( $pod )
$search->find( $pod, @search_dirs )
$self->contains_pod( $file )
Pod::Simple::Search - find POD documents in directory trees
use Pod::Simple::Search; my $name2path = Pod::Simple::Search->new->limit_glob('LWP::*')->survey; print "Looky see what I found: ", join(' ', sort keys %$name2path), "\n";
print "LWPUA docs = ", Pod::Simple::Search->new->find('LWP::UserAgent') || "?", "\n";
Pod::Simple::Search is a class that you use for running searches for Pod files. An object of this class has several attributes (mostly options for controlling search options), and some methods for searching based on those attributes.
The way to use this class is to make a new object of this class,
set any options, and then call one of the search options
(probably survey
or find
). The sections below discuss the
syntaxes for doing all that.
This class provides the one constructor, called new
.
It takes no parameters:
use Pod::Simple::Search; my $search = Pod::Simple::Search->new;
This class defines several methods for setting (and, occasionally, reading) the contents of an object. With two exceptions (discussed at the end of this section), these attributes are just for controlling the way searches are carried out.
Note that each of these return $self
when you call them as
$self->whatever(value)
. That's so that you can chain
together set-attribute calls like this:
my $name2path = Pod::Simple::Search->new -> inc(0) -> verbose(1) -> callback(\&blab) ->survey(@there);
...which works exactly as if you'd done this:
my $search = Pod::Simple::Search->new; $search->inc(0); $search->verbose(1); $search->callback(\&blab); my $name2path = $search->survey(@there);
PERL5LIB
environment
as this is prepended to @INC by the Perl interpreter itself.
This attribute's default value is TRUE. If you want to search
only specific directories, set $self->inc(0)
before calling
$inc->survey or $inc->find.
warn
) notes about what they're doing as they do it.
This option may be useful for debugging a pod-related module.
This attribute's default value is zero, meaning that no warn
messages
are produced. (Setting verbose to 1 turns on some messages, and setting
it to 2 turns on even more messages, i.e., makes the following search(es)
even more verbose than 1 would make them.)
("/etc/perljunk/File/Crunk.pm", "File::Crunk")
would
be in @_
.)
The callback routine's return value is not used for anything.
This attribute's default value is false, meaning that no callback is called.
Specifically: Turning on this option will disable the heuristics for seeing only files with Perl-like extensions, omitting subdirectories that are numeric but do not match the current Perl interpreter's version ID, suppressing site_perl as a module hierarchy name, etc.
somepathdir/Foo/Bar.pm
already in this
search, then it won't bother looking at a somelaterpathdir/Foo/Bar.pm
later on in that search, because that file is merely a ``shadow''. But if
you turn on $self->shadows(1)
, then these ``shadow'' files are
inspected too, and are noted in the pathname2podname return hash.
This attribute's default value is false; and normally you won't need to turn it on.
limit_glob
attribute is used instead.
$search->limit_glob("File::Find::R*")
is the same as the combination of the search options
$search->limit_re("^File::Find::R") -> dir_prefix("File::Find")
.
Normally you don't need to know about the dir_prefix
option, but I
include it in case it might prove useful for someone somewhere.
(Implementationally, searching with limit_glob ends up setting limit_re and usually dir_prefix.)
reach
method and a done
method. This is meant for reporting
progress during the search, if you don't want to use a simple
callback.
Normally you don't need to know about the progress
option, but I
include it in case it might prove useful for someone somewhere.
While a search is in progress, the progress object's reach
and
done
methods are called like this:
# Every time a file is being scanned for pod: $progress->reach($count, "Scanning $file"); ++$count;
# And then at the end of the search: $progress->done("Noted $count Pod files total");
Internally, we often set this to an object of class Pod::Simple::Progress. That class is probably undocumented, but you may wish to look at its source.
survey
method, as discussed in the next section.
survey
method, as discussed in the next section.
Once you've actually set any options you want (if any), you can go ahead and use the following methods to search for Pod files in particular ways.
$search->survey( @directories )
The method survey
searches for POD documents in a given set of
files and/or directories. This runs the search according to the various
options set by the accessors above. (For example, if the inc
attribute
is on, as it is by default, then the perl @INC directories are implicitly
added to the list of directories (if any) that you specify.)
The return value of survey
is two hashes:
name2path
path2name
Besides saving these hashes as the hashref attributes
name2path
and path2name
, calling this function also returns
these hashrefs. In list context, the return value of
$search->survey
is the list (\%name2path, \%path2name)
.
In scalar context, the return value is \%name2path
.
Or you can just call this in void context.
Regardless of calling context, calling survey
saves
its results in its name2path
and path2name
attributes.
E.g., when searching in $HOME/perl5lib, the file $HOME/perl5lib/MyModule.pm would get the POD name MyModule, whereas $HOME/perl5lib/Myclass/Subclass.pm would be Myclass::Subclass. The name information can be used for POD translators.
Only text files containing at least one valid POD command are found.
In verbose mode, a warning is printed if shadows are found (i.e., more than one POD file with the same POD name is found, e.g. CPAN.pm in different directories). This usually indicates duplicate occurrences of modules in the @INC search path, which is occasionally inadvertent (but is often simply a case of a user's path dir having a more recent version than the system's general path dirs in general.)
The options to this argument is a list of either directories that are
searched recursively, or files. (Usually you wouldn't specify files,
but just dirs.) Or you can just specify an empty-list, as in
$name2path; with the inc
option on, as it is by default.
The POD names of files are the plain basenames with any Perl-like
extension (.pm, .pl, .pod) stripped, and path separators replaced by
::
's.
Calling Pod::Simple::Search->search(...)
is short for
Pod::Simple::Search->new->search(...). That is, a throwaway object
with default attribute values is used.
$search->simplify_name( $str )
The method simplify_name is equivalent to basename, but also strips Perl-like extensions (.pm, .pl, .pod) and extensions like .bat, .cmd on Win32 and OS/2, or .com on VMS, respectively.
$search->find( $pod )
$search->find( $pod, @search_dirs )
Returns the location of a Pod file, given a Pod/module/script name
(like ``Foo::Bar'' or ``perlvar'' or ``perldoc''), and an idea of
what files/directories to look in.
It searches according to the various options set by the accessors above.
(For example, if the inc
attribute is on, as it is by default, then
the perl @INC directories are implicitly added to the list of
directories (if any) that you specify.)
This returns the full path of the first occurrence to the file. Package names (eg 'A::B') are automatically converted to directory names in the selected directory. Additionally, '.pm', '.pl' and '.pod' are automatically appended to the search as required. (So, for example, under Unix, ``A::B'' is converted to ``somedir/A/B.pm'', ``somedir/A/B.pod'', or ``somedir/A/B.pl'', as appropriate.)
If no such Pod file is found, this method returns undef.
If any of the given search directories contains a pod/ subdirectory, then it is searched. (That's how we manage to find perlfunc, for example, which is usually in pod/perlfunc in most Perl dists.)
The verbose
and inc
attributes influence the behavior of this
search; notably, inc
, if true, adds @INC and also
$Config::Config{'scriptdir'} to the list of directories to search.
It is common to simply say $filename = Pod::Simple::Search-> new
->find("perlvar")
so that just the @INC (well, and scriptdir)
directories are searched. (This happens because the inc
attribute is true by default.)
Calling Pod::Simple::Search->find(...)
is short for
Pod::Simple::Search->new->find(...). That is, a throwaway object
with default attribute values is used.
$self->contains_pod( $file )
Returns true if the supplied filename (not POD module) contains some Pod documentation.
Questions or discussion about POD and Pod::Simple should be sent to the pod-people@perl.org mail list. Send an empty email to pod-people-subscribe@perl.org to subscribe.
This module is managed in an open GitHub repository, https://github.com/perl-pod/pod-simple/. Feel free to fork and contribute, or to clone git://github.com/perl-pod/pod-simple.git and send patches!
Patches against Pod::Simple are welcome. Please send bug reports to <bug-pod-simple@rt.cpan.org>.
Copyright (c) 2002 Sean M. Burke.
This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.
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.
Pod::Simple was created by Sean M. Burke <sburke@cpan.org> with code borrowed
from Marek Rouchal's the Pod::Find manpage, which in turn heavily borrowed code from
Nick Ing-Simmons' PodToHtml
.
But don't bother him, he's retired.
Pod::Simple is maintained by:
Pod::Simple::Search - find POD documents in directory trees |