++ed by:
14 non-PAUSE users
Author image Adam Kennedy


YAML::Tiny - Read/Write YAML files with as little code as possible


The YAML specification is huge. Really, really huge. It contains all the functionality of XML, except with flexibility and choice, which makes it easier to read, but with a formal specification that is more complex than XML.

The original pure-Perl implementation YAML costs just over 4 megabytes of memory to load. Just like with Windows .ini files (3 meg to load) and CSS (3.5 meg to load) the situation is just asking for a YAML::Tiny module, an incomplete but correct and usable subset of the functionality, in as little code as possible.

Like the other ::Tiny modules, YAML::Tiny will have no non-core dependencies, not require a compiler, and be back-compatible to at least perl 5.005_03, and ideally 5.004.


    # In your file
    rootproperty: blah
      one: two
      three: four
      Foo: Bar
      empty: ~
    # In your program
    use YAML::Tiny;
    # Create a YAML file
    my $yaml = YAML::Tiny->new;
    # Open the config
    $yaml = YAML::Tiny->read( 'file.yml' );
    # Reading properties
    my $root = $yaml->[0]->{rootproperty};
    my $one  = $yaml->[0]->{section}->{one};
    my $Foo  = $yaml->[0]->{section}->{Foo};
    # Changing data
    $yaml->[0]->{newsection} = { this => 'that' }; # Add a section
    $yaml->[0]->{section}->{Foo} = 'Not Bar!';     # Change a value
    delete $yaml->[0]->{section};                  # Delete a value or section
    # Add an entire document
    $yaml->[1] = [ 'foo', 'bar', 'baz' ];
    # Save the file
    $yaml->write( 'file.conf' );


YAML::Tiny is a perl class for reading and writing YAML-style files, written with as little code as possible, reducing load time and memory overhead.

Most of the time it is accepted that Perl applications use a lot of memory and modules. The ::Tiny family of modules is specifically intended to provide an ultralight and zero-dependency alternative to many more-thorough standard modules.

This module is primarily for reading human-written files (like simple config files) and generating very simple human-readable files. Note that I said human-readable and not geek-readable. The sort of files that your average manager or secretary should be able to look at and make sense of.

YAML::Tiny does not generate comments, it won't necesarily preserve the order of your hashes, and it will normalise if reading in and writing out again.

It only supports a very basic subset of the full YAML specification.

Usage is targetted at files like Perl's META.yml, for which a small and easily-embeddable module is extremely attractive.

Features will only be added if they are human readable, and can be written in a few lines of code. Please don't be offended if your request is refused. Someone has to draw the line, and for YAML::Tiny that someone is me.

If you need something with more power move up to YAML (4 megabytes of memory overhead) or YAML::Syck (275k, but requires libsyck and a C compiler).

To restate, YAML::Tiny does not preserve your comments, whitespace, or the order of your YAML data. But it should round-trip from Perl structure to file and back again just fine.


This section of the documentation provides a specification for "YAML Tiny", a subset of the YAML specification.

It is based on and described comparatively to the YAML 1.1 Working Draft 2004-12-28 specification, located at http://yaml.org/spec/current.html.

Terminology and chapter numbers are based on that specification.

1. Introduction and Goals

The purpose of the YAML Tiny specification is to describe a useful subset of the YAML specification that can be used for typical document-oriented uses such as configuration files and simple data structure dumps.

Many specification elements that add flexibility or extensibility are intentionally removed, as is support for complex datastructures, class and object-orientation.

In general, YAML Tiny targets only those data structures available in JSON, with the additional limitation that only simple keys are supported.

As a result, all possible YAML Tiny documents should be able to be transformed into an equivalent JSON document, although the reverse is not necesarily true (but will be true in simple cases).

As a result of these simplifications the YAML Tiny specification should be implementable in a relatively small amount of code in any language that supports Perl Compatible Regular Expressions (PCRE).

2. Introduction

YAML Tiny supports three data structures. These are scalars (in a variety of forms), block-form sequences and block-form mappings. Flow-style sequences and mappings are not supported, with some minor exceptions detailed later.

The use of three dashes "---" to indicate the start of a new document is supported, and multiple documents per file/stream is allowed.

Both line and inline comments are supported.

Scalars are supported via the plain style, single quote and double quote, as well as literal-style and folded-style multi-line scalars.

The use of tags is not supported.

The use of anchors and aliases is not supported.

The use of directives is supported only for the %YAML directive.

3. Processing YAML Tiny Information


The YAML specification dictates three-phase serialization and three-phase deserialization.

The YAML Tiny specification does not mandate any particular methodology or mechanism for parsing.

Any compliant parser is only required to parse a single document at a time. The ability to support streaming documents is optional and most likely non-typical.

Because anchors and aliases are not supported, the resulting representation graph is thus directed but (unlike the main YAML specification) acyclic.

Circular references/pointers are not possible, and any YAML Tiny serializer detecting a circulars should error with an appropriate message.

Presentation Stream

YAML Tiny is notionally unicode, but support for unicode is required if the underlying language or system being used to implement a parser does not support Unicode. If unicode is encountered in this case an error should be returned.

Loading Failure Points

YAML Tiny parsers and emitters are not expected to recover from adapt to errors. The specific error modality of any implementation is not dictated (return codes, exceptions, etc) but is expected to be consistant.

4. Syntax

Character Set

YAML Tiny streams are implemented primarily using the ASCII character set, although the use of Unicode inside strings is allowed if support by the implementation.

Specific YAML Tiny encoded document types aiming for maximum compatibility should restrict themselves to ASCII.

The escaping and unescaping of the 8-bit YAML escapes is required.

The escaping and unescaping of 16-bit and 32-bit YAML escapes is not required.

Indicator Characters

Support for the "~" null/undefined indicator is required.

Implementations may represent this as appropriate for the underlying language.

Support for the "-" block sequence indicator is required.

Support for the "?" mapping key indicator is not required.

Support for the ":" mapping value indicator is required.

Support for the "," flow collection indicator is not required.

Support for the "[" flow sequence indicator is not required, with one exception (detailed below).

Support for the "]" flow sequence indicator is not required, with one exception (detailed below).

Support for the "{" flow mapping indicator is not required, with one exception (detailed below).

Support for the "}" flow mapping indicator is not required, with one exception (detailed below).

Support for the "#" comment indicator is required.

Support for the "&" anchor indicator is not required.

Support for the "*" alias indicator is not required.

Support for the "!" tag indicator is not required.

Support for the "|" literal block indicator is required.

Support for the ">" folded block indicator is required.

Support for the "'" single quote indicator is required.

Support for the """ double quote indicator is required.

Support for the "%" directive indicator is required, but only for the special case of a %YAML version directive before the "---" document header, or on the same line as the document header.

For example:

  %YAML 1.1
  - A sequence with a single element

Special Exception:

To provide the ability to support empty sequences and mappings, support for the constructs [] (empty sequence) and {} (empty mapping) are required.

For example,

  %YAML 1.1
  # A document consisting of only an empty mapping
  --- {}
  # A document consisting of only an empty sequence
  --- []
  # A document consisting of an empty mapping within a sequence
  - foo
  - {}
  - bar

Syntax Primitives

Other than the empty sequence and mapping cases described above, YAML Tiny supports only the indentation-based block-style group of contexts.

All five scalar contexts are supported.

Indentation spaces work as per the YAML specification in all cases.

Comments work as per the YAML specification in all simple cases. Support for indented multi-line comments is not required.

Seperation spaces work as per the YAML specification in all cases.

YAML Tiny Character Stream

The only directive supported by the YAML Tiny specification is the %YAML language/version identifier. Although detected, this directive will have no control over the parsing itself.

The parser must recognise both the YAML 1.0 and YAML 1.1+ formatting of this directive (as well as the commented form, although no explicit code should be needed to deal with this case, being a comment anyway)

That is, all of the following should be supported.

  --- #YAML:1.0
  - foo

  - foo

  % YAML 1.1
  - foo

Support for the %TAG directive is not required.

Support for additional directives is not required.

Support for the document boundary marker "---" is required.

Support for the document boundary market "..." is not required.

If necesary, a document boundary should simply by indicated with a "---" marker, with not preceding "..." marker.

Support for empty streams (containing no documents) is required.

Support for implicit document starts is required.

That is, the following must be equivalent.

 # Full form
 %YAML 1.1
 foo: bar

 # Implicit form
 foo: bar


Support for nodes optional anchor and tag properties are not required.

Support for node anchors is not required.

Supprot for node tags is not required.

Support for alias nodes is not required.

Support for flow nodes is not required.

Support for block nodes is required.

Scalar Styles

Support for all five scalar styles are required as per the YAML specification, although support for quoted scalars spanning more than one line is not required.

Support for the chomping indicators on multi-line scalar styles is required.

Collection Styles

Support for block-style sequences is required.

Support for flow-style sequences is not required.

Support for block-style mappings is required.

Support for flow-style mappings is not required.

Both sequences and mappings should be able to be arbitrarily nested.

Support for plain-style mapping keys is required.

Support for quoted keys in mappings is not required.

Support for "?"-indicated explicit keys is not required.

Here endeth the specification.

Additional Perl-Specific Notes

For some Perl applications, it's important to know if you really have a number and not a string.

That is, in some contexts is important that 3 the number is distinctive from "3" the string.

Because even Perl itself is not trivially able to understand the difference (certainly without XS-based modules) Perl implementations of the YAML Tiny specification are not required to retain the distinctiveness of 3 vs "3".



The constructor new creates and returns an empty YAML::Tiny object.

read $filename

The read constructor reads a YAML file, and returns a new YAML::Tiny object containing the contents of the file.

Returns the object on success, or undef on error.

When read fails, YAML::Tiny sets an error message internally you can recover via YAML::Tiny->errstr. Although in some cases a failed read will also set the operating system error variable $!, not all errors do and you should not rely on using the $! variable.

read_string $string;

The read_string method takes as argument the contents of a YAML file (a YAML document) as a string and returns the YAML::Tiny object for it.

write $filename

The write method generates the file content for the properties, and writes it to disk to the filename specified.

Returns true on success or undef on error.


Generates the file content for the object and returns it as a string.


When an error occurs, you can retrieve the error message either from the $YAML::Tiny::errstr variable, or using the errstr() method.


YAML::Tiny implements a number of functions to add compatibility with the YAML API. These should be a drop-in replacement, except that YAML::Tiny will not export functions by default, and so you will need to explicitly import the functions.


  my $string = Dump(list-of-Perl-data-structures);

Turn Perl data into YAML. This function works very much like Data::Dumper::Dumper().

It takes a list of Perl data strucures and dumps them into a serialized form.

It returns a string containing the YAML stream.

The structures can be references or plain scalars.


  my @documents = Load(string-containing-a-YAML-stream);

Turn YAML into Perl data. This is the opposite of Dump.

Just like Storable's thaw() function or the eval() function in relation to Data::Dumper.

It parses a string containing a valid YAML stream into a list of Perl data structures.

freeze() and thaw()

Aliases to Dump() and Load() for Storable fans. This will also allow YAML::Tiny to be plugged directly into modules like POE.pm, that use the freeze/thaw API for internal serialization.

DumpFile(filepath, list)

Writes the YAML stream to a file instead of just returning a string.


Reads the YAML stream from a file instead of a string.


Bugs should be reported via the CPAN bug tracker at


For other issues, or commercial enhancement or support, please contact Adam Kennedy directly.


Adam Kennedy <adamk@cpan.org>


YAML, YAML::Syck, Config::Tiny, CSS::Tiny, http://use.perl.org/~Alias/journal/29427, http://ali.as/


Copyright 2006 - 2009 Adam Kennedy.

This program is free software; you can redistribute it and/or modify it under the same terms as Perl itself.

The full text of the license can be found in the LICENSE file included with this module.