This is a working draft of an upcoming specification. It is a work in progress and may be modified before final release. You should not rely on the content of this draft or cite it as an authority.

YAML Ain’t Markup Language (YAML™) version 1.3

Revision 1.3.0 (2022-02-06)

Copyright presently by YAML Language Development Team1
Copyright 2001-2009 by Oren Ben-Kiki, Clark Evans, Ingy döt Net

This document may be freely copied, provided it is not modified.

Status of this Document

This is the YAML specification v1.3.0. It defines the YAML 1.3 data language.

The YAML specification is often seen as overly complicated for something which appears to be so simple. Even though YAML often is used for software configuration, it has always been and will continue to be a complete data serialization language. Future YAML plans are focused on making the language and ecosystem more powerful and reliable while simultaneously simplifying the development process for implementers.

In addition to this specification, there is companion documentation intended to guide YAML framework implementers and YAML language users. This documentation can continue to evolve and expand continually between published revisions of this specification.



YAML™ (rhymes with “camel”) is a human-friendly, cross language, Unicode based data serialization language designed around the common native data types of dynamic programming languages. It is broadly useful for programming needs ranging from configuration files to internet messaging to object persistence to data auditing and visualization. Together with the Unicode standard for characters2, this specification provides all the information necessary to understand YAML version 1.3 and to create programs that process YAML information.


Part I. Introduction

Chapter 1. Introduction to YAML

YAML (a recursive acronym for “YAML Ain’t Markup Language”) is a data serialization language designed to be human-friendly and work well with modern programming languages for common everyday tasks. This specification is both an introduction to the YAML language and the concepts supporting it. It is also a complete specification of the information needed to develop applications for processing YAML.

Open, interoperable and readily understandable tools have advanced computing immensely. YAML was designed from the start to be useful and friendly to people working with data. It uses Unicode printable characters, some of which provide structural information and the rest containing the data itself. YAML achieves a unique cleanness by minimizing the amount of structural characters and allowing the data to show itself in a natural and meaningful way. For example, indentation may be used for structure, colons separate key/value pairs and dashes are used to create “bulleted” lists.

There are many kinds of data structures, but they can all be adequately represented with three basic primitives: mappings (hashes/dictionaries), sequences (arrays/lists) and scalars (strings/numbers). YAML leverages these primitives and adds a simple typing system and aliasing mechanism to form a complete language for serializing any native data structure. While most programming languages can use YAML for data serialization, YAML excels in working with those languages that are fundamentally built around the three basic primitives. These include common dynamic languages such as JavaScript, Perl, PHP, Python and Ruby.

There are hundreds of different languages for programming, but only a handful of languages for storing and transferring data. Even though its potential is virtually boundless, YAML was specifically created to work well for common use cases such as: configuration files, log files, interprocess messaging, cross-language data sharing, object persistence and debugging of complex data structures. When data is easy to view and understand, programming becomes a simpler task.

1.1. Goals

The design goals for YAML are, in decreasing priority:

  1. YAML should be easily readable by humans.
  2. YAML data should be portable between programming languages.
  3. YAML should match the native data structures of dynamic languages.
  4. YAML should have a consistent model to support generic tools.
  5. YAML should support one-pass processing.
  6. YAML should be expressive and extensible.
  7. YAML should be easy to implement and use.

1.2. YAML History

The YAML 1.0 specification was published in early 2004 by by Clark Evans, Oren Ben-Kiki, and Ingy döt Net after 3 years of collaborative design work through the yaml-core mailing list3. The project was initially rooted in Clark and Oren’s work on the SML-DEV4 mailing list (for simplifying XML) and Ingy’s plain text serialization module5 for Perl. The language took a lot of inspiration from many other technologies and formats that preceded it.

The first YAML framework was written in Perl in 2001 and Ruby was the first language to ship a YAML framework as part of its core language distribution in 2003.

The YAML 1.16 specification was published in 2005. Around this time, the developers became aware of JSON7. By sheer coincidence, JSON was almost a complete subset of YAML (both syntactically and semantically).

In 2006, Kyrylo Simonov produced PyYAML8 and LibYAML9. A lot of the YAML frameworks in various programming languages are built over LibYAML and many others have looked to PyYAML as a solid reference for their implementations.

The YAML 1.210 specification was published in 2009. Its primary focus was making YAML a strict superset of JSON. It also removed many of the problematic implicit typing recommendations.

Since the release of the 1.2 specification, YAML adoption has continued to grow, and many large-scale projects use it as their primary interface language. In 2020, the new YAML language design team began meeting regularly to discuss improvements to the YAML language and specification; to better meet the needs and expectations of its users and use cases.

The YAML 1.2.2 specification, published in October 2021, was the first step in YAML’s rejuvenated development journey. YAML is now more popular than it has ever been, but there is a long list of things that need to be addressed for it to reach its full potential. The YAML design team is focused on making YAML as good as possible.

1.3. Terminology

The key words “MUST”, “MUST NOT”, “REQUIRED”, “SHALL”, “SHALL NOT”, “SHOULD”, “SHOULD NOT”, “RECOMMENDED”, “MAY”, and “OPTIONAL” in this document are to be interpreted as described in RFC 211911.

The rest of this document is arranged as follows. Chapter 2 provides a short preview of the main YAML features. Chapter 3 describes the YAML information model and the processes for converting from and to this model and the YAML text format. The bulk of the document, chapters 4, 5, 6, 7, 8 and 9, formally define this text format. Finally, chapter 10 recommends basic YAML schemas.

Chapter 2. Language Overview

This section provides a quick glimpse into the expressive power of YAML. It is not expected that the first-time reader grok all of the examples. Rather, these selections are used as motivation for the remainder of the specification.

2.1. Collections

YAML’s block collections use indentation for scope and begin each entry on its own line. Block sequences indicate each entry with a dash and space (“- ”). Mappings use a colon and space (“: ”) to mark each key/value pair. Comments begin with an octothorpe (also called a “hash”, “sharp”, “pound” or “number sign” - “#”).

Example 2.1. Sequence of Scalars (ball players)

- Mark McGwire
- Sammy Sosa
- Ken Griffey

Example 2.2. Mapping Scalars to Scalars (player statistics)

hr:  65    # Home runs
avg: 0.278 # Batting average
rbi: 147   # Runs Batted In

Example 2.3. Mapping Scalars to Sequences (ball clubs in each league)

- Boston Red Sox
- Detroit Tigers
- New York Yankees
- New York Mets
- Chicago Cubs
- Atlanta Braves

Example 2.4. Sequence of Mappings (players’ statistics)

  name: Mark McGwire
  hr:   65
  avg:  0.278
  name: Sammy Sosa
  hr:   63
  avg:  0.288

YAML also has flow styles, using explicit indicators rather than indentation to denote scope. The flow sequence is written as a comma separated list within square brackets. In a similar manner, the flow mapping uses curly braces.

Example 2.5. Sequence of Sequences

- [name        , hr, avg  ]
- [Mark McGwire, 65, 0.278]
- [Sammy Sosa  , 63, 0.288]

Example 2.6. Mapping of Mappings

Mark McGwire: {hr: 65, avg: 0.278}
Sammy Sosa: {
    hr: 63,
    avg: 0.288,

2.2. Structures

YAML uses three dashes (“---”) to separate directives from document content. This also serves to signal the start of a document if no directives are present. Three dots ( “...”) indicate the end of a document without starting a new one, for use in communication channels.

Example 2.7. Two Documents in a Stream (each with a leading comment)

# Ranking of 1998 home runs
- Mark McGwire
- Sammy Sosa
- Ken Griffey

# Team ranking
- Chicago Cubs
- St Louis Cardinals

Example 2.8. Play by Play Feed from a Game

time: 20:03:20
player: Sammy Sosa
action: strike (miss)
time: 20:03:47
player: Sammy Sosa
action: grand slam

Repeated nodes (objects) are first identified by an anchor (marked with the ampersand - “&”) and are then aliased (referenced with an asterisk - “*”) thereafter.

Example 2.9. Single Document with Two Comments

hr: # 1998 hr ranking
- Mark McGwire
- Sammy Sosa
# 1998 rbi ranking
- Sammy Sosa
- Ken Griffey

Example 2.10. Node for “Sammy Sosa” appears twice in this document

- Mark McGwire
# Following node labeled SS
- &SS Sammy Sosa
- *SS # Subsequent occurrence
- Ken Griffey

A question mark and space (“? ”) indicate a complex mapping key. Within a block collection, key/value pairs can start immediately following the dash, colon or question mark.

Example 2.11. Mapping between Sequences

? - Detroit Tigers
  - Chicago cubs
: - 2001-07-23

? [ New York Yankees,
    Atlanta Braves ]
: [ 2001-07-02, 2001-08-12,
    2001-08-14 ]

Example 2.12. Compact Nested Mapping

# Products purchased
- item    : Super Hoop
  quantity: 1
- item    : Basketball
  quantity: 4
- item    : Big Shoes
  quantity: 1

2.3. Scalars

Scalar content can be written in block notation, using a literal style (indicated by “|”) where all line breaks are significant. Alternatively, they can be written with the folded style (denoted by “>”) where each line break is folded to a space unless it ends an empty or a more-indented line.

Example 2.13. In literals, newlines are preserved

--- |
  // ||  ||__

Example 2.14. In the folded scalars, newlines become spaces

--- >
  Mark McGwire's
  year was crippled
  by a knee injury.

Example 2.15. Folded newlines are preserved for “more indented” and blank lines

--- >
 Sammy Sosa completed another
 fine season with great stats.

   63 Home Runs
   0.288 Batting Average

 What a year!

Example 2.16. Indentation determines scope

name: Mark McGwire
accomplishment: >
  Mark set a major league
  home run record in 1998.
stats: |
  65 Home Runs
  0.278 Batting Average

YAML’s flow scalars include the plain style (most examples thus far) and two quoted styles. The double-quoted style provides escape sequences. The single-quoted style is useful when escaping is not needed. All flow scalars can span multiple lines; line breaks are always folded.

Example 2.17. Quoted Scalars

unicode: "Sosa did fine.\u263A"
control: "\b1998\t1999\t2000\n"
hex esc: "\x0d\x0a is \r\n"

single: '"Howdy!" he cried.'
quoted: ' # Not a ''comment''.'
tie-fighter: '|\-*-/|'

Example 2.18. Multi-line Flow Scalars

  This unquoted scalar
  spans many lines.

quoted: "So does this
  quoted scalar.\n"

2.4. Tags

In YAML, untagged nodes are given a type depending on the application. The examples in this specification generally use the seq, map and str types from the fail safe schema. A few examples also use the int, float and null types from the JSON schema.

Example 2.19. Integers

canonical: 12345
decimal: +12345
octal: 0o14
hexadecimal: 0xC

Example 2.20. Floating Point

canonical: 1.23015e+3
exponential: 12.3015e+02
fixed: 1230.15
negative infinity: -.inf
not a number: .nan

Example 2.21. Miscellaneous

booleans: [ true, false ]
string: '012345'

Example 2.22. Timestamps

canonical: 2001-12-15T02:59:43.1Z
iso8601: 2001-12-14t21:59:43.10-05:00
spaced: 2001-12-14 21:59:43.10 -5
date: 2002-12-14

Explicit typing is denoted with a tag using the exclamation point (“!”) symbol. Global tags are URIs and may be specified in a tag shorthand notation using a handle. Application-specific local tags may also be used.

Example 2.23. Various Explicit Tags

not-date: !!str 2002-04-28

picture: !!binary |

application specific tag: !something |
 The semantics of the tag
 above may be different for
 different documents.

Example 2.24. Global Tags

%TAG !,2002:
--- !shape
  # Use the ! handle for presenting
- !circle
  center: &ORIGIN {x: 73, y: 129}
  radius: 7
- !line
  start: *ORIGIN
  finish: { x: 89, y: 102 }
- !label
  start: *ORIGIN
  color: 0xFFEEBB
  text: Pretty vector drawing.

Example 2.25. Unordered Sets

# Sets are represented as a
# Mapping where each key is
# associated with a null value
--- !!set
? Mark McGwire
? Sammy Sosa
? Ken Griffey

Example 2.26. Ordered Mappings

# Ordered maps are represented as
# A sequence of mappings, with
# each mapping having one key
--- !!omap
- Mark McGwire: 65
- Sammy Sosa: 63
- Ken Griffey: 58

2.5. Full Length Example

Below are two full-length examples of YAML. The first is a sample invoice; the second is a sample log file.

Example 2.27. Invoice

--- !<,2002:invoice>
invoice: 34843
date   : 2001-01-23
bill-to: &id001
  given  : Chris
  family : Dumars
    lines: |
      458 Walkman Dr.
      Suite #292
    city    : Royal Oak
    state   : MI
    postal  : 48046
ship-to: *id001
- sku         : BL394D
  quantity    : 4
  description : Basketball
  price       : 450.00
- sku         : BL4438H
  quantity    : 1
  description : Super Hoop
  price       : 2392.00
tax  : 251.42
total: 4443.52
  Late afternoon is best.
  Backup contact is Nancy
  Billsmer @ 338-4338.

Example 2.28. Log File

Time: 2001-11-23 15:01:42 -5
User: ed
  This is an error message
  for the log file
Time: 2001-11-23 15:02:31 -5
User: ed
  A slightly different error
Date: 2001-11-23 15:03:17 -5
User: ed
  Unknown variable "bar"
- file:
  line: 23
  code: |
    x = MoreObject("345\n")
- file:
  line: 58
  code: |-
    foo = bar

2.6. Processes and Models

YAML is both a text format and a method for presenting any native data structure in this format. Therefore, this specification defines two concepts: a class of data objects called YAML representations and a syntax for presenting YAML representations as a series of characters, called a YAML stream.

A YAML processor is a tool for converting information between these complementary views. It is assumed that a YAML processor does its work on behalf of another module, called an application. This chapter describes the information structures a YAML processor must provide to or obtain from the application.

YAML information is used in two ways: for machine processing and for human consumption. The challenge of reconciling these two perspectives is best done in three distinct translation stages: representation, serialization and presentation. Representation addresses how YAML views native data structures to achieve portability between programming environments. Serialization concerns itself with turning a YAML representation into a serial form, that is, a form with sequential access constraints. Presentation deals with the formatting of a YAML serialization as a series of characters in a human-friendly manner.

Part II. Processes

Chapter 3. Processes

Translating between native data structures and a character stream is done in several logically distinct stages, each with a well defined input and output data model, as shown in the following diagram:

Figure 3.1. Processing Overview

Processing Overview

A YAML processor need not expose the serialization or representation stages. It may translate directly between native data structures and a character stream (dump and load in the diagram above). However, such a direct translation should take place so that the native data structures are constructed only from information available in the representation. In particular, mapping key order, comments and tag handles should not be referenced during construction.

3.1. Dump

Dumping native data structures to a character stream is done using the following three stages:

Representing Native Data Structures
YAML represents any native data structure using three node kinds: sequence - an ordered series of entries; mapping - an unordered association of unique keys to values; and scalar - any datum with opaque structure presentable as a series of Unicode characters.

Combined, these primitives generate directed graph structures. These primitives were chosen because they are both powerful and familiar: the sequence corresponds to a Perl array and a Python list, the mapping corresponds to a Perl hash table and a Python dictionary. The scalar represents strings, integers, dates and other atomic data types.

Each YAML node requires, in addition to its kind and content, a tag specifying its data type. Type specifiers are either global URIs or are local in scope to a single application. For example, an integer is represented in YAML with a scalar plus the global,2002:int”. Similarly, an invoice object, particular to a given organization, could be represented as a mapping together with the local tag!invoice”. This simple model can represent any data structure independent of programming language.

Serializing the Representation Graph
For sequential access mediums, such as an event callback API, a YAML representation must be serialized to an ordered tree. Since in a YAML representation, mapping keys are unordered and nodes may be referenced more than once (have more than one incoming “arrow”), the serialization process is required to impose an ordering on the mapping keys and to replace the second and subsequent references to a given node with place holders called aliases. YAML does not specify how these serialization details are chosen. It is up to the YAML processor to come up with human-friendly key order and anchor names, possibly with the help of the application. The result of this process, a YAML serialization tree, can then be traversed to produce a series of event calls for one-pass processing of YAML data.
Presenting the Serialization Tree
The final output process is presenting the YAML serializations as a character stream in a human-friendly manner. To maximize human readability, YAML offers a rich set of stylistic options which go far beyond the minimal functional needs of simple data storage. Therefore the YAML processor is required to introduce various presentation details when creating the stream, such as the choice of node styles, how to format scalar content, the amount of indentation, which tag handles to use, the node tags to leave unspecified, the set of directives to provide and possibly even what comments to add. While some of this can be done with the help of the application, in general this process should be guided by the preferences of the user.

3.2. Load

Loading native data structures from a character stream is done using the following three stages:

Parsing the Presentation Stream
Parsing is the inverse process of presentation, it takes a stream of characters and produces a serialization tree. Parsing discards all the details introduced in the presentation process, reporting only the serialization tree. Parsing can fail due to ill-formed input.
Composing the Representation Graph
Composing takes a serialization tree and produces a representation graph. Composing discards all the details introduced in the serialization process, producing only the representation graph. Composing can fail due to any of several reasons, detailed below.
Constructing Native Data Structures
The final input process is constructing native data structures from the YAML representation. Construction must be based only on the information available in the representation and not on additional serialization or presentation details such as comments, directives, mapping key order, node styles, scalar content format, indentation levels etc. Construction can fail due to the unavailability of the required native data types.

Part III. Information Models

Chapter 4. Information Models

This section specifies the formal details of the results of the above processes. To maximize data portability between programming languages and implementations, users of YAML should be mindful of the distinction between serialization or presentation properties and those which are part of the YAML representation. Thus, while imposing a order on mapping keys is necessary for flattening YAML representations to a sequential access medium, this serialization detail must not be used to convey application level information. In a similar manner, while indentation technique and a choice of a node style are needed for the human readability, these presentation details are neither part of the YAML serialization nor the YAML representation. By carefully separating properties needed for serialization and presentation, YAML representations of application information will be consistent and portable between various programming environments.

The following diagram summarizes the three information models. Full arrows denote composition, hollow arrows denote inheritance, “1” and “*” denote “one” and “many” relationships. A single “+” denotes serialization details, a double “++” denotes presentation details.

Figure 4.1. Information Models

Information Models

4.1. Representation Graph

YAML’s representation of native data structure is a rooted, connected, directed graph of tagged nodes. By “directed graph” we mean a set of nodes and directed edges (“arrows”), where each edge connects one node to another (see a formal directed graph definition12). All the nodes must be reachable from the root node via such edges. Note that the YAML graph may include cycles and a node may have more than one incoming edge.

Nodes that are defined in terms of other nodes are collections; nodes that are independent of any other nodes are scalars. YAML supports two kinds of collection nodes: sequences and mappings. Mapping nodes are somewhat tricky because their keys are unordered and must be unique.

Figure 4.2. Representation Model

Representation Model

4.1.1. Nodes

A YAML node represents a single native data structure. Such nodes have content of one of three kinds: scalar, sequence or mapping. In addition, each node has a tag which serves to restrict the set of possible values the content can have.

The content of a scalar node is an opaque datum that can be presented as a series of zero or more Unicode characters.
The content of a sequence node is an ordered series of zero or more nodes. In particular, a sequence may contain the same node more than once. It could even contain itself.
The content of a mapping node is an unordered set of key/value node pairs, with the restriction that each of the keys is unique. YAML places no further restrictions on the nodes. In particular, keys may be arbitrary nodes, the same node may be used as the value of several key/value pairs and a mapping could even contain itself as a key or a value.

4.1.2. Tags

YAML represents type information of native data structures with a simple identifier, called a tag. Global tags are URIs and hence globally unique across all applications. The “tag:” URI scheme13 is recommended for all global YAML tags. In contrast, local tags are specific to a single application. Local tags start with “!”, are not URIs and are not expected to be globally unique. YAML provides a “TAG” directive to make tag notation less verbose; it also offers easy migration from local to global tags. To ensure this, local tags are restricted to the URI character set and use URI character escaping.

YAML does not mandate any special relationship between different tags that begin with the same substring. Tags ending with URI fragments (containing “#”) are no exception; tags that share the same base URI but differ in their fragment part are considered to be different, independent tags. By convention, fragments are used to identify different “variants” of a tag, while “/” is used to define nested tag “namespace” hierarchies. However, this is merely a convention and each tag may employ its own rules. For example, Perl tags may use “::” to express namespace hierarchies, Java tags may use “.”, etc.

YAML tags are used to associate meta information with each node. In particular, each tag must specify the expected node kind (scalar, sequence or mapping). Scalar tags must also provide a mechanism for converting formatted content to a canonical form for supporting equality testing. Furthermore, a tag may provide additional information such as the set of allowed content values for validation, a mechanism for tag resolution or any other data that is applicable to all of the tag’s nodes.

4.1.3. Node Comparison

Since YAML mappings require key uniqueness, representations must include a mechanism for testing the equality of nodes. This is non-trivial since YAML allows various ways to format scalar content. For example, the integer eleven can be written as “0o13” (octal) or “0xB” (hexadecimal). If both notations are used as keys in the same mapping, only a YAML processor which recognizes integer formats would correctly flag the duplicate key as an error.

Canonical Form
YAML supports the need for scalar equality by requiring that every scalar tag must specify a mechanism for producing the canonical form of any formatted content. This form is a Unicode character string which also presents the same content and can be used for equality testing.
Two nodes must have the same tag and content to be equal. Since each tag applies to exactly one kind, this implies that the two nodes must have the same kind to be equal.

Two scalars are equal only when their tags and canonical forms are equal character-by-character. Equality of collections is defined recursively.

Two sequences are equal only when they have the same tag and length and each node in one sequence is equal to the corresponding node in the other sequence.

Two mappings are equal only when they have the same tag and an equal set of keys and each key in this set is associated with equal values in both mappings.

Different URI schemes may define different rules for testing the equality of URIs. Since a YAML processor cannot be reasonably expected to be aware of them all, it must resort to a simple character-by-character comparison of tags to ensure consistency. This also happens to be the comparison method defined by the “tag:” URI scheme. Tags in a YAML stream must therefore be presented in a canonical way so that such comparison would yield the correct results.

If a node has itself as a descendant (via an alias), then determining the equality of that node is implementation-defined.

A YAML processor may treat equal scalars as if they were identical.

A mapping’s keys are unique if no two keys are equal to each other. Obviously, identical nodes are always considered equal.

4.2. Serialization Tree

To express a YAML representation using a serial API, it is necessary to impose an order on mapping keys and employ alias nodes to indicate a subsequent occurrence of a previously encountered node. The result of this process is a serialization tree, where each node has an ordered set of children. This tree can be traversed for a serial event-based API. Construction of native data structures from the serial interface should not use key order or anchor names for the preservation of application data.

Figure 4.3. Serialization Model

Serialization Model

4.2.1. Mapping Key Order

In the representation model, mapping keys do not have an order. To serialize a mapping, it is necessary to impose an ordering on its keys. This order is a serialization detail and should not be used when composing the representation graph (and hence for the preservation of application data). In every case where node order is significant, a sequence must be used. For example, an ordered mapping can be represented as a sequence of mappings, where each mapping is a single key/value pair. YAML provides convenient compact notation for this case.

4.2.2. Anchors and Aliases

In the representation graph, a node may appear in more than one collection. When serializing such data, the first occurrence of the node is identified by an anchor. Each subsequent occurrence is serialized as an alias node which refers back to this anchor. Otherwise, anchor names are a serialization detail and are discarded once composing is completed. When composing a representation graph from serialized events, an alias event refers to the most recent event in the serialization having the specified anchor. Therefore, anchors need not be unique within a serialization. In addition, an anchor need not have an alias node referring to it.

4.3. Presentation Stream

A YAML presentation is a stream of Unicode characters making use of styles, scalar content formats, comments, directives and other presentation details to present a YAML serialization in a human readable way. YAML allows several serialization trees to be contained in the same YAML presentation stream, as a series of documents separated by markers.

Figure 4.4. Presentation Model

Presentation Model

4.3.1. Node Styles

Each node is presented in some style, depending on its kind. The node style is a presentation detail and is not reflected in the serialization tree or representation graph. There are two groups of styles. Block styles use indentation to denote structure. In contrast, flow styles rely on explicit indicators.

YAML provides a rich set of scalar styles. Block scalar styles include the literal style and the folded style. Flow scalar styles include the plain style and two quoted styles, the single-quoted style and the double-quoted style. These styles offer a range of trade-offs between expressive power and readability.

Normally, block sequences and mappings begin on the next line. In some cases, YAML also allows nested block collections to start in-line for a more compact notation. In addition, YAML provides a compact notation for flow mappings with a single key/value pair, nested inside a flow sequence. These allow for a natural “ordered mapping” notation.

Figure 4.5. Kind/Style Combinations

Kind/Style Combinations

4.3.2. Scalar Formats

YAML allows scalars to be presented in several formats. For example, the integer “11” might also be written as “0xB”. Tags must specify a mechanism for converting the formatted content to a canonical form for use in equality testing. Like node style, the format is a presentation detail and is not reflected in the serialization tree and representation graph.


Comments are a presentation detail and must not have any effect on the serialization tree or representation graph. In particular, comments are not associated with a particular node. The usual purpose of a comment is to communicate between the human maintainers of a file. A typical example is comments in a configuration file. Comments must not appear inside scalars, but may be interleaved with such scalars inside collections.

4.3.4. Directives

Each document may be associated with a set of directives. A directive has a name and an optional sequence of parameters. Directives are instructions to the YAML processor and like all other presentation details are not reflected in the YAML serialization tree or representation graph. This version of YAML defines two directives, “YAML” and “TAG”. All other directives are reserved for future versions of YAML.

Chapter 5. Loading Failure Points

The process of loading native data structures from a YAML stream has several potential failure points. The character stream may be ill-formed, aliases may be unidentified, unspecified tags may be unresolvable, tags may be unrecognized, the content may be invalid, mapping keys may not be unique and a native type may be unavailable. Each of these failures results with an incomplete loading.

A partial representation need not resolve the tag of each node and the canonical form of formatted scalar content need not be available. This weaker representation is useful for cases of incomplete knowledge of the types used in the document.

In contrast, a complete representation specifies the tag of each node and provides the canonical form of formatted scalar content, allowing for equality testing. A complete representation is required in order to construct native data structures.

Figure 5.1. Loading Failure Points

Loading Failure Points

5.1. Well-Formed Streams and Identified Aliases

A well-formed character stream must match the BNF productions specified in the following chapters. Successful loading also requires that each alias shall refer to a previous node identified by the anchor. A YAML processor should reject ill-formed streams and unidentified aliases. A YAML processor may recover from syntax errors, possibly by ignoring certain parts of the input, but it must provide a mechanism for reporting such errors.

5.2. Resolved Tags

Typically, most tags are not explicitly specified in the character stream. During parsing, nodes lacking an explicit tag are given a non-specific tag: “!” for non-plain scalars and “?” for all other nodes. Composing a complete representation requires each such non-specific tag to be resolved to a specific tag, be it a global tag or a local tag.

Resolving the tag of a node must only depend on the following three parameters: (1) the non-specific tag of the node, (2) the path leading from the root to the node and (3) the content (and hence the kind) of the node. When a node has more than one occurrence (using aliases), tag resolution must depend only on the path to the first (anchored) occurrence of the node.

Note that resolution must not consider presentation details such as comments, indentation and node style. Also, resolution must not consider the content of any other node, except for the content of the key nodes directly along the path leading from the root to the resolved node. Finally, resolution must not consider the content of a sibling node in a collection or the content of the value node associated with a key node being resolved.

These rules ensure that tag resolution can be performed as soon as a node is first encountered in the stream, typically before its content is parsed. Also, tag resolution only requires referring to a relatively small number of previously parsed nodes. Thus, in most cases, tag resolution in one-pass processors is both possible and practical.

YAML processors should resolve nodes having the “!” non-specific tag as “,2002:seq”, “,2002:map” or “,2002:str” depending on their kind. This tag resolution convention allows the author of a YAML character stream to effectively “disable” the tag resolution process. By explicitly specifying a “!” non-specific tag property, the node would then be resolved to a “vanilla” sequence, mapping or string, according to its kind.

Application specific tag resolution rules should be restricted to resolving the “?” non-specific tag, most commonly to resolving plain scalars. These may be matched against a set of regular expressions to provide automatic resolution of integers, floats, timestamps and similar types. An application may also match the content of mapping nodes against sets of expected keys to automatically resolve points, complex numbers and similar types. Resolved sequence node types such as the “ordered mapping” are also possible.

That said, tag resolution is specific to the application. YAML processors should therefore provide a mechanism allowing the application to override and expand these default tag resolution rules.

If a document contains unresolved tags, the YAML processor is unable to compose a complete representation graph. In such a case, the YAML processor may compose a partial representation, based on each node’s kind and allowing for non-specific tags.

5.3. Recognized and Valid Tags

To be valid, a node must have a tag which is recognized by the YAML processor and its content must satisfy the constraints imposed by this tag. If a document contains a scalar node with an unrecognized tag or invalid content, only a partial representation may be composed. In contrast, a YAML processor can always compose a complete representation for an unrecognized or an invalid collection, since collection equality does not depend upon knowledge of the collection’s data type. However, such a complete representation cannot be used to construct a native data structure.

5.4. Available Tags

In a given processing environment, there need not be an available native type corresponding to a given tag. If a node’s tag is unavailable, a YAML processor will not be able to construct a native data structure for it. In this case, a complete representation may still be composed and an application may wish to use this representation directly.

Part IV. Syntax

Chapter 6. Character Productions

6.1. Character Set

To ensure readability, YAML streams use only the printable subset of the Unicode character set. The allowed character range explicitly excludes the C0 control block14 x00-x1F (except for TAB x09, LF x0A and CR x0D which are allowed), DEL x7F, the C1 control block x80-x9F (except for NEL x85 which is allowed), the surrogate block15 xD800-xDFFF, xFFFE and xFFFF.

On input, a YAML processor must accept all characters in this printable subset.

On output, a YAML processor must only produce only characters in this printable subset. ? Characters outside this set must be presented using escape sequences. In addition, any allowed characters known to be non-printable should also be escaped.

Note: This isn’t mandatory since a full implementation would require extensive character property tables.

To ensure JSON compatibility, YAML processors must allow all non-C0 characters inside quoted scalars. ? To ensure readability, non-printable characters should be escaped on output, even inside such scalars.

Note: JSON quoted scalars cannot span multiple lines or contain tabs, but YAML quoted scalars can.

6.2. Character Encodings

All characters mentioned in this specification are Unicode code points. Each such code point is written as one or more bytes depending on the character encoding used. Note that in UTF-16, characters above xFFFF are written as four bytes, using a surrogate pair.

The character encoding is a presentation detail and must not be used to convey content information.

On input, a YAML processor must support the UTF-8 and UTF-16 character encodings. For JSON compatibility, the UTF-32 encodings must also be supported.

If a character stream begins with a byte order mark, the character encoding will be taken to be as indicated by the byte order mark. Otherwise, the stream must begin with an ASCII character. This allows the encoding to be deduced by the pattern of null (x00) characters.

Byte order marks may appear at the start of any document, however all documents in the same stream must use the same character encoding.

To allow for JSON compatibility, byte order marks are also allowed inside quoted scalars. For readability, such content byte order marks should be escaped on output.

The encoding can therefore be deduced by matching the first few bytes of the stream with the following table rows (in order):

  Byte0 Byte1 Byte2 Byte3 Encoding
Explicit BOM x00 x00 xFE xFF UTF-32BE
ASCII first character x00 x00 x00 any UTF-32BE
Explicit BOM xFF xFE x00 x00 UTF-32LE
ASCII first character any x00 x00 x00 UTF-32LE
Explicit BOM xFE xFF     UTF-16BE
ASCII first character x00 any     UTF-16BE
Explicit BOM xFF xFE     UTF-16LE
ASCII first character any x00     UTF-16LE
Explicit BOM xEF xBB xBF   UTF-8
Default         UTF-8

The recommended output encoding is UTF-8. If another encoding is used, it is recommended that an explicit byte order mark be used, even if the first stream character is ASCII.

For more information about the byte order mark and the Unicode character encoding schemes see the Unicode FAQ16.

In the examples, byte order mark characters are displayed as “”.

Example 6.1. Byte Order Mark

# Comment only.

# This stream contains no
# documents, only comments.

Example 6.2. Invalid Byte Order Mark

- Invalid use of BOM

- Inside a document.
 A BOM must not appear
 inside a document.

6.3. Indicator Characters

Indicators are characters that have special semantics.

-” (x2D, hyphen) denotes a block sequence entry.

?” (x3F, question mark) denotes a mapping key.

:” (x3A, colon) denotes a mapping value.

Example 6.3. Block Structure Indicators

- one
- two
  ? sky
  : blue
  sea : green
{ "sequence": [
    "two" ],
  "mapping": {
    "sky": "blue",
    "sea": "green" } }


  • sequence entry indicator
  • mapping key
  • mapping value

,” (x2C, comma) ends a flow collection entry.

[” (x5B, left bracket) starts a flow sequence.

]” (x5D, right bracket) ends a flow sequence.

{” (x7B, left brace) starts a flow mapping.

}” (x7D, right brace) ends a flow mapping.

Example 6.4. Flow Collection Indicators

sequence: [ one, two, ]
mapping: { sky: blue, sea: green }
{ "sequence": [ "one", "two" ],
    { "sky": "blue", "sea": "green" } }


  • sequence start, sequence end
  • mapping start, mapping end
  • collection entry

#” (x23, octothorpe, hash, sharp, pound, number sign) denotes a comment.

Example 6.5. Comment Indicator

# Comment only.

# This stream contains no
# documents, only comments.


  • comment

&” (x26, ampersand) denotes a node’s anchor property.

*” (x2A, asterisk) denotes an alias node.

The “!” (x21, exclamation) is used for specifying node tags. It is used to denote tag handles used in tag directives and tag properties; to denote local tags; and as the non-specific tag for non-plain scalars.

Example 6.6. Node Property Indicators

anchored: !local &anchor value
alias: *anchor
{ "anchored": !local &A1 "value",
  "alias": *A1 }


  • tag
  • anchor
  • alias

|” (7C, vertical bar) denotes a literal block scalar.

>” (x3E, greater than) denotes a folded block scalar.

Example 6.7. Block Scalar Indicators

literal: |
folded: >
{ "literal": "some\ntext\n",
  "folded": "some text\n" }


  • literal
  • folded

'” (x27, apostrophe, single quote) surrounds a single-quoted flow scalar.

"” (x22, double quote) surrounds a double-quoted flow scalar.

Example 6.8. Quoted Scalar Indicators

single: 'text'
double: "text"
{ "single": "text",
  "double": "text" }


  • single quote
  • double quote

%” (x25, percent) denotes a directive line.

Example 6.9. Directive Indicator

%YAML 1.3
--- text


  • directive

The “@” (x40, at) and “`” (x60, grave accent) are reserved for future use.

Example 6.10. Invalid use of Reserved Indicators

commercial-at: @text
grave-accent: `text
 Reserved indicators can't
 start a plain scalar.

The “[”, “]”, “{”, “}” and “,” indicators denote structure in flow collections. They are therefore forbidden in some cases, to avoid ambiguity in several constructs. This is handled on a case-by-case basis by the relevant productions.

6.4. Line Break Characters

The line feed (x0A) and carriage return (0x0D) are line break characters.

All other characters, including the form feed (x0C), are considered to be non-break characters. Note that these include the non-ASCII line breaks: next line (x85), line separator (x2028) and paragraph separator (x2029).

YAML version 1.1 did support the above non-ASCII line break characters; however, JSON does not. Hence, to ensure JSON compatibility, YAML treats them as non-break characters as of version 1.2. YAML 1.3 processors parsing a version 1.1 document should therefore treat these line breaks as non-break characters, with an appropriate warning.

Line breaks are interpreted differently by different systems and have multiple widely used formats.

Line breaks inside scalar content must be normalized by the YAML processor. Each such line break must be parsed into a single line feed character. The original line break format is a presentation detail and must not be used to convey content information.

Outside scalar content, YAML allows any line break to be used to terminate lines.

On output, a YAML processor is free to emit line breaks using whatever convention is most appropriate.

In the examples, line breaks are sometimes displayed using the “” glyph for clarity.

Example 6.11. Line Break Characters

  Line break (no glyph)
  Line break (glyphed)

"Line break (no glyph)\nLine break (glyphed)\n"


6.5. White Space Characters

YAML recognizes two white space characters: space and tab.

The rest of the (printable) non-break characters are considered to be non-space characters.

In the examples, tab characters are displayed as the glyph “”. Space characters are sometimes displayed as the glyph “·” for clarity.

Example 6.12. Tabs and Spaces

# Tabs and spaces
quoted:·"Quoted "
··void main() {
··printf("Hello, world!\n");
{ "quoted": "Quoted \t",
  "block": "void main()
    {\n\tprintf(\"Hello, world!\\n\");\n}\n" }


  • space
  • tab

6.6. Miscellaneous Characters

The YAML syntax productions make use of the following additional character classes:

  • A decimal digit for numbers.
  • A hexadecimal digit for escape sequences.
  • ASCII letter (alphabetic) characters.
  • Word (alphanumeric) characters for identifiers.
  • URI characters for tags, as defined in the URI specification17.

By convention, any URI characters other than the allowed printable ASCII characters are first encoded in UTF-8 and then each byte is escaped using the “%” character. The YAML processor must not expand such escaped characters. Tag characters must be preserved and compared exactly as presented in the YAML stream, without any processing.

The “!” character is used to indicate the end of a named tag handle; hence its use in tag shorthands is restricted. In addition, such shorthands must not contain the “[”, “]”, “{”, “}” and “,” characters. These characters would cause ambiguity with flow collection structures.

6.7. Escaped Characters

All non-printable characters must be escaped. YAML escape sequences use the “\” notation common to most modern computer languages. Each escape sequence must be parsed into the appropriate Unicode character. The original escape sequence is a presentation detail and must not be used to convey content information.

Note that escape sequences are only interpreted in double-quoted scalars. In all other scalar styles, the “\” character has no special meaning and non-printable characters are not available.

YAML escape sequences are a superset of C’s escape sequences:

Escape Character
\0 Null (x00)
\a Bell (x07)
\b Backspace (x08)
\t Horizontal Tabulation (x09)
\n New Line (x0A)
\v Vertical Tabulation (x0B)
\f Form Feed (x0C)
\r Carriage Return (x0D)
\e Escape (x1B)
\ Space (x20)
\" Quotation Mark (x22)
\/ Solidus (x22)
\\ Reverse Solidus (x22)
\N Next Line (x85)
\_ No-Break Space (x85)
\L Line Separator (x2028)
\P Paragraph Separator (x2029)

An escaped tab character (x09) evaluates to a tab character.

The \x, \u, and \U escape sequences are followed by two, four, and eight hexadecimal digits respectively. Those digits are interpreted as a Unicode code point. The value of the escape is the character at that code point.

Example 6.13. Escaped Characters

- "Fun with \\"
- "\" \a \b \e \f"
- "\n \r \t \v \0"
- "\  \_ \N \L \P \
  \x41 \u0041 \U00000041"
[ "Fun with \\",
  "\" \u0007 \b \u001b \f",
  "\n \r \t \u000b \u0000",
  "\u0020 \u00a0 \u0085 \u2028 \u2029 A A A" ]

Example 6.14. Invalid Escaped Characters

Bad escapes:
- c is an invalid escaped character.
- q and - are invalid hex digits.

Chapter 7. Structural Productions

7.1. Indentation Spaces

In YAML block styles, structure is determined by indentation. In general, indentation is defined as a zero or more space characters at the start of a line. ?

To maintain portability, tab characters must not be used in indentation, since different systems treat tabs differently. Note that most modern editors may be configured so that pressing the tab key results in the insertion of an appropriate number of spaces.

The amount of indentation is a presentation detail and must not be used to convey content information.

A block style construct is terminated when encountering a line which is less indented than the construct. The productions use the notation “indentation-spaces-less-than(n)” and “indentation-spaces-less-or-equal(n)” to express this.

Each node must be indented further than its parent node. All sibling nodes must use the exact same indentation level. However the content of each sibling node may be further indented independently.

Example 7.1. Indentation Spaces

··# Leading comment line spaces are
···# neither content nor indentation.
Not indented:
·By one space: |
····By four
·Flow style: [    # Leading spaces
···By two,        # in flow style
··Also by two,    # are neither
··→Still by two   # content nor
····]             # indentation.
{ "Not indented": {
    "By one space": "By four\n  spaces\n",
    "Flow style": [
      "By two",
      "Also by two",
      "Still by two" ] } }


The “-”, “?” and “:” characters used to denote block collection entries are perceived by people to be part of the indentation. This is handled on a case-by-case basis by the relevant productions.

Example 7.2. Indentation Indicators

{ "a":
  [ "b",
    [ "c",
      "d" ] ] }


7.2. Separation Spaces

Outside indentation and scalar content, YAML uses white space characters for separation between tokens within a line. Note that such white space may safely include tab characters.

Separation spaces are a presentation detail and must not be used to convey content information.

Example 7.3. Separation Spaces

- -·baz
[ { "foo": "bar" },
  [ "baz",
    "baz" ] ]

7.3. Line Prefixes

Inside scalar content, each line begins with a non-content line prefix. This prefix always includes the indentation. For flow scalar styles it additionally includes all leading white space, which may contain tab characters.

Line prefixes are a presentation detail and must not be used to convey content information.

Example 7.4. Line Prefixes

plain: text
quoted: "text
block: |
{ "plain": "text lines",
  "quoted": "text lines",
  "block": "text\n \tlines\n" }

7.4. Empty Lines

An empty line line consists of the non-content prefix followed by a line break.

The semantics of empty lines depend on the scalar style they appear in. This is handled on a case-by-case basis by the relevant productions.

Example 7.5. Empty Lines

  "Empty line
  as a line feed"
Chomping: |
  Clipped empty lines
{ "Folding": "Empty line\nas a line feed",
  "Chomping": "Clipped empty lines\n" }

7.5. Line Folding

Line folding allows long lines to be broken for readability, while retaining the semantics of the original long line. If a line break is followed by an empty line, it is trimmed; the first line break is discarded and the rest are retained as content.

Otherwise (the following line is not empty), the line break is converted to a single space (x20).

A folded non-empty line may end with either of the above line breaks.

Example 7.6. Line Folding


"trimmed\n\n\nas space"


The above rules are common to both the folded block style and the scalar flow styles. Folding does distinguish between these cases in the following way:

Block Folding
In the folded block style, the final line break and trailing empty lines are subject to chomping and are never folded. In addition, folding does not apply to line breaks surrounding text lines that contain leading white space. Note that such a more-indented line may consist only of such leading white space.

The combined effect of the block line folding rules is that each “paragraph” is interpreted as a line, empty lines are interpreted as a line feed and the formatting of more-indented lines is preserved.

Example 7.7. Block Folding


"foo \n\n\t bar\n\nbaz\n"


Flow Folding
Folding in flow styles provides more relaxed semantics. Flow styles typically depend on explicit indicators rather than indentation to convey structure. Hence spaces preceding or following the text in a line are a presentation detail and must not be used to convey content information. Once all such spaces have been discarded, all line breaks are folded without exception.

The combined effect of the flow line folding rules is that each “paragraph” is interpreted as a line, empty lines are interpreted as line feeds and text can be freely more-indented without affecting the content information.

Example 7.8. Flow Folding


··baz "
" foo\nbar\nbaz "



An explicit comment is marked by a “#” indicator. Comments are a presentation detail and must not be used to convey content information.

Comments must be separated from other tokens by white space characters.

Note: To ensure JSON compatibility, YAML processors must allow for the omission of the final comment line break of the input stream. However, as this confuses many tools, YAML processors should terminate the stream with an explicit line break on output.

In this example, the end of the stream is displayed using the “” glyph for clarity.

Example 7.9. Separated Comment

key:····# Comment
{ "key": "value" }

Outside scalar content, comments may appear on a line of their own, independent of the indentation level. Note that outside scalar content, a line containing only white space characters is taken to be a comment line.

Example 7.10. Comment Lines

··# Comment↵

# This stream contains no
# documents, only comments.

In most cases, when a line may end with a comment, YAML allows it to be followed by additional comment lines. The only exception is a comment ending a block scalar header.

Example 7.11. Multi-Line Comments

key:····# Comment↵
········# lines↵

{ "key": "value" }

7.7. Separation Lines

Implicit keys are restricted to a single line. In all other cases, YAML allows tokens to be separated by multi-line (possibly empty) comments.

Note that structures following multi-line comment separation must be properly indented, even though there is no such restriction on the separation comment lines themselves.

Example 7.12. Separation Lines

# Statistics:
··hr:··# Home runs
··avg:·# Average
{ { "first": "Sammy",
    "last": "Sosa" }: {
    "hr": 65,
    "avg": 0.278 } }

7.8. Directives

Directives are instructions to the YAML processor. This specification defines two directives, “YAML” and “TAG”, and reserves all other directives for future use. There is no way to define private directives. This is intentional.

Directives are a presentation detail and must not be used to convey content information.

Each directive is specified on a separate non-indented line starting with the “%” indicator, followed by the directive name and a list of parameters. The semantics of these parameters depends on the specific directive. A YAML processor should ignore unknown directives with an appropriate warning.

Example 7.13. Reserved Directives

%FOO  bar baz # Should be ignored
               # with a warning.
--- "foo"

7.8.1. “YAML” Directives

The “YAML” directive specifies the version of YAML the document conforms to. This specification defines version “1.3”, including recommendations for YAML 1.1 processing.

A version 1.3 YAML processor must accept documents with an explicit “%YAML 1.3” directive, as well as documents lacking a “YAML” directive. Such documents are assumed to conform to the 1.3 version specification. Documents with a “YAML” directive specifying a higher minor version (e.g. “%YAML 1.3”) should be processed with an appropriate warning. Documents with a “YAML” directive specifying a higher major version (e.g. “%YAML 2.0”) should be rejected with an appropriate error message.

A version 1.3 YAML processor must also accept documents with an explicit “%YAML 1.1” directive. Note that version 1.3 is mostly a superset of version 1.1, defined for the purpose of ensuring JSON compatibility. Hence a version 1.3 processor should process version 1.1 documents as if they were version 1.3, giving a warning on points of incompatibility (handling of non-ASCII line breaks, as described above).

Example 7.14. “YAML” directive

%YAML 1.3 # Attempt parsing
           # with a warning

It is an error to specify more than one “YAML” directive for the same document, even if both occurrences give the same version number.

Example 7.15. Invalid Repeated YAML directive

%YAML 1.3
%YAML 1.2
The YAML directive must only be
given at most once per document.

7.8.2. “TAG” Directives

The “TAG” directive establishes a tag shorthand notation for specifying node tags. Each “TAG” directive associates a handle with a prefix. This allows for compact and readable tag notation.

Example 7.16. “TAG” directive

%TAG !yaml!,2002:
!yaml!str "foo"

It is an error to specify more than one “TAG” directive for the same handle in the same document, even if both occurrences give the same prefix.

Example 7.17. Invalid Repeated TAG directive

%TAG ! !foo
%TAG ! !foo
The TAG directive must only
be given at most once per
handle in the same document. Tag Handles

The tag handle exactly matches the prefix of the affected tag shorthand. There are three tag handle variants:

Primary Handle
The primary tag handle is a single “!” character. This allows using the most compact possible notation for a single “primary” name space. By default, the prefix associated with this handle is “!”. Thus, by default, shorthands using this handle are interpreted as local tags.

It is possible to override the default behavior by providing an explicit “TAG” directive, associating a different prefix for this handle. This provides smooth migration from using local tags to using global tags by the simple addition of a single “TAG” directive.

Example 7.18. Primary Tag Handle

# Private
!foo "bar"
# Global
%TAG !,2000:app/
!foo "bar"
!<!foo> "bar"
!<,2000:app/foo> "bar"
Secondary Handle
The secondary tag handle is written as “!!”. This allows using a compact notation for a single “secondary” name space. By default, the prefix associated with this handle is “,2002:”.

It is possible to override this default behavior by providing an explicit “TAG” directive associating a different prefix for this handle.

Example 7.19. Secondary Tag Handle

%TAG !!,2000:app/
!!int 1 - 3 # Interval, not integer
!<,2000:app/int> "1 - 3"
Named Handles
A named tag handle surrounds a non-empty name with “!” characters. A handle name must not be used in a tag shorthand unless an explicit “TAG” directive has associated some prefix with it.

The name of the handle is a presentation detail and must not be used to convey content information. In particular, the YAML processor need not preserve the handle name once parsing is completed.

Example 7.20. Tag Handles

%TAG !e!,2000:app/
!e!foo "bar"
!<,2000:app/foo> "bar" Tag Prefixes

There are two tag prefix variants:

Local Tag Prefix
If the prefix begins with a “!” character, shorthands using the handle are expanded to a local tag. Note that such a tag is intentionally not a valid URI and its semantics are specific to the application. In particular, two documents in the same stream may assign different semantics to the same local tag.

Example 7.21. Local Tag Prefix

%TAG !m! !my-
--- # Bulb here
!m!light fluorescent
%TAG !m! !my-
--- # Color here
!m!light green
!<!my-light> "fluorescent"
!<!my-light> "green"
Global Tag Prefix
If the prefix begins with a character other than “!”, it must be a valid URI prefix, and should contain at least the scheme. Shorthands using the associated handle are expanded to globally unique URI tags and their semantics is consistent across applications. In particular, every document in every stream must assign the same semantics to the same global tag.

Example 7.22. Global Tag Prefix

%TAG !e!,2000:app/
- !e!foo "bar"
- !<,2000:app/foo> "bar"

7.9. Node Properties

Each node may have two optional properties, anchor and tag, in addition to its content. Node properties may be specified in any order before the node’s content. Either or both may be omitted.

Example 7.23. Node Properties

!!str &a1 "foo":
  !!str bar
&a2 baz : *a1
{ &B1 "foo": "bar",
  "baz": *B1 }

7.9.1. Node Tags

The tag property identifies the type of the native data structure presented by the node. A tag is denoted by the “!” indicator.

Verbatim Tags
A tag may be written verbatim by surrounding it with the “<” and “>” characters. In this case, the YAML processor must deliver the verbatim tag as-is to the application. In particular, verbatim tags are not subject to tag resolution. A verbatim tag must either begin with a “!” (a local tag) or be a valid URI (a global tag).

Example 7.24. Verbatim Tags

!<,2002:str> foo :
  !<!bar> baz
{ "foo": !<!bar> "baz" }


Example 7.25. Invalid Verbatim Tags

- !<!> foo
- !<$:?> bar
- Verbatim tags aren't resolved,
  so ! is invalid.
- The $:? tag is neither a global
  URI tag nor a local tag starting
  with '!'.
Tag Shorthands
A tag shorthand consists of a valid tag handle followed by a non-empty suffix. The tag handle must be associated with a prefix, either by default or by using a “TAG” directive. The resulting parsed tag is the concatenation of the prefix and the suffix and must either begin with “!” (a local tag) or be a valid URI (a global tag).

The choice of tag handle is a presentation detail and must not be used to convey content information. In particular, the tag handle may be discarded once parsing is completed.

The suffix must not contain any “!” character. This would cause the tag shorthand to be interpreted as having a named tag handle. In addition, the suffix must not contain the “[”, “]”, “{”, “}” and “,” characters. These characters would cause ambiguity with flow collection structures. If the suffix needs to specify any of the above restricted characters, they must be escaped using the “%” character. This behavior is consistent with the URI character escaping rules (specifically, section 2.3 of URI RFC).

Example 7.26. Tag Shorthands

%TAG !e!,2000:app/
- !local foo
- !!str bar
- !e!tag%21 baz
[ !<!local> "foo",
  !<,2002:str> "bar",
  !<,2000:app/tag!> "baz" ]


Example 7.27. Invalid Tag Shorthands

%TAG !e! tag:example,2000:app/
- !e! foo
- !h!bar baz
- The !e! handle has no suffix.
- The !h! handle wasn't declared.
Non-Specific Tags
If a node has no tag property, it is assigned a non-specific tag that needs to be resolved to a specific one. This non-specific tag is “!” for non-plain scalars and “?” for all other nodes. This is the only case where the node style has any effect on the content information.

It is possible for the tag property to be explicitly set to the “!” non-specific tag. By convention, this “disables” tag resolution, forcing the node to be interpreted as “,2002:seq”, “,2002:map” or “,2002:str”, according to its kind.

There is no way to explicitly specify the “?” non-specific tag. This is intentional.

Example 7.28. Non-Specific Tags

# Assuming conventional resolution:
- "12"
- 12
- ! 12
[ "12",
  "12" ]

7.9.2. Node Anchors

An anchor is denoted by the “&” indicator. It marks a node for future reference. An alias node can then be used to indicate additional inclusions of the anchored node. An anchored node need not be referenced by any alias nodes; in particular, it is valid for all nodes to be anchored.

Note that as a serialization detail, the anchor name is preserved in the serialization tree. However, it is not reflected in the representation graph and must not be used to convey content information. In particular, the YAML processor need not preserve the anchor name once the representation is composed.

Anchor names must not contain the “[”, “]”, “{”, “}” and “,” characters. These characters would cause ambiguity with flow collection structures.

Example 7.29. Node Anchors

First occurrence: &anchor Value
Second occurrence: *anchor
{ "First occurrence": &A "Value",
  "Second occurrence": *A }

Chapter 8. Flow Style Productions

YAML’s flow styles can be thought of as the natural extension of JSON to cover folding long content lines for readability, tagging nodes to control construction of native data structures and using anchors and aliases to reuse constructed object instances.

8.1. Alias Nodes

Subsequent occurrences of a previously serialized node are presented as alias nodes. The first occurrence of the node must be marked by an anchor to allow subsequent occurrences to be presented as alias nodes.

An alias node is denoted by the “*” indicator. The alias refers to the most recent preceding node having the same anchor. It is an error for an alias node to use an anchor that does not previously occur in the document. It is not an error to specify an anchor that is not used by any alias node.

Note that an alias node must not specify any properties or content, as these were already specified at the first occurrence of the node.

Example 8.1. Alias Nodes

First occurrence: &anchor Foo
Second occurrence: *anchor
Override anchor: &anchor Bar
Reuse anchor: *anchor
{ "First occurrence": &A "Foo",
  "Override anchor": &B "Bar",
  "Second occurrence": *A,
  "Reuse anchor": *B }

8.2. Empty Nodes

YAML allows the node content to be omitted in many cases. Nodes with empty content are interpreted as if they were plain scalars with an empty value. Such nodes are commonly resolved to a “null” value.

In the examples, empty scalars are sometimes displayed as the glyph “°” for clarity. Note that this glyph corresponds to a position in the characters stream rather than to an actual character.

Example 8.2. Empty Content

  foo : !!str°,
  !!str° : bar,
{ "foo": "",
  "": "bar" }


Both the node’s properties and node content are optional. This allows for a completely empty node. Completely empty nodes are only valid when following some explicit indication for their existence.

Example 8.3. Completely Empty Flow Nodes

  ? foo :°,
  °: bar,
{ "foo": null,
  null : "bar" }


8.3. Flow Scalar Styles

YAML provides three flow scalar styles: double-quoted, single-quoted and plain (unquoted). Each provides a different trade-off between readability and expressive power.

The scalar style is a presentation detail and must not be used to convey content information, with the exception that plain scalars are distinguished for the purpose of tag resolution.

8.3.1. Double-Quoted Style

The double-quoted style is specified by surrounding “"” indicators. This is the only style capable of expressing arbitrary strings, by using “\escape sequences. This comes at the cost of having to escape the “\” and “"” characters.

Double-quoted scalars are restricted to a single line when contained inside an implicit key.

Example 8.4. Double Quoted Implicit Keys

"implicit block key" : [
  "implicit flow key" : value,
{ "implicit block key":
  [ { "implicit flow key": "value" } ] }

In a multi-line double-quoted scalar, line breaks are subject to flow line folding, which discards any trailing white space characters. It is also possible to escape the line break character. In this case, the escaped line break is excluded from the content and any trailing white space characters that precede the escaped line break are preserved. Combined with the ability to escape white space characters, this allows double-quoted lines to be broken at arbitrary positions.

Example 8.5. Double Quoted Line Breaks

to a space,→↵
to a line feed, or·→\↵
"folded to a space,\nto a line feed, or \t \tnon-content"

All leading and trailing white space characters on each line are excluded from the content. Each continuation line must therefore contain at least one non-space character. Empty lines, if any, are consumed as part of the line folding.

Example 8.6. Double Quoted Lines

"·1st non-empty

·2nd non-empty·
3rd non-empty·"
" 1st non-empty\n2nd non-empty 3rd non-empty "

8.3.2. Single-Quoted Style

The single-quoted style is specified by surrounding “'” indicators. Therefore, within a single-quoted scalar, such characters need to be repeated. This is the only form of escaping performed in single-quoted scalars. In particular, the “\” and “"” characters may be freely used. This restricts single-quoted scalars to printable characters. In addition, it is only possible to break a long single-quoted line where a space character is surrounded by non-spaces.

Example 8.7. Single Quoted Characters

'here''s to "quotes"'
"here's to \"quotes\""

Single-quoted scalars are restricted to a single line when contained inside a implicit key.

Example 8.8. Single Quoted Implicit Keys

'implicit block key' : [
  'implicit flow key' : value,
{ "implicit block key":
  [ { "implicit flow key": "value" } ] }

All leading and trailing white space characters are excluded from the content. Each continuation line must therefore contain at least one non-space character. Empty lines, if any, are consumed as part of the line folding.

Example 8.9. Single Quoted Lines

'·1st non-empty

·2nd non-empty·
3rd non-empty·'
" 1st non-empty\n2nd non-empty 3rd non-empty "

8.3.3. Plain Style

The plain (unquoted) style has no identifying indicators and provides no form of escaping. It is therefore the most readable, most limited and most context sensitive style. In addition to a restricted character set, a plain scalar must not be empty or contain leading or trailing white space characters. It is only possible to break a long plain line where a space character is surrounded by non-spaces.

Plain scalars must not begin with most indicators, as this would cause ambiguity with other YAML constructs. However, the “:”, “?” and “-indicators may be used as the first character if followed by a non-space “safe” character, as this causes no ambiguity.

Plain scalars must never contain the “: ” and “ #” character combinations. Such combinations would cause ambiguity with mapping key/value pairs and comments. In addition, inside flow collections, or when used as implicit keys, plain scalars must not contain the “[”, “]”, “{”, “}” and “,” characters. These characters would cause ambiguity with flow collection structures.

Example 8.10. Plain Characters

# Outside flow collection:
- ::vector
- ": - ()"
- Up, up, and away!
- -123
# Inside flow collection:
- [ ::vector,
  ": - ()",
  "Up, up and away!",
  -123, ]
[ "::vector",
  ": - ()",
  "Up, up, and away!",
  [ "::vector",
    ": - ()",
    "Up, up, and away!",
    "" ] ]


Plain scalars are further restricted to a single line when contained inside an implicit key.

Example 8.11. Plain Implicit Keys

implicit block key : [
  implicit flow key : value,
{ "implicit block key":
  [ { "implicit flow key": "value" } ] }

All leading and trailing white space characters are excluded from the content. Each continuation line must therefore contain at least one non-space character. Empty lines, if any, are consumed as part of the line folding.

Example 8.12. Plain Lines

1st non-empty

·2nd non-empty·
3rd non-empty
"1st non-empty\n2nd non-empty 3rd non-empty"

8.4. Flow Collection Styles

A flow collection may be nested within a block collection ([FLOW-OUT context]), nested within another flow collection ([FLOW-IN context]) or be a part of an implicit key ([FLOW-KEY context] or [BLOCK-KEY context]). Flow collection entries are terminated by the “,” indicator. The final “,” may be omitted. This does not cause ambiguity because flow collection entries can never be completely empty.

8.4.1. Flow Sequences

Flow sequence content is denoted by surrounding “[” and “]” characters.

Sequence entries are separated by a “,” character.

Example 8.13. Flow Sequence

- [ one, two, ]
- [three ,four]
[ [ "one",
    "two" ],
  [ "three",
    "four" ] ]


Any flow node may be used as a flow sequence entry. In addition, YAML provides a compact notation for the case where a flow sequence entry is a mapping with a single key/value pair.

Example 8.14. Flow Sequence Entries

 quoted", 'single
 text, [ nested ],
single: pair,
[ "double quoted",
  "single quoted",
  "plain text",
  [ "nested" ],
  { "single": "pair" } ]

8.4.2. Flow Mappings

Flow mappings are denoted by surrounding “{” and “}” characters.

Mapping entries are separated by a “,” character.

Example 8.15. Flow Mappings

- { one : two , three: four , }
- {five: six,seven : eight}
[ { "one": "two",
    "three": "four" },
  { "five": "six",
    "seven": "eight" } ]


If the optional “?” mapping key indicator is specified, the rest of the entry may be completely empty.

Example 8.16. Flow Mapping Entries

? explicit: entry,
implicit: entry,
{ "explicit": "entry",
  "implicit": "entry",
  null: null }

Normally, YAML insists the “:” mapping value indicator be separated from the value by white space. A benefit of this restriction is that the “:” character can be used inside plain scalars, as long as it is not followed by white space. This allows for unquoted URLs and timestamps. It is also a potential source for confusion as “a:1” is a plain scalar and not a key/value pair.

Note that the value may be completely empty since its existence is indicated by the “:”.

Example 8.17. Flow Mapping Separate Values

omitted value:°,
°:·omitted key,
{ "unquoted": "separate",
  "": null,
  "omitted value": null,
  null: "omitted key" }

To ensure JSON compatibility, if a key inside a flow mapping is JSON-like, YAML allows the following value to be specified adjacent to the “:”. This causes no ambiguity, as all JSON-like keys are surrounded by indicators. However, as this greatly reduces readability, YAML processors should separate the value from the “:” on output, even in this case.

Example 8.18. Flow Mapping Adjacent Values

{ "adjacent": "value",
  "readable": "value",
  "empty": null }

A more compact notation is usable inside flow sequences, if the mapping contains a single key/value pair. This notation does not require the surrounding “{” and “}” characters. Note that it is not possible to specify any node properties for the mapping in this case.

Example 8.19. Single Pair Flow Mappings

foo: bar
[ { "foo": "bar" } ]

If the “?” indicator is explicitly specified, parsing is unambiguous and the syntax is identical to the general case.

Example 8.20. Single Pair Explicit Entry

? foo
 bar : baz
[ { "foo bar": "baz" } ]

If the “?” indicator is omitted, parsing needs to see past the implicit key to recognize it as such. To limit the amount of lookahead required, the “:” indicator must appear at most 1024 Unicode characters beyond the start of the key. In addition, the key is restricted to a single line.

Note that YAML allows arbitrary nodes to be used as keys. In particular, a key may be a sequence or a mapping. Thus, without the above restrictions, practical one-pass parsing would have been impossible to implement.

Example 8.21. Single Pair Implicit Entries

- [ YAML·: separate ]
- [ °: empty key entry ]
- [ {JSON: like}:adjacent ]
[ [ { "YAML": "separate" } ],
  [ { null: "empty key entry" } ],
  [ { { "JSON": "like" }: "adjacent" } ] ]

Example 8.22. Invalid Implicit Keys

[ foo
 bar: invalid,
 "foo_...>1K characters..._bar": invalid ]
- The foo bar key spans multiple lines
- The key is too long

8.5. Flow Nodes

JSON-like flow styles all have explicit start and end indicators. The only flow style that does not have this property is the plain scalar. Note that none of the “JSON-like” styles is actually acceptable by JSON. Even the double-quoted style is a superset of the JSON string format.

Example 8.23. Flow Content

- [ a, b ]
- { a: b }
- "a"
- 'b'
- c
[ [ "a", "b" ],
  { "a": "b" },
  "c" ]

A complete flow node also has optional node properties, except for alias nodes which refer to the anchored node properties.

Example 8.24. Flow Nodes

- !!str "a"
- 'b'
- &anchor "c"
- *anchor
- !!str°
[ "a",
  "" ]

Chapter 9. Block Style Productions

YAML’s block styles employ indentation rather than indicators to denote structure. This results in a more human readable (though less compact) notation.

9.1. Block Scalar Styles

YAML provides two block scalar styles, literal and folded. Each provides a different trade-off between readability and expressive power.

9.1.1. Block Scalar Headers

Block scalars are controlled by a few indicators given in a header preceding the content itself. This header is followed by a non-content line break with an optional comment. This is the only case where a comment must not be followed by additional comment lines.

Example 9.1. Block Scalar Header

- | # Empty header↵
- >1 # Indentation indicator↵
- |+ # Chomping indicator↵

- >1- # Both indicators↵
[ "literal\n",
  " folded\n",
  " strip" ] Block Indentation Indicator

Every block scalar has a content indentation level. The content of the block scalar excludes a number of leading spaces on each line up to the content indentation level.

If a block scalar has an indentation indicator, then the content indentation level of the block scalar is equal to the indentation level of the block scalar plus the integer value of the indentation indicator character.

If no indentation indicator is given, then the content indentation level is equal to the number of leading spaces on the first non-empty line of the contents. If there is no non-empty line then the content indentation level is equal to the number of spaces on the longest line.

It is an error if any non-empty line does not begin with a number of spaces greater than or equal to the content indentation level.

It is an error for any of the leading empty lines to contain more spaces than the first non-empty line.

A YAML processor should only emit an explicit indentation indicator for cases where detection will fail.

Example 9.2. Block Indentation Indicator

- |°
- >°
··# detected
- |1
- >°
[ "detected\n",
  "\n\n# detected\n",
  " explicit\n",
  "\t\ndetected\n" ]

Example 9.3. Invalid Block Scalar Indentation Indicators

- |
- >
- |2
- A leading all-space line must
  not have too many spaces.
- A following text line must
  not be less indented.
- The text is less indented
  than the indicated level. Block Chomping Indicator

Chomping controls how final line breaks and trailing empty lines are interpreted. YAML provides three chomping methods:

Stripping is specified by the “-” chomping indicator. In this case, the final line break and any trailing empty lines are excluded from the scalar’s content.
Clipping is the default behavior used if no explicit chomping indicator is specified. In this case, the final line break character is preserved in the scalar’s content. However, any trailing empty lines are excluded from the scalar’s content.
Keeping is specified by the “+” chomping indicator. In this case, the final line break and any trailing empty lines are considered to be part of the scalar’s content. These additional lines are not subject to folding.

The chomping method used is a presentation detail and must not be used to convey content information.

The interpretation of the final line break of a block scalar is controlled by the chomping indicator specified in the block scalar header.

Example 9.4. Chomping Final Line Break

strip: |-
clip: |
keep: |+
{ "strip": "text",
  "clip": "text\n",
  "keep": "text\n" }

The interpretation of the trailing empty lines following a block scalar is also controlled by the chomping indicator specified in the block scalar header.

Explicit comment lines may follow the trailing empty lines. To prevent ambiguity, the first such comment line must be less indented than the block scalar content. Additional comment lines, if any, are not so restricted. This is the only case where the indentation of comment lines is constrained.

Example 9.5. Chomping Trailing Lines

# Strip
  # Comments:
strip: |-
  # text↵
·# Clip
··# comments:

clip: |
  # text↵
·# Keep
··# comments:

keep: |+
  # text↵

·# Trail
··# comments.
{ "strip": "# text",
  "clip": "# text\n",
  "keep": "# text\n\n" }

If a block scalar consists only of empty lines, then these lines are considered as trailing lines and hence are affected by chomping.

Example 9.6. Empty Scalar Chomping

strip: >-

clip: >

keep: |+

{ "strip": "",
  "clip": "",
  "keep": "\n" }

9.1.2. Literal Style

The literal style is denoted by the “|” indicator. It is the simplest, most restricted and most readable scalar style.

Example 9.7. Literal Scalar



Inside literal scalars, all (indented) characters are considered to be content, including white space characters. Note that all line break characters are normalized. In addition, empty lines are not folded, though final line breaks and trailing empty lines are chomped.

There is no way to escape characters inside literal scalars. This restricts them to printable characters. In addition, there is no way to break a long literal line.

Example 9.8. Literal Content

··text·# Comment

9.1.3. Folded Style

The folded style is denoted by the “>” indicator. It is similar to the literal style; however, folded scalars are subject to line folding.

Example 9.9. Folded Scalar


"folded text\n"

Folding allows long lines to be broken anywhere a single space character separates two non-space characters.

Example 9.10. Folded Lines



   * bullet

   * list
   * lines


# Comment
"\nfolded line\nnext line\n  \
* bullet\n \n  * list\n  \
* lines\n\nlast line\n"

(The following three examples duplicate this example, each highlighting different productions.)

Lines starting with white space characters (more-indented lines) are not folded.

Example 9.11. More Indented Lines



···* bullet

···* list
···* lines↵


# Comment
"\nfolded line\nnext line\n  \
* bullet\n \n  * list\n  \
* lines\n\nlast line\n"

Line breaks and empty lines separating folded and more-indented lines are also not folded.

Example 9.12. Empty Separation Lines



   * bullet

   * list
   * lines


# Comment
"\nfolded line\nnext line\n  \
* bullet\n \n  * list\n  \
* lines\n\nlast line\n"


The final line break and trailing empty lines if any, are subject to chomping and are never folded.

Example 9.13. Final Empty Lines



   * bullet

   * list
   * lines


# Comment
"\nfolded line\nnext line\n  \
* bullet\n \n  * list\n  \
* lines\n\nlast line\n"

9.2. Block Collection Styles

For readability, block collections styles are not denoted by any indicator. Instead, YAML uses a lookahead method, where a block collection is distinguished from a plain scalar only when a key/value pair or a sequence entry is seen.

9.2.1. Block Sequences

A block sequence is simply a series of nodes, each denoted by a leading “-” indicator. The “-” indicator must be separated from the node by white space. This allows “-” to be used as the first character in a plain scalar if followed by a non-space character (e.g. “-42”).

Example 9.14. Block Sequence

block sequence:
··- one↵
  - two : three↵
{ "block sequence": [
    { "two": "three" } ] }

The entry node may be either completely empty, be a nested block node or use a compact in-line notation. The compact notation may be used when the entry is itself a nested block collection. In this case, both the “-” indicator and the following spaces are considered to be part of the indentation of the nested collection. Note that it is not possible to specify node properties for such a collection.

Example 9.15. Block Sequence Entry Types

-° # Empty
- |
 block node
-·- one # Compact
··- two # sequence
- one: two # Compact mapping
[ null,
  "block node\n",
  [ "one", "two" ],
  { "one": "two" } ]

9.2.2. Block Mappings

A Block mapping is a series of entries, each presenting a key/value pair.

Example 9.16. Block Mappings

block mapping:
·key: value↵
{ "block mapping": {
    "key": "value" } }

If the “?” indicator is specified, the optional value node must be specified on a separate line, denoted by the “:” indicator. Note that YAML allows here the same compact in-line notation described above for block sequence entries.

Example 9.17. Explicit Block Mapping Entries

? explicit key # Empty value↵°
? |
  block key↵
:·- one # Explicit compact
··- two # block value↵
{ "explicit key": null,
  "block key\n": [
    "two" ] }

If the “?” indicator is omitted, parsing needs to see past the implicit key, in the same way as in the single key/value pair flow mapping. Hence, such keys are subject to the same restrictions; they are limited to a single line and must not span more than 1024 Unicode characters.

In this case, the value may be specified on the same line as the implicit key. Note however that in block mappings the value must never be adjacent to the “:”, as this greatly reduces readability and is not required for JSON compatibility (unlike the case in flow mappings).

There is no compact notation for in-line values. Also, while both the implicit key and the value following it may be empty, the “:” indicator is mandatory. This prevents a potential ambiguity with multi-line plain scalars.

Example 9.18. Implicit Block Mapping Entries

plain key: in-line value
°:° # Both empty
"quoted key":
- entry
{ "plain key": "in-line value",
  null: null,
  "quoted key": [ "entry" ] }

A compact in-line notation is also available. This compact notation may be nested inside block sequences and explicit block mapping entries. Note that it is not possible to specify node properties for such a nested mapping.

Example 9.19. Compact Block Mappings

- sun: yellow↵
- ? earth: blue↵
  : moon: white↵
[ { "sun": "yellow" },
  { { "earth": "blue" }:
      { "moon": "white" } } ]

9.2.3. Block Nodes

YAML allows flow nodes to be embedded inside block collections (but not vice-versa). Flow nodes must be indented by at least one more space than the parent block collection. Note that flow nodes may begin on a following line.

It is at this point that parsing needs to distinguish between a plain scalar and an implicit key starting a nested block mapping.

Example 9.20. Block Node Types

··"flow in block"↵>
 Block scalar↵!!map # Block collection
  foo : bar↵
[ "flow in block",
  "Block scalar\n",
  { "foo": "bar" } ]

The block node’s properties may span across several lines. In this case, they must be indented by at least one more space than the block collection, regardless of the indentation of the block collection entries.

Example 9.21. Block Scalar Nodes

literal: |2
{ "literal": "value",
  "folded": !<!foo> "value" }

Since people perceive the “-” indicator as indentation, nested block sequences may be indented by one less space to compensate, except, of course, if nested inside another block sequence ([BLOCK-OUT context] versus [BLOCK-IN context]).

Example 9.22. Block Collection Nodes

sequence: !!seq
- entry
- !!seq
 - nested
mapping: !!map
 foo: bar
{ "sequence": [
    [ "nested" ] ],
  "mapping": { "foo": "bar" } }

Chapter 10. Document Stream Productions

10.1. Documents

A YAML character stream may contain several documents. Each document is completely independent from the rest.

10.1.1. Document Prefix

A document may be preceded by a prefix specifying the character encoding and optional comment lines. Note that all documents in a stream must use the same character encoding. However it is valid to re-specify the encoding using a byte order mark for each document in the stream.

The existence of the optional prefix does not necessarily indicate the existence of an actual document.

Example 10.1. Document Prefix

⇔# Comment
# lines

10.1.2. Document Markers

Using directives creates a potential ambiguity. It is valid to have a “%” character at the start of a line (e.g. as the first character of the second line of a plain scalar). How, then, to distinguish between an actual directive and a content line that happens to start with a “%” character?

The solution is the use of two special marker lines to control the processing of directives, one at the start of a document and one at the end.

At the start of a document, lines beginning with a “%” character are assumed to be directives. The (possibly empty) list of directives is terminated by a directives end marker line. Lines following this marker can safely use “%” as the first character.

At the end of a document, a document end marker line is used to signal the parser to begin scanning for directives again.

The existence of this optional document suffix does not necessarily indicate the existence of an actual following document.

Obviously, the actual content lines are therefore forbidden to begin with either of these markers.

Example 10.2. Document Markers

%YAML 1.3
... # Suffix

10.1.3. Bare Documents

A bare document does not begin with any directives or marker lines. Such documents are very “clean” as they contain nothing other than the content. In this case, the first non-comment line may not start with a “%” first character.

Document nodes are indented as if they have a parent indented at -1 spaces. Since a node must be more indented than its parent node, this allows the document’s node to be indented at zero or more spaces.

Example 10.3. Bare Documents

# No document
%!PS-Adobe-2.0 # Not the first line
"Bare document"


10.1.4. Explicit Documents

An explicit document begins with an explicit directives end marker line but no directives. Since the existence of the document is indicated by this marker, the document itself may be completely empty.

Example 10.4. Explicit Documents

{ matches
% : 20 }
# Empty
{ "matches %": 20 }

10.1.5. Directives Documents

A directives document begins with some directives followed by an explicit directives end marker line.

Example 10.5. Directives Documents

%YAML 1.3
--- |
%YAML 1.3
# Empty

10.2. Streams

A YAML stream consists of zero or more documents. Subsequent documents require some sort of separation marker line. If a document is not terminated by a document end marker line, then the following document must begin with a directives end marker line.

Example 10.6. Stream

# Empty
%YAML 1.3
matches %: 20
{ "matches %": 20 }

A sequence of bytes is a well-formed stream if, taken as a whole, it complies with the above yaml-stream production.

Part V. Appendices

Appendix A. Grammar

The following chapters formally define the syntax of YAML character streams, using parameterized BNF productions.

Each BNF production is both named and numbered for easy reference. Whenever possible, basic structures are specified before the more complex structures using them in a “bottom up” fashion.

The productions are accompanied by examples which are presented in a two-pane side-by-side format. The left-hand side is the YAML example and the right-hand side is an alternate YAML view of the example. The right-hand view uses JSON when possible. Otherwise it uses a YAML form that is as close to JSON as possible.

A.1. Production Syntax

Productions are defined using the syntax production-name ::= term, where a term is either:

An atomic term
  • A quoted string ("abc"), which matches that concatenation of characters. A single character is usually written with single quotes ('a').
  • A hexadecimal number (x0A), which matches the character at that Unicode code point.
  • A range of hexadecimal numbers ([x20-x7E]), which matches any character whose Unicode code point is within that range.
  • The name of a production (yaml-character), which matches that production.
A lookaround
  • [ lookahead = term ], which matches the empty string if term would match.
  • [ lookahead ≠ term ], which matches the empty string if term would not match.
  • [ lookbehind = term ], which matches the empty string if term would match beginning at any prior point on the line and ending at the current position.
A special production
  • <start-of-line>, which matches the empty string at the beginning of a line.
  • <end-of-input>, matches the empty string at the end of the input.
A parenthesized term
Matches its contents.
A concatenation
Is term-one term-two, which matches term-one followed by term-two.
A alternation
Is term-one | term-two, which matches the term-one if possible, or term-two otherwise.
A quantified term:
  • term?, which matches (term | "").
  • term*, which matches (term term* | "").
  • term+, which matches (term term*).

Note: Quantified terms are always greedy.

The order of precedence is parenthesization, then quantification, then concatenation, then alternation.

Some lines in a production definition might have a comment like:

production-a ::=
  production-b      # clarifying comment

These comments are meant to be informative only. For instance a comment that says # not followed by non-ws char just means that you should be aware that actual production rules will behave as described even though it might not be obvious from the content of that particular production alone.

A.1.1. Production Parameters

Some productions have parameters in parentheses after the name, such as line-prefix-spaces(n,c). A parameterized production is shorthand for a (infinite) series of productions, each with a fixed value for each parameter.

For instance, this production:

production-a(n) ::= production-b(n)

Is shorthand for:

production-a(0) ::= production-b(0)
production-a(1) ::= production-b(1)

And this production:

production-a(n) ::=
  ( production-b(n+m) production-c(n+m) )+

Is shorthand for:

production-a(0) ::=
    ( production-b(0) production-c(0) )+
  | ( production-b(1) production-c(1) )+
  | …
production-a(1) ::=
    ( production-b(1) production-c(1) )+
  | ( production-b(2) production-c(2) )+
  | …

The parameters are as follows:

Indentation: n or m
May be any natural number, including zero. n may also be -1.
Context: c
This parameter allows productions to tweak their behavior according to their surrounding. YAML supports two groups of contexts, distinguishing between block styles and flow styles.

May be any of the following values:

  • BLOCK-IN – inside block context
  • BLOCK-OUT – outside block context
  • BLOCK-KEY – inside block key context
  • FLOW-IN – inside flow context
  • FLOW-OUT – outside flow context
  • FLOW-KEY – inside flow key context
(Block) Chomping: t

The line break chomping behavior for flow scalars. May be any of the following values:

  • STRIP – remove all trailing newlines
  • CLIP – remove all trailing newlines except the first
  • KEEP – retain all trailing newlines

A.1.2. Production Naming Conventions

To make it easier to follow production combinations, production names use a prefix-style naming convention. Each production is given a prefix based on the type of characters it begins and ends with.


A production matching no characters.


A production starting and ending with a special character.


A production matching a single line break.


A production starting and ending with a non-break character.


A production starting and ending with a white space character.


A production starting and ending with a non-space character.


A production matching complete line(s).


A production starting with an X- character and ending with a Y- character, where X- and Y- are any of the above prefixes.

X+, X-Y+

A production as above, with the additional property that the matched content indentation level is greater than the specified n parameter.

A.2. Document Stream Productions

The YAML 1.3 syntax productions are presented here in a top down fashion.

A.2.1. Streams

The starting production (yaml-stream) matches an entire YAML character stream.

yaml-stream ::=
    | byte-order-mark
    | comment-line
    | start-indicator-and-document
document-prefix ::=
document-suffix ::=
document-start-indicator ::=
document-end-indicator ::=
  "..."                             # Not followed by non-ws char

A.2.2. Documents

any-document ::=
  | start-indicator-and-document
  | bare-document
directives-and-document ::=
start-indicator-and-document ::=
    | (
bare-document ::=
  /* Excluding forbidden-content */
directive-line ::=
    | tag-directive-line
    | reserved-directive-line
forbidden-content ::=
    | document-end-indicator
    | blank-character
A.2.2.1. Block Nodes
block-node(n,c) ::=
  | flow-node-in-a-block-node(n)

XXX The following two should be moved into above

block-node-in-a-block-node(n,c) ::=
  | block-collection(n,c)
flow-node-in-a-block-node(n) ::=
block-collection(n,c) ::=
    | block-mapping(n)
block-sequence-context(n,BLOCK-OUT) ::= block-sequence(n-1)
block-sequence-context(n,BLOCK-IN)  ::= block-sequence(n)
block-scalar(n,c) ::=
    | block-folded-scalar(n)
A.2.2.2. Block Mappings
block-mapping(n) ::=
block-mapping-entry(n) ::=
  | block-mapping-implicit-entry(n)
block-mapping-explicit-entry(n) ::=
    | empty-node
block-mapping-explicit-key(n) ::=
  '?'                               # Not followed by non-ws char
block-mapping-explicit-value(n) ::=
  ':'                               # Not followed by non-ws char
block-mapping-implicit-entry(n) ::=
    | empty-node
block-mapping-implicit-key ::=
  | implicit-yaml-key(BLOCK-KEY)
block-mapping-implicit-value(n) ::=
  ':'                               # Not followed by non-ws char
    | (
compact-mapping(n) ::=
A.2.2.3. Block Sequences
block-sequence(n) ::=
block-sequence-entry(n) ::=
  [ lookahead ≠ non-space-character ]
block-indented-node(n,c) ::=
        | compact-mapping(n+1+m)
  | block-node(n,c)
  | (
compact-sequence(n) ::=

A.2.3. Block Scalar Styles

See Production Parameters for the definition of the t variable.

A.2.3.1. Literal Style
block-literal-scalar(n) ::=
literal-scalar-content(n,t) ::=
literal-scalar-line-content(n) ::=
literal-scalar-next-line(n) ::=
A.2.3.2. Folded Style
block-folded-scalar(n) ::=
folded-scalar-content(n,t) ::=
folded-scalar-lines-different-indentation(n) ::=
folded-scalar-lines-same-indentation(n) ::=
    | folded-scalar-spaced-lines(n)
folded-scalar-lines(n) ::=
folded-scalar-spaced-lines(n) ::=
folded-scalar-text(n) ::=
line-break-and-empty-lines(n) ::=
folded-scalar-spaced-text(n) ::=
A.2.3.3. Block Scalar Headers
block-scalar-indicators(t) ::=
    | (
block-scalar-indentation-indicator ::=
block-scalar-chomping-indicator(STRIP) ::= '-'
block-scalar-chomping-indicator(KEEP)  ::= '+'
block-scalar-chomping-indicator(CLIP)  ::= ""
block-scalar-chomp-last(STRIP) ::= line-break | <end-of-input>
block-scalar-chomp-last(CLIP)  ::= break-as-line-feed | <end-of-input>
block-scalar-chomp-last(KEEP)  ::= break-as-line-feed | <end-of-input>
block-scalar-chomp-empty(n,STRIP) ::= line-strip-empty(n)
block-scalar-chomp-empty(n,CLIP)  ::= line-strip-empty(n)
block-scalar-chomp-empty(n,KEEP)  ::= line-keep-empty(n)
line-strip-empty(n) ::=
line-keep-empty(n) ::=
line-trail-comments(n) ::=

A.3. Flow Style Productions

A.3.1. Flow Nodes

flow-node(n,c) ::=
  | flow-content(n,c)
  | (
        | empty-node
flow-content(n,c) ::=
  | flow-json-content(n,c)
flow-yaml-content(n,c) ::=
flow-json-content(n,c) ::=
  | flow-mapping(n,c)
  | single-quoted-scalar(n,c)
  | double-quoted-scalar(n,c)
A.3.1.1. Flow Mappings
flow-mapping(n,c) ::=
flow-mapping-entries(n,c) ::=
flow-mapping-entry(n,c) ::=
      '?'                           # Not followed by non-ws char
  | flow-mapping-implicit-entry(n,c)
flow-mapping-explicit-entry(n,c) ::=
  | (
flow-mapping-implicit-entry(n,c) ::=
  | flow-mapping-empty-key-entry(n,c)
  | flow-mapping-json-key-entry(n,c)
flow-mapping-yaml-key-entry(n,c) ::=
    | empty-node
flow-mapping-empty-key-entry(n,c) ::=
flow-mapping-separate-value(n,c) ::=
  [ lookahead ≠ non-space-plain-scalar-character(c) ]
    | empty-node
flow-mapping-json-key-entry(n,c) ::=
    | empty-node
flow-mapping-adjacent-value(n,c) ::=
    | empty-node
flow-pair(n,c) ::=
      '?'                           # Not followed by non-ws char
  | flow-pair-entry(n,c)
flow-pair-entry(n,c) ::=
  | flow-mapping-empty-key-entry(n,c)
  | flow-pair-json-key-entry(n,c)
flow-pair-yaml-key-entry(n,c) ::=
flow-pair-json-key-entry(n,c) ::=
implicit-yaml-key(c) ::=
  /* At most 1024 characters altogether */
implicit-json-key(c) ::=
  /* At most 1024 characters altogether */
flow-yaml-node(n,c) ::=
  | flow-yaml-content(n,c)
  | (
        | empty-node
flow-json-node(n,c) ::=
A.3.1.2. Flow Sequences
flow-sequence(n,c) ::=
flow-sequence-entries(n,c) ::=
flow-sequence-entry(n,c) ::=
  | flow-node(n,c)

A.3.2. Flow Scalar Styles

A.3.2.1. Double-Quoted Style
double-quoted-scalar(n,c) ::=
double-quoted-text(n,BLOCK-KEY) ::= double-quoted-one-line
double-quoted-text(n,FLOW-KEY)  ::= double-quoted-one-line
double-quoted-text(n,FLOW-OUT)  ::= double-quoted-multi-line(n)
double-quoted-text(n,FLOW-IN)   ::= double-quoted-multi-line(n)
double-quoted-multi-line(n) ::=
    | blank-character*
double-quoted-one-line ::=
double-quoted-first-line ::=
double-quoted-next-line(n) ::=
    | flow-folded-whitespace(n)
      | blank-character*
non-space-double-quoted-character ::=
  - blank-character
non-break-double-quoted-character ::=
  | (
      - '\'
      - '"'
double-quoted-line-continuation(n) ::=
A.3.2.2. Flow Collection Context
flow-mapping-context(n,FLOW-OUT)  ::= flow-sequence-entries(n,FLOW-IN)
flow-mapping-context(n,FLOW-IN)   ::= flow-sequence-entries(n,FLOW-IN)
flow-mapping-context(n,BLOCK-KEY) ::= flow-sequence-entries(n,FLOW-KEY)
flow-mapping-context(n,FLOW-KEY)  ::= flow-sequence-entries(n,FLOW-KEY)
flow-sequence-context(n,FLOW-OUT)  ::= flow-sequence-entries(n,FLOW-IN)
flow-sequence-context(n,FLOW-IN)   ::= flow-sequence-entries(n,FLOW-IN)
flow-sequence-context(n,BLOCK-KEY) ::= flow-sequence-entries(n,FLOW-KEY)
flow-sequence-context(n,FLOW-KEY)  ::= flow-sequence-entries(n,FLOW-KEY)
A.3.2.3. Single-Quoted Style
single-quoted-scalar(n,c) ::=
single-quoted-text(BLOCK-KEY) ::= single-quoted-one-line
single-quoted-text(FLOW-KEY)  ::= single-quoted-one-line
single-quoted-text(FLOW-OUT)  ::= single-quoted-multi-line(n)
single-quoted-text(FLOW-IN)   ::= single-quoted-multi-line(n)
single-quoted-multi-line(n) ::=
    | blank-character*
single-quoted-one-line ::=
single-quoted-first-line ::=
single-quoted-next-line(n) ::=
      | blank-character*
non-space-single-quoted-character ::=
  - blank-character
non-break-single-quoted-character ::=
  | (
      - "'"
single-quoted-escaped-single-quote ::=
A.3.2.4. Plain Style
flow-plain-scalar(n,FLOW-OUT)  ::= plain-scalar-multi-line(n,FLOW-OUT)
flow-plain-scalar(n,FLOW-IN)   ::= plain-scalar-multi-line(n,FLOW-IN)
flow-plain-scalar(n,BLOCK-KEY) ::= plain-scalar-single-line(BLOCK-KEY)
flow-plain-scalar(n,FLOW-KEY)  ::= plain-scalar-single-line(FLOW-KEY)
plain-scalar-multi-line(n,c) ::=
plain-scalar-single-line(c) ::=
plain-scalar-next-line(n,c) ::=
plain-scalar-line-characters(c) ::=
plain-scalar-first-character(c) ::=
      - '?'                         # Mapping key
      - ':'                         # Mapping value
      - '-'                         # Sequence entry
      - '{'                         # Mapping start
      - '}'                         # Mapping end
      - '['                         # Sequence start
      - ']'                         # Sequence end
      - ','                         # Entry separator
      - '#'                         # Comment
      - '&'                         # Anchor
      - '*'                         # Alias
      - '!'                         # Tag
      - '|'                         # Literal scalar
      - '>'                         # Folded scalar
      - "'"                         # Single quote
      - '"'                         # Double quote
      - '%'                         # Directive
      - '@'                         # Reserved
      - '`'                         # Reserved
  | (
      ( '?' | ':' | '-' )
      [ lookahead = non-space-plain-scalar-character(c) ]
plain-scalar-characters(c) ::=
      - ':'
      - '#'
  | (
      [ lookbehind = non-space-character ]
  | (
      [ lookahead = non-space-plain-scalar-character(c) ]
non-space-plain-scalar-character(FLOW-OUT)  ::= block-plain-scalar-character
non-space-plain-scalar-character(FLOW-IN)   ::= flow-plain-scalar-character
non-space-plain-scalar-character(BLOCK-KEY) ::= block-plain-scalar-character
non-space-plain-scalar-character(FLOW-KEY)  ::= flow-plain-scalar-character
block-plain-scalar-character ::=
flow-plain-scalar-character ::=
  - flow-collection-indicators

A.3.3. Other Nodes

alias-node ::=
empty-node ::=

A.4. Structural Productions

A.4.1. Indentation Spaces

indentation-spaces(0) ::=

# When n≥0
indentation-spaces(n+1) ::=
indentation-spaces-less-than(1) ::=

# When n≥1
indentation-spaces-less-than(n+1) ::=
  | ""
indentation-spaces-less-or-equal(0) ::=

# When n≥0
indentation-spaces-less-or-equal(n+1) ::=
  | ""

A.4.2. Line Prefixes

line-prefix-spaces(n,BLOCK-OUT) ::= indentation-spaces-exact(n)
line-prefix-spaces(n,BLOCK-IN)  ::= indentation-spaces-exact(n)
line-prefix-spaces(n,FLOW-OUT)  ::= indentation-spaces-plus-maybe-more(n)
line-prefix-spaces(n,FLOW-IN)   ::= indentation-spaces-plus-maybe-more(n)
indentation-spaces-exact(n) ::=
indentation-spaces-plus-maybe-more(n) ::=

A.4.3. Line Folding

flow-folded-whitespace(n) ::=
folded-whitespace(n,c) ::=
  | break-as-space

A.4.4. Comments

comment-lines ::=
  | <start-of-line>
comment-line ::=
blanks-and-comment-line ::=
comment-content ::=

A.4.5. Empty Lines

empty-line(n,c) ::=
    | indentation-spaces-less-than(n)

A.4.6. Separation Lines and Blanks

separation-characters(n,BLOCK-OUT) ::= separation-lines(n)
separation-characters(n,BLOCK-IN)  ::= separation-lines(n)
separation-characters(n,FLOW-OUT)  ::= separation-lines(n)
separation-characters(n,FLOW-IN)   ::= separation-lines(n)
separation-characters(n,BLOCK-KEY) ::= separation-blanks
separation-characters(n,FLOW-KEY)  ::= separation-blanks
separation-lines(n) ::=
  | separation-blanks
separation-blanks ::=
  | <start-of-line>

A.4.7. Directives

A %YAML directive line.

yaml-directive-line ::=
yaml-version-number ::=
reserved-directive-line ::=
directive-name ::=
directive-parameter ::=
tag-directive-line ::=
tag-handle ::=
  | secondary-tag-handle
  | primary-tag-handle
named-tag-handle ::=
secondary-tag-handle ::=
primary-tag-handle ::=
tag-prefix ::=
  | global-tag-prefix
local-tag-prefix ::=
global-tag-prefix ::=

A.4.8. Node Properties

node-properties(n,c) ::=
  | (
anchor-property ::=
anchor-name ::=

XXX Change to anchor-character ::= flow-plain-scalar-character ?

anchor-character ::=
  - flow-collection-indicators
tag-property ::=
  | shorthand-tag
  | non-specific-tag
verbatim-tag ::=
shorthand-tag ::=
non-specific-tag ::=

A.5. Character Productions

A.5.1. Character Encoding

byte-order-mark ::=

A.5.2. Valid Character Sets

All characters allowed in YAML.

yaml-character ::=
                                    # 8 bit
    x09                             # Tab
  | x0A                             # Line feed
  | x0D                             # Carriage return
  | [x20-x7E]                       # Printable ASCII
                                    # 16 bit
  | x85                             # Next line (NEL)
  | [xA0-xD7FF]                     # Basic multilingual plane (BMP)
  | [xE000-xFFFD]                   # Additional unicode areas
  | [x010000-x10FFFF]               # 32 bit

All characters allowed in JSON.

json-character ::=
    x09                             # Tab
  | [x20-x10FFFF]                   # Non-C0-control characters
non-space-character ::=
  - blank-character
non-break-character ::=
  - x0A
  - x0D
  - byte-order-mark
blank-character ::=
    x20                             # Space
  | x09                             # Tab
space-character ::=

A.5.3. Line Ending Productions

line-ending ::=
  | <end-of-input>
break-as-space ::=
break-as-line-feed ::=
line-break ::=
      x0D                           # Carriage return
      x0A                           # Line feed
  | x0D
  | x0A

A.5.4. Indicator Characters

flow-collection-indicators ::=
    ','                             # Flow collection separator
  | '{'                             # Flow mapping start
  | '}'                             # Flow mapping end
  | '['                             # Flow sequence start
  | ']'                             # Flow sequence end

A.5.5. Miscellaneous Character Sets

See Escaped Characters.

double-quoted-scalar-escape-character ::=
    | 'a'
    | 'b'
    | 't' | x09
    | 'n'
    | 'v'
    | 'f'
    | 'r'
    | 'e'
    | x20
    | '"'
    | '/'
    | '\'
    | 'N'
    | '_'
    | 'L'
    | 'P'
    | ( 'x' hexadecimal-digit{2} )
    | ( 'u' hexadecimal-digit{4} )
    | ( 'U' hexadecimal-digit{8} )
tag-character ::=
  - '!'
  - flow-collection-indicators
uri-character ::=
  | word-character
  | '#'
  | ';'
  | '/'
  | '?'
  | ':'
  | '@'
  | '&'
  | '='
  | '+'
  | '$'
  | ','
  | '_'
  | '.'
  | '!'
  | '~'
  | '*'
  | "'"
  | '('
  | ')'
  | '['
  | ']'
word-character ::=
  | ascii-alpha-character
  | '-'
hexadecimal-digit ::=
  | [x41-x46]                       # A-F
  | [x61-x66]                       # a-f
decimal-digit ::=
  [x30-x39]                         # 0-9
decimal-digit-1-9 ::=
  [x31-x39]                         # 1-9
ascii-alpha-character ::=
    [x41-x5A]                       # A-Z
  | [x61-x7A]                       # a-z