YAML: Do I need quotes for strings in YAML?
SyntaxYamlQuotesSyntax Problem Overview
I am trying to write a YAML dictionary for internationalisation of a Rails project. I am a little confused though, as in some files I see strings in double-quotes and in some without. A few points to consider:
- example 1 - all strings use double quotes;
- example 2 - no strings (except the last two) use quotes;
- the YAML cookbook says: Enclosing strings in double quotes allows you to use escaping to represent ASCII and Unicode characters. Does this mean I need to use double quotes only when I want to escape some characters? If yes - why do they use double quotes everywhere in the first example - only for the sake of unity / stylistic reasons?
- the last two lines of example 2 use
!
- the non-specific tag, while the last two lines of the first example don't - and they both work.
My question is: what are the rules for using the different types of quotes in YAML?
Could it be said that:
- in general, you don't need quotes;
- if you want to escape characters use double quotes;
- use
!
with single quotes, when... ?!?
Syntax Solutions
Solution 1 - Syntax
After a brief review of the YAML cookbook cited in the question and some testing, here's my interpretation:
- In general, you don't need quotes.
- Use quotes to force a string, e.g. if your key or value is
10
but you want it to return a String and not a Fixnum, write'10'
or"10"
. - Use quotes if your value includes special characters, (e.g.
:
,{
,}
,[
,]
,,
,&
,*
,#
,?
,|
,-
,<
,>
,=
,!
,%
,@
,\
). - Single quotes let you put almost any character in your string, and won't try to parse escape codes.
'\n'
would be returned as the string\n
. - Double quotes parse escape codes.
"\n"
would be returned as a line feed character. - The exclamation mark introduces a method, e.g.
!ruby/sym
to return a Ruby symbol.
Seems to me that the best approach would be to not use quotes unless you have to, and then to use single quotes unless you specifically want to process escape codes.
Update
"Yes" and "No" should be enclosed in quotes (single or double) or else they will be interpreted as TrueClass and FalseClass values:
en:
yesno:
'yes': 'Yes'
'no': 'No'
Solution 2 - Syntax
While Mark's answer nicely summarizes when the quotes are needed according to the YAML language rules, I think what many of the developers/administrators are asking themselves, when working with strings in YAML, is "what should be my rule of thumb for handling the strings?"
It may sound subjective, but the number of rules you have to remember, if you want to use the quotes only when they are really needed as per the language spec, is somewhat excessive for such a simple thing as specifying one of the most common datatypes. Don't get me wrong, you will eventually remember them when working with YAML regularly, but what if you use it occasionally, and you didn't develop automatism for writing YAML? Do you really want to spend time remembering all the rules just to specify the string correctly?
The whole point of the "rule of thumb" is to save the cognitive resource and to handle a common task without thinking about it. Our "CPU" time can arguably be used for something more useful than handling the strings correctly.
From this - pure practical - perspective, I think the best rule of thumb is to single quote the strings. The rationale behind it:
- Single quoted strings work for all scenarios, except when you need to use escape sequences.
- The only special character you have to handle within a single-quoted string is the single quote itself.
These are just 2 rules to remember for some occasional YAML user, minimizing the cognitive effort.
Solution 3 - Syntax
There have been some great answers to this question. However, I would like to extend them and provide some context from the new official YAML v1.2.2 specification (released October 1st 2021) which is the "true source" to all things considering YAML.
There are three different styles that can be used to represent strings, each of them with their own (dis-)advantages:
> YAML provides three flow scalar styles: double-quoted, single-quoted and plain (unquoted). Each provides a different trade-off between readability and expressive power.
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.
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.
Plain (unquoted) 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.
TL;DR
With that being said, according to the official YAML specification one should:
-
Whenever applicable use the unquoted style since it is the most readable.
-
Use the single-quoted style (
'
) if characters such as"
and\
are being used inside the string to avoid escpaing them and therefore improve readability. -
Use the double-quoted style (
"
) when the first two options aren't sufficient, i.e. in scenarios where more complex line breaks are required or non-printable characters are needed.
Solution 4 - Syntax
Strings in yaml only need quotation if (the beginning of) the value can be misinterpreted as a data type or the value contains a ":" (because it could get misinterpreted as key).
For example
foo: '{{ bar }}'
needs quotes, because it can be misinterpreted as datatype dict
, but
foo: barbaz{{ bam }}
does not, since it does not begin with a critical char. Next,
foo: '123'
needs quotes, because it can be misinterpreted as datatype int
, but
foo: bar1baz234
bar: 123baz
Does not, because it can not be misinterpreted as int
foo: 'yes'
needs quotes, because it can be misinterpreted as datatype bool
foo: "bar:baz:bam"
needs quotes, because the value can be misinterpreted as key.
These are just examples. Using yamllint
helps avoiding to start values with a wrong token
foo@bar:/tmp$ yamllint test.yaml
test.yaml
3:4 error syntax error: found character '@' that cannot start any token (syntax)
and is a must, if working productively with yaml.
Quoting all strings as some suggest, is like using brackets in python. It is bad practice, harms readability and throws away the beautiful feature of not having to quote strings.
Solution 5 - Syntax
I had this concern when working on a Rails application with Docker.
My most preferred approach is to generally not use quotes. This includes not using quotes for:
- variables like
${RAILS_ENV}
- values separated by a colon (:) like
postgres-log:/var/log/postgresql
- other strings values
I, however, use double-quotes for integer
values that need to be converted to strings like:
- docker-compose version like
version: "3.8"
- port numbers like
"8080:8080"
- image
"traefik:v2.2.1"
However, for special cases like booleans
, floats
, integers
, and other cases, where using double-quotes for the entry values could be interpreted as strings
, please do not use double-quotes.
Here's a sample docker-compose.yml
file to explain this concept:
version: "3"
services:
traefik:
image: "traefik:v2.2.1"
command:
- --api.insecure=true # Don't do that in production
- --providers.docker=true
- --providers.docker.exposedbydefault=false
- --entrypoints.web.address=:80
ports:
- "80:80"
- "8080:8080"
volumes:
- /var/run/docker.sock:/var/run/docker.sock:ro
That's all.
I hope this helps
Solution 6 - Syntax
If you are trying to escape a string in pytest tavern, !raw
could be helpful to avoid parsing of strings to yaml:
some: !raw "{test: 123}"
Check for more info: https://tavern.readthedocs.io/en/latest/basics.html#type-conversions
Solution 7 - Syntax
Here's a small function (not optimized for performance) that quotes your strings with single quotes if needed and tests if the result could be unmarshalled into the original value: https://go.dev/play/p/AKBzDpVz9hk. Instead of testing for the rules it simply uses the marshaller itself and checks if the marshalled and unmmarshalled value matches the original version.
func yamlQuote(value string) string {
input := fmt.Sprintf("key: %s", value)
var res struct {
Value string `yaml:"key"`
}
if err := yaml.Unmarshal([]byte(input), &res); err != nil || value != res.Value {
quoted := strings.ReplaceAll(value, `'`, `''`)
return fmt.Sprintf("'%s'", quoted)
}
return value
}