Skip to content

number suffix type annotations #513

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 10 commits into
base: main
Choose a base branch
from
Open
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
165 changes: 134 additions & 31 deletions draft-marchan-kdl2.md
Original file line number Diff line number Diff line change
Expand Up @@ -117,10 +117,11 @@ Being a node-oriented language means that the real core component of any KDL
document is the "node". Every node must have a name, which must be a
String ({{string}}).

The name may be preceded by a Type Annotation ({{type-annotation}}) to further
clarify its type, particularly in relation to its parent node. (For example,
clarifying that a particular `date` child node is for the _publication_ date,
rather than the last-modified date, with `(published)date`.)
The name may be preceded by a Prefix Type Annotation
({{prefix-type-annotation}}) to further clarify its type, particularly in
relation to its parent node. (For example, clarifying that a particular `date`
child node is for the _publication_ date, rather than the last-modified date,
with `(published)date`.)

Following the name are zero or more Arguments ({{argument}}) or
Properties ({{property}}), separated by either whitespace ({{whitespace}}) or a
Expand Down Expand Up @@ -247,28 +248,124 @@ Values _MUST_ be either Arguments ({{argument}}) or values of
Properties ({{property}}). Only String ({{string}}) values may be used as
Node ({{node}}) names or Property ({{property}}) keys.

Values (both as arguments and in properties) _MAY_ be prefixed by a single
Type Annotation ({{type-annotation}}).
Values (both as arguments and in properties) _MAY_ include a single Type
Annotation ({{type-annotation}}).

## Type Annotation

A type annotation is a prefix to any Node Name ({{node}}) or Value ({{value}}) that
includes a _suggestion_ of what type the value is _intended_ to be treated as,
or as a _context-specific elaboration_ of the more generic type the node name
indicates.

Type annotations are written as a set of `(` and `)` with a single
String ({{string}}) in it. It may contain Whitespace after the `(` and before
the `)`, and may be separated from its target by Whitespace.
A type annotation is a String ({{string}}) value attached to any Node Name
({{node}}) or Value ({{value}}) that includes a _suggestion_ of what type the
value is _intended_ to be treated as, or as a _context-specific elaboration_ of
the more generic type the node name indicates.

KDL does not specify any restrictions on what implementations might do with
these annotations. They are free to ignore them, or use them to make decisions
about how to interpret a value.
about how to interpret a value. That said, KDL does reserve certain well-known
strings for what would be their intended purpose, for the sake of
interoperability ({{reserved-type-annotations}}).

Additionally, the following type annotations MAY be recognized by KDL parsers
and, if used, SHOULD interpret these types as follows:
There are two kinds of Type Annotation syntaxes in KDL: Prefix Type Annotations
({{prefix-type-annotation}}) and Suffix Type Annotations
({{suffix-type-annotation}}).

### Examples

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Just for readability purposes, I think it's worth mentioning the # escape hatch here with a reference to the "Explicit Suffix Type Annotation" section, at least when it comes to types like u32. For example, maybe:

To remove ambiguity, some suffixes must be prefixed with #: for example, 10.0u8 is invalid, but 10.0#u8 is. The full list of rules for invalid suffixes is clarified in the "Explicit Suffix Type Annotation" section.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I dropped a suggestion for this.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Ah, GitHub collapsed it as "outdated" because they're bad

~~~kdl
node 123u8
node 0#b 20b 50GiB
node prop=(regex).*
(published)date "1970-01-01"
(contributor)person name="Foo McBar"
~~~

### Prefix Type Annotation

Prefix Type Annotations are written as a set of `(` and `)` with a single String
({{string}}) in it. It may contain Whitespace after the `(` and before the `)`,
and may be separated from its target by Whitespace. Unlike the other annotation
types, any String type may be used.

### Suffix Type Annotation

When a Value ({{value}}) is a Number ({{number}}), it's possible to attach the
type annotation as a "suffix", instead of prepending it between `(` and `)`.
This makes it possible to, for example, write `10px`, `10.5%`, `512GiB`, etc.,
which are equivalent to `(px)10`, `(%)5`, and `(GiB)512`, respectively.

Most suffixes can be appended directly to the number (a Bare Suffix Type
Annotation ({{bare-suffix-type-annotation}})), as shown in the previous
paragraph. To avoid parsing ambiguity, there are some restrictions on this; an
Explicit Suffix Type Annotation ({{explicit-suffix-type-annotation}}) avoids all
these restrictions by using an additional `#` to explicitly indicate it. For
example, `0bytes` is invalid, but `0#bytes` is valid and equivalent to
`(bytes)0`. See Bare Suffix Type Annotation ({{bare-suffix-type-annotation}})
for the full list of restrictions.

An implementation that finds BOTH a parenthesized Prefix Type Annotation
({{prefix-type-annotation}}) and a Suffix Type Annotation
({{suffix-type-annotation}}) on the same Number ({{number}}) MUST yield a syntax
error.

Suffixes MUST BE plain Identifier Strings ({{identifier-string}}). No other
String ({{string}}) syntax is acceptable.

#### Bare Suffix Type Annotation

When a Value ({{value}}) is a Number ({{number}}) that meets certain criteria,
it's possible to append an Identifier String ({{identifier-string}}), and ONLY
an Identifier String, as a suffix directly to the Number, as its Type Annotation
({{type-annotation}}). The criteria are as follows:

* The Number MUST be a Decimal (that is, it MUST NOT start with `0b`, `0o`, or
`0x`). Additionally, the tokens `0b`, `0o`, and `0x` MUST be treated as syntax
errors (incomplete non-decimal numbers).
* It MUST NOT have an exponent part (e.g. `5.2e+3`).
* The Identifier String used for the type itself MUST NOT start with either `.` or `,`.
* As part of the exponential restriction, the suffix MUST NOT match
`[eE]([-+]|[0-9])` (e.g. `5.2e+` SHOULD be considered a "bad exponential", and
MUST NOT parse as `(e+)5.2`).

### Reserved Type Annotations for Numbers Without Decimals:
For example, the following are all illegal:

* `10,000` (suffix would start with `,`)
* `10e0n` (suffix on an exponential)
* `0xyz` (starts with reserved hexadecimal prefix)
* `0b` (starts with reserved binary prefix)
* `5e+oops` (looks too close to an exponential)
* `1.2.3-abc` (suffix would start with `.` AND Identifier Strings can't start
with `.<digit>`)

Whereas the following are all legal:

* `0u8` = `(u8)0`
* `5em` = `(em)5`, the `e` is not followed by a digit.
* `1xyz` = `(xyz)1`, no longer starts with `0` as above.
* `20b` = `(b)20`, "20 bytes", no longer starts with just `0` as above.

If the desired suffix would violate any of the above rules, either Prefix Type
Annotations ({{prefix-type-annotation}}) or Explicit Suffix Type Annotations
({{explicit-suffix-type-annotation}}) may be used.

#### Explicit Suffix Type Annotation

Any Number ({{number}}) may have a `#` appended to it, followed by any valid
Identifier String ({{identifier-string}}). This is an Explicit Suffix Type
Annotation ({{suffix-type-annotation}}) syntax without any of the added
restrictions of Bare Suffix Type Annotations ({{bare-suffix-type-annotation}}),
which can be a useful escape hatch. For example: `0#b` is invalid syntax without
the `#` prefix.

Note that, unlike Bare Suffix Type Annotations
({{bare-suffix-type-annotation}}), Explicit Suffixes may be used with ALL Number
({{number}}) formats (hexadecimal, decimal, octal, and binary). For example,
`0x1234#u32` is valid.

### Reserved Type Annotations

#### For Numbers Without Decimal Parts

Additionally, the following type annotations MAY be recognized by KDL parsers
and, if used, SHOULD interpret these types as follows.

Signed integers of various sizes (the number is the bit size):

Expand All @@ -291,7 +388,7 @@ Platform-dependent integer types, both signed and unsigned:
- `isize`
- `usize`

### Reserved Type Annotations for Numbers With Decimals:
#### For Numbers With Decimal Parts

IEEE 754 floating point numbers, both single (32) and double (64) precision:

Expand All @@ -303,7 +400,7 @@ IEEE 754-2008 decimal floating point numbers
- `decimal64`
- `decimal128`

### Reserved Type Annotations for Strings:
#### For Strings

- `date-time`: ISO8601 date/time format.
- `time`: "Time" section of ISO8601.
Expand Down Expand Up @@ -331,15 +428,6 @@ IEEE 754-2008 decimal floating point numbers
- `base64`: A Base64-encoded string, denoting arbitrary binary data.
- `base85`: An [Ascii85](https://en.wikipedia.org/wiki/Ascii85)-encoded string, denoting arbitrary binary data.

### Examples

~~~kdl
node (u8)123
node prop=(regex).*
(published)date "1970-01-01"
(contributor)person name="Foo McBar"
~~~

## String

Strings in KDL represent textual UTF-8 Values ({{value}}). A String is either an
Expand Down Expand Up @@ -951,8 +1039,9 @@ node-children := '{' nodes final-node? '}'
node-terminator := single-line-comment | newline | ';' | eof

prop := string node-space* '=' node-space* value
value := type? node-space* (string | number | keyword)
value := normal-value | suffixed-decimal
type := '(' node-space* string node-space* ')'
normal-value := type? node-space* (string | number | keyword)

// Strings
string := identifier-string | quoted-string | raw-string ¶
Expand Down Expand Up @@ -1013,12 +1102,26 @@ multi-line-raw-string-body :=
// Numbers
number := keyword-number | hex | octal | binary | decimal

decimal := sign? integer ('.' integer)? exponent?
decimal := significand exponent?
suffixed-decimal := significand (
bare-type-suffix
| (exponent? explicit-type-suffix)
)
significand := sign? significand-initial integer? ('.' integer)?
exponent := ('e' | 'E') sign? integer
integer := digit (digit | '_')*
significand-initial = digit
- '0b'
- '0o'
- '0x'
Comment on lines +1113 to +1116
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This use of - isn't quite right - it's not subtracting from the LHS (the digit), it's imposing a wider restriction on the digit and its following character (which in this grammar would come from the start of bare-type-suffix).

I think this should instead be handled as a cut point on the hex/octal/binary productions, so that if you see their prefix you're committed. (We'll also have to specify that the ordering of | alternates is significant when cuts are involved, so normal-value is parsed first and gets to claim that syntax space with a cut before suffixed-decimal can run and claim it.)

digit := [0-9]
sign := '+' | '-'

bare-type-suffix := unambiguous-bare-type-suffix | e-bare-type-suffix
unambiguous-bare-type-suffix := (identifier-char - ('.' | ',' | 'e' | 'E')) identifier-char*
e-bare-type-suffix := 'e' | 'E' | ( ((identifier-char - ('.' | ',')) identifier-char) - (('e' | 'E') (sign | digit)) ) identifier-char*
explicit-type-suffix := '#' identifier-string

hex := sign? '0x' hex-digit (hex-digit | '_')*
octal := sign? '0o' [0-7] [0-7_]*
binary := sign? '0b' ('0' | '1') ('0' | '1' | '_')*
Expand Down
1 change: 1 addition & 0 deletions tests/test_cases/expected_kdl/suffix_type_bare.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node (abc)123
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node (u8)123
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node (abc)123
1 change: 1 addition & 0 deletions tests/test_cases/expected_kdl/suffix_type_explicit.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node (abc)123
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node (abc)12300000
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node (abc)5 (def)83 (zzz)291
1 change: 0 additions & 1 deletion tests/test_cases/input/bare_ident_numeric_fail.kdl

This file was deleted.

1 change: 0 additions & 1 deletion tests/test_cases/input/bare_ident_numeric_sign_fail.kdl

This file was deleted.

1 change: 1 addition & 0 deletions tests/test_cases/input/exponent_no_digit_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 1.0e-
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123abc
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_binary_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 0b0101hi
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_binaryish_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 0b
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_comma_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123,abc
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123e5e5
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123e5abc
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 1234e+foo
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_hex_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 0x123nope
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_hexish_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 0xohno
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_keyword_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node #truefoo #false#bar #nullish #nan#no #inf123 #-inf#123
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_letter_number.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123u8
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_octal_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 0o123nope
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node "foo"bar
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node #"foo"#bar
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_period_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123.abc
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123"astring"
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 456#"rawstring"#
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_redundant_fail.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node (abc)123def
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_bare_underscore.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123_abc
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_explicit.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123#abc
1 change: 1 addition & 0 deletions tests/test_cases/input/suffix_type_explicit_exponent.kdl
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123e5#abc
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123#true 456#false 789#null 012#nan 345#inf 678#-inf
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 0b0101#abc 0o123#def 0x123#zzz
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123#123
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 123#"string"
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node 456##"rawstring"#
Original file line number Diff line number Diff line change
@@ -0,0 +1 @@
node (abc)123#def