diff options
author | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-17 12:19:03 +0000 |
---|---|---|
committer | Daniel Baumann <daniel.baumann@progress-linux.org> | 2024-04-17 12:19:03 +0000 |
commit | 64d98f8ee037282c35007b64c2649055c56af1db (patch) | |
tree | 5492bcf97fce41ee1c0b1cc2add283f3e66cdab0 /vendor/time/src/format_description/parse | |
parent | Adding debian version 1.67.1+dfsg1-1. (diff) | |
download | rustc-64d98f8ee037282c35007b64c2649055c56af1db.tar.xz rustc-64d98f8ee037282c35007b64c2649055c56af1db.zip |
Merging upstream version 1.68.2+dfsg1.
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'vendor/time/src/format_description/parse')
-rw-r--r-- | vendor/time/src/format_description/parse/ast.rs | 278 | ||||
-rw-r--r-- | vendor/time/src/format_description/parse/format_item.rs | 386 | ||||
-rw-r--r-- | vendor/time/src/format_description/parse/lexer.rs | 159 | ||||
-rw-r--r-- | vendor/time/src/format_description/parse/mod.rs | 193 |
4 files changed, 1016 insertions, 0 deletions
diff --git a/vendor/time/src/format_description/parse/ast.rs b/vendor/time/src/format_description/parse/ast.rs new file mode 100644 index 000000000..6977cc9cd --- /dev/null +++ b/vendor/time/src/format_description/parse/ast.rs @@ -0,0 +1,278 @@ +//! AST for parsing format descriptions. + +use alloc::string::String; +use alloc::vec::Vec; +use core::iter; +use core::iter::Peekable; + +use super::{lexer, Error, Location, Span}; + +/// One part of a complete format description. +#[allow(variant_size_differences)] +pub(super) enum Item<'a> { + /// A literal string, formatted and parsed as-is. + Literal { + /// The string itself. + value: &'a [u8], + /// Where the string originates from in the format string. + _span: Span, + }, + /// A sequence of brackets. The first acts as the escape character. + EscapedBracket { + /// The first bracket. + _first: Location, + /// The second bracket. + _second: Location, + }, + /// Part of a type, along with its modifiers. + Component { + /// Where the opening bracket was in the format string. + _opening_bracket: Location, + /// Whitespace between the opening bracket and name. + _leading_whitespace: Option<Whitespace<'a>>, + /// The name of the component. + name: Name<'a>, + /// The modifiers for the component. + modifiers: Vec<Modifier<'a>>, + /// Whitespace between the modifiers and closing bracket. + _trailing_whitespace: Option<Whitespace<'a>>, + /// Where the closing bracket was in the format string. + _closing_bracket: Location, + }, +} + +/// Whitespace within a component. +pub(super) struct Whitespace<'a> { + /// The whitespace itself. + pub(super) _value: &'a [u8], + /// Where the whitespace was in the format string. + pub(super) span: Span, +} + +/// The name of a component. +pub(super) struct Name<'a> { + /// The name itself. + pub(super) value: &'a [u8], + /// Where the name was in the format string. + pub(super) span: Span, +} + +/// A modifier for a component. +pub(super) struct Modifier<'a> { + /// Whitespace preceding the modifier. + pub(super) _leading_whitespace: Whitespace<'a>, + /// The key of the modifier. + pub(super) key: Key<'a>, + /// Where the colon of the modifier was in the format string. + pub(super) _colon: Location, + /// The value of the modifier. + pub(super) value: Value<'a>, +} + +/// The key of a modifier. +pub(super) struct Key<'a> { + /// The key itself. + pub(super) value: &'a [u8], + /// Where the key was in the format string. + pub(super) span: Span, +} + +/// The value of a modifier. +pub(super) struct Value<'a> { + /// The value itself. + pub(super) value: &'a [u8], + /// Where the value was in the format string. + pub(super) span: Span, +} + +/// Parse the provided tokens into an AST. +pub(super) fn parse<'a>( + tokens: impl Iterator<Item = lexer::Token<'a>>, +) -> impl Iterator<Item = Result<Item<'a>, Error>> { + let mut tokens = tokens.peekable(); + iter::from_fn(move || { + Some(match tokens.next()? { + lexer::Token::Literal { value, span } => Ok(Item::Literal { value, _span: span }), + lexer::Token::Bracket { + kind: lexer::BracketKind::Opening, + location, + } => { + // escaped bracket + if let Some(&lexer::Token::Bracket { + kind: lexer::BracketKind::Opening, + location: second_location, + }) = tokens.peek() + { + tokens.next(); // consume + Ok(Item::EscapedBracket { + _first: location, + _second: second_location, + }) + } + // component + else { + parse_component(location, &mut tokens) + } + } + lexer::Token::Bracket { + kind: lexer::BracketKind::Closing, + location: _, + } => unreachable!( + "internal error: closing bracket should have been consumed by `parse_component`", + ), + lexer::Token::ComponentPart { + kind: _, + value: _, + span: _, + } => unreachable!( + "internal error: component part should have been consumed by `parse_component`", + ), + }) + }) +} + +/// Parse a component. This assumes that the opening bracket has already been consumed. +fn parse_component<'a>( + opening_bracket: Location, + tokens: &mut Peekable<impl Iterator<Item = lexer::Token<'a>>>, +) -> Result<Item<'a>, Error> { + let leading_whitespace = if let Some(&lexer::Token::ComponentPart { + kind: lexer::ComponentKind::Whitespace, + value, + span, + }) = tokens.peek() + { + tokens.next(); // consume + Some(Whitespace { + _value: value, + span, + }) + } else { + None + }; + + let name = if let Some(&lexer::Token::ComponentPart { + kind: lexer::ComponentKind::NotWhitespace, + value, + span, + }) = tokens.peek() + { + tokens.next(); // consume + Name { value, span } + } else { + let span = leading_whitespace.map_or_else( + || Span { + start: opening_bracket, + end: opening_bracket, + }, + |whitespace| whitespace.span.shrink_to_end(), + ); + return Err(Error { + _inner: span.error("expected component name"), + public: crate::error::InvalidFormatDescription::MissingComponentName { + index: span.start_byte(), + }, + }); + }; + + let mut modifiers = Vec::new(); + let trailing_whitespace = loop { + let whitespace = if let Some(&lexer::Token::ComponentPart { + kind: lexer::ComponentKind::Whitespace, + value, + span, + }) = tokens.peek() + { + tokens.next(); // consume + Whitespace { + _value: value, + span, + } + } else { + break None; + }; + + if let Some(&lexer::Token::ComponentPart { + kind: lexer::ComponentKind::NotWhitespace, + value, + span, + }) = tokens.peek() + { + tokens.next(); // consume + + let colon_index = match value.iter().position(|&b| b == b':') { + Some(index) => index, + None => { + return Err(Error { + _inner: span.error("modifier must be of the form `key:value`"), + public: crate::error::InvalidFormatDescription::InvalidModifier { + value: String::from_utf8_lossy(value).into_owned(), + index: span.start_byte(), + }, + }); + } + }; + let key = &value[..colon_index]; + let value = &value[colon_index + 1..]; + + if key.is_empty() { + return Err(Error { + _inner: span.shrink_to_start().error("expected modifier key"), + public: crate::error::InvalidFormatDescription::InvalidModifier { + value: String::new(), + index: span.start_byte(), + }, + }); + } + if value.is_empty() { + return Err(Error { + _inner: span.shrink_to_end().error("expected modifier value"), + public: crate::error::InvalidFormatDescription::InvalidModifier { + value: String::new(), + index: span.shrink_to_end().start_byte(), + }, + }); + } + + modifiers.push(Modifier { + _leading_whitespace: whitespace, + key: Key { + value: key, + span: span.subspan(..colon_index), + }, + _colon: span.start.offset(colon_index), + value: Value { + value, + span: span.subspan(colon_index + 1..), + }, + }); + } else { + break Some(whitespace); + } + }; + + let closing_bracket = if let Some(&lexer::Token::Bracket { + kind: lexer::BracketKind::Closing, + location, + }) = tokens.peek() + { + tokens.next(); // consume + location + } else { + return Err(Error { + _inner: opening_bracket.error("unclosed bracket"), + public: crate::error::InvalidFormatDescription::UnclosedOpeningBracket { + index: opening_bracket.byte, + }, + }); + }; + + Ok(Item::Component { + _opening_bracket: opening_bracket, + _leading_whitespace: leading_whitespace, + name, + modifiers, + _trailing_whitespace: trailing_whitespace, + _closing_bracket: closing_bracket, + }) +} diff --git a/vendor/time/src/format_description/parse/format_item.rs b/vendor/time/src/format_description/parse/format_item.rs new file mode 100644 index 000000000..53146d522 --- /dev/null +++ b/vendor/time/src/format_description/parse/format_item.rs @@ -0,0 +1,386 @@ +//! Typed, validated representation of a parsed format description. + +use alloc::string::String; + +use super::{ast, Error}; + +/// Parse an AST iterator into a sequence of format items. +pub(super) fn parse<'a>( + ast_items: impl Iterator<Item = Result<ast::Item<'a>, Error>>, +) -> impl Iterator<Item = Result<Item<'a>, Error>> { + ast_items.map(|ast_item| ast_item.and_then(Item::from_ast)) +} + +/// A description of how to format and parse one part of a type. +#[allow(variant_size_differences)] +pub(super) enum Item<'a> { + /// A literal string. + Literal(&'a [u8]), + /// Part of a type, along with its modifiers. + Component(Component), +} + +impl Item<'_> { + /// Parse an AST item into a format item. + pub(super) fn from_ast(ast_item: ast::Item<'_>) -> Result<Item<'_>, Error> { + Ok(match ast_item { + ast::Item::Component { + _opening_bracket: _, + _leading_whitespace: _, + name, + modifiers, + _trailing_whitespace: _, + _closing_bracket: _, + } => Item::Component(component_from_ast(&name, &modifiers)?), + ast::Item::Literal { value, _span: _ } => Item::Literal(value), + ast::Item::EscapedBracket { + _first: _, + _second: _, + } => Item::Literal(b"["), + }) + } +} + +impl<'a> From<Item<'a>> for crate::format_description::FormatItem<'a> { + fn from(item: Item<'a>) -> Self { + match item { + Item::Literal(literal) => Self::Literal(literal), + Item::Component(component) => Self::Component(component.into()), + } + } +} + +impl From<Item<'_>> for crate::format_description::OwnedFormatItem { + fn from(item: Item<'_>) -> Self { + match item { + Item::Literal(literal) => Self::Literal(literal.to_vec().into_boxed_slice()), + Item::Component(component) => Self::Component(component.into()), + } + } +} + +/// Declare the `Component` struct. +macro_rules! component_definition { + ($vis:vis enum $name:ident { + $($variant:ident = $parse_variant:literal { + $($field:ident = $parse_field:literal: + Option<$field_type:ty> => $target_field:ident),* $(,)? + }),* $(,)? + }) => { + $vis enum $name { + $($variant($variant),)* + } + + $($vis struct $variant { + $($field: Option<$field_type>),* + })* + + $(impl $variant { + /// Parse the component from the AST, given its modifiers. + fn with_modifiers(modifiers: &[ast::Modifier<'_>]) -> Result<Self, Error> { + let mut this = Self { + $($field: None),* + }; + + for modifier in modifiers { + $(if modifier.key.value.eq_ignore_ascii_case($parse_field) { + this.$field = <$field_type>::from_modifier_value(&modifier.value)?; + continue; + })* + return Err(Error { + _inner: modifier.key.span.error("invalid modifier key"), + public: crate::error::InvalidFormatDescription::InvalidModifier { + value: String::from_utf8_lossy(modifier.key.value).into_owned(), + index: modifier.key.span.start_byte(), + } + }); + } + + Ok(this) + } + })* + + impl From<$name> for crate::format_description::Component { + fn from(component: $name) -> Self { + match component {$( + $name::$variant($variant { $($field),* }) => { + $crate::format_description::component::Component::$variant( + $crate::format_description::modifier::$variant {$( + $target_field: $field.unwrap_or_default().into() + ),*} + ) + } + )*} + } + } + + /// Parse a component from the AST, given its name and modifiers. + fn component_from_ast( + name: &ast::Name<'_>, + modifiers: &[ast::Modifier<'_>], + ) -> Result<Component, Error> { + $(if name.value.eq_ignore_ascii_case($parse_variant) { + return Ok(Component::$variant($variant::with_modifiers(&modifiers)?)); + })* + Err(Error { + _inner: name.span.error("invalid component"), + public: crate::error::InvalidFormatDescription::InvalidComponentName { + name: String::from_utf8_lossy(name.value).into_owned(), + index: name.span.start_byte(), + }, + }) + } + } +} + +// Keep in alphabetical order. +component_definition! { + pub(super) enum Component { + Day = b"day" { + padding = b"padding": Option<Padding> => padding, + }, + Hour = b"hour" { + padding = b"padding": Option<Padding> => padding, + base = b"repr": Option<HourBase> => is_12_hour_clock, + }, + Minute = b"minute" { + padding = b"padding": Option<Padding> => padding, + }, + Month = b"month" { + padding = b"padding": Option<Padding> => padding, + repr = b"repr": Option<MonthRepr> => repr, + case_sensitive = b"case_sensitive": Option<MonthCaseSensitive> => case_sensitive, + }, + OffsetHour = b"offset_hour" { + sign_behavior = b"sign": Option<SignBehavior> => sign_is_mandatory, + padding = b"padding": Option<Padding> => padding, + }, + OffsetMinute = b"offset_minute" { + padding = b"padding": Option<Padding> => padding, + }, + OffsetSecond = b"offset_second" { + padding = b"padding": Option<Padding> => padding, + }, + Ordinal = b"ordinal" { + padding = b"padding": Option<Padding> => padding, + }, + Period = b"period" { + case = b"case": Option<PeriodCase> => is_uppercase, + case_sensitive = b"case_sensitive": Option<PeriodCaseSensitive> => case_sensitive, + }, + Second = b"second" { + padding = b"padding": Option<Padding> => padding, + }, + Subsecond = b"subsecond" { + digits = b"digits": Option<SubsecondDigits> => digits, + }, + Weekday = b"weekday" { + repr = b"repr": Option<WeekdayRepr> => repr, + one_indexed = b"one_indexed": Option<WeekdayOneIndexed> => one_indexed, + case_sensitive = b"case_sensitive": Option<WeekdayCaseSensitive> => case_sensitive, + }, + WeekNumber = b"week_number" { + padding = b"padding": Option<Padding> => padding, + repr = b"repr": Option<WeekNumberRepr> => repr, + }, + Year = b"year" { + padding = b"padding": Option<Padding> => padding, + repr = b"repr": Option<YearRepr> => repr, + base = b"base": Option<YearBase> => iso_week_based, + sign_behavior = b"sign": Option<SignBehavior> => sign_is_mandatory, + }, + } +} + +/// Get the target type for a given enum. +macro_rules! target_ty { + ($name:ident $type:ty) => { + $type + }; + ($name:ident) => { + $crate::format_description::modifier::$name + }; +} + +/// Get the target value for a given enum. +macro_rules! target_value { + ($name:ident $variant:ident $value:expr) => { + $value + }; + ($name:ident $variant:ident) => { + $crate::format_description::modifier::$name::$variant + }; +} + +// TODO use `#[derive(Default)]` on enums once MSRV is 1.62 (NET 2022-12-30) +/// Simulate `#[derive(Default)]` on enums. +macro_rules! derived_default_on_enum { + ($type:ty; $default:expr) => {}; + ($attr:meta $type:ty; $default:expr) => { + impl Default for $type { + fn default() -> Self { + $default + } + } + }; +} + +/// Declare the various modifiers. +/// +/// For the general case, ordinary syntax can be used. Note that you _must_ declare a default +/// variant. The only significant change is that the string representation of the variant must be +/// provided after the variant name. For example, `Numerical = b"numerical"` declares a variant +/// named `Numerical` with the string representation `b"numerical"`. This is the value that will be +/// used when parsing the modifier. The value is not case sensitive. +/// +/// If the type in the public API does not have the same name as the type in the internal +/// representation, then the former must be specified in parenthesis after the internal name. For +/// example, `HourBase(bool)` has an internal name "HourBase", but is represented as a boolean in +/// the public API. +/// +/// By default, the internal variant name is assumed to be the same as the public variant name. If +/// this is not the case, the qualified path to the variant must be specified in parenthesis after +/// the internal variant name. For example, `Twelve(true)` has an internal variant name "Twelve", +/// but is represented as `true` in the public API. +macro_rules! modifier { + ($( + enum $name:ident $(($target_ty:ty))? { + $( + $(#[$attr:meta])? + $variant:ident $(($target_value:expr))? = $parse_variant:literal + ),* $(,)? + } + )+) => {$( + enum $name { + $($variant),* + } + + $(derived_default_on_enum! { + $($attr)? $name; $name::$variant + })* + + impl $name { + /// Parse the modifier from its string representation. + fn from_modifier_value(value: &ast::Value<'_>) -> Result<Option<Self>, Error> { + $(if value.value.eq_ignore_ascii_case($parse_variant) { + return Ok(Some(Self::$variant)); + })* + Err(Error { + _inner: value.span.error("invalid modifier value"), + public: crate::error::InvalidFormatDescription::InvalidModifier { + value: String::from_utf8_lossy(value.value).into_owned(), + index: value.span.start_byte(), + }, + }) + } + } + + impl From<$name> for target_ty!($name $($target_ty)?) { + fn from(modifier: $name) -> Self { + match modifier { + $($name::$variant => target_value!($name $variant $($target_value)?)),* + } + } + } + )+}; +} + +// Keep in alphabetical order. +modifier! { + enum HourBase(bool) { + Twelve(true) = b"12", + #[default] + TwentyFour(false) = b"24", + } + + enum MonthCaseSensitive(bool) { + False(false) = b"false", + #[default] + True(true) = b"true", + } + + enum MonthRepr { + #[default] + Numerical = b"numerical", + Long = b"long", + Short = b"short", + } + + enum Padding { + Space = b"space", + #[default] + Zero = b"zero", + None = b"none", + } + + enum PeriodCase(bool) { + Lower(false) = b"lower", + #[default] + Upper(true) = b"upper", + } + + enum PeriodCaseSensitive(bool) { + False(false) = b"false", + #[default] + True(true) = b"true", + } + + enum SignBehavior(bool) { + #[default] + Automatic(false) = b"automatic", + Mandatory(true) = b"mandatory", + } + + enum SubsecondDigits { + One = b"1", + Two = b"2", + Three = b"3", + Four = b"4", + Five = b"5", + Six = b"6", + Seven = b"7", + Eight = b"8", + Nine = b"9", + #[default] + OneOrMore = b"1+", + } + + enum WeekNumberRepr { + #[default] + Iso = b"iso", + Sunday = b"sunday", + Monday = b"monday", + } + + enum WeekdayCaseSensitive(bool) { + False(false) = b"false", + #[default] + True(true) = b"true", + } + + enum WeekdayOneIndexed(bool) { + False(false) = b"false", + #[default] + True(true) = b"true", + } + + enum WeekdayRepr { + Short = b"short", + #[default] + Long = b"long", + Sunday = b"sunday", + Monday = b"monday", + } + + enum YearBase(bool) { + #[default] + Calendar(false) = b"calendar", + IsoWeek(true) = b"iso_week", + } + + enum YearRepr { + #[default] + Full = b"full", + LastTwo = b"last_two", + } +} diff --git a/vendor/time/src/format_description/parse/lexer.rs b/vendor/time/src/format_description/parse/lexer.rs new file mode 100644 index 000000000..e405ea8c8 --- /dev/null +++ b/vendor/time/src/format_description/parse/lexer.rs @@ -0,0 +1,159 @@ +//! Lexer for parsing format descriptions. + +use core::iter; + +use super::{Location, Span}; + +/// A token emitted by the lexer. There is no semantic meaning at this stage. +pub(super) enum Token<'a> { + /// A literal string, formatted and parsed as-is. + Literal { + /// The string itself. + value: &'a [u8], + /// Where the string was in the format string. + span: Span, + }, + /// An opening or closing bracket. May or may not be the start or end of a component. + Bracket { + /// Whether the bracket is opening or closing. + kind: BracketKind, + /// Where the bracket was in the format string. + location: Location, + }, + /// One part of a component. This could be its name, a modifier, or whitespace. + ComponentPart { + /// Whether the part is whitespace or not. + kind: ComponentKind, + /// The part itself. + value: &'a [u8], + /// Where the part was in the format string. + span: Span, + }, +} + +/// What type of bracket is present. +pub(super) enum BracketKind { + /// An opening bracket: `[` + Opening, + /// A closing bracket: `]` + Closing, +} + +/// Indicates whether the component is whitespace or not. +pub(super) enum ComponentKind { + #[allow(clippy::missing_docs_in_private_items)] + Whitespace, + #[allow(clippy::missing_docs_in_private_items)] + NotWhitespace, +} + +/// Attach [`Location`] information to each byte in the iterator. +fn attach_location(iter: impl Iterator<Item = u8>) -> impl Iterator<Item = (u8, Location)> { + let mut line = 1; + let mut column = 1; + let mut byte_pos = 0; + + iter.map(move |byte| { + let location = Location { + line, + column, + byte: byte_pos, + }; + column += 1; + byte_pos += 1; + + if byte == b'\n' { + line += 1; + column = 1; + } + + (byte, location) + }) +} + +/// Parse the string into a series of [`Token`]s. +pub(super) fn lex(mut input: &[u8]) -> impl Iterator<Item = Token<'_>> { + let mut depth: u8 = 0; + let mut iter = attach_location(input.iter().copied()).peekable(); + let mut second_bracket_location = None; + + iter::from_fn(move || { + // There is a flag set to emit the second half of an escaped bracket pair. + if let Some(location) = second_bracket_location.take() { + return Some(Token::Bracket { + kind: BracketKind::Opening, + location, + }); + } + + Some(match iter.next()? { + (b'[', location) => { + if let Some((_, second_location)) = iter.next_if(|&(byte, _)| byte == b'[') { + // escaped bracket + second_bracket_location = Some(second_location); + input = &input[2..]; + } else { + // opening bracket + depth += 1; + input = &input[1..]; + } + + Token::Bracket { + kind: BracketKind::Opening, + location, + } + } + // closing bracket + (b']', location) if depth > 0 => { + depth -= 1; + input = &input[1..]; + Token::Bracket { + kind: BracketKind::Closing, + location, + } + } + // literal + (_, start_location) if depth == 0 => { + let mut bytes = 1; + let mut end_location = start_location; + + while let Some((_, location)) = iter.next_if(|&(byte, _)| byte != b'[') { + end_location = location; + bytes += 1; + } + + let value = &input[..bytes]; + input = &input[bytes..]; + Token::Literal { + value, + span: Span::start_end(start_location, end_location), + } + } + // component part + (byte, start_location) => { + let mut bytes = 1; + let mut end_location = start_location; + let is_whitespace = byte.is_ascii_whitespace(); + + while let Some((_, location)) = iter.next_if(|&(byte, _)| { + byte != b'[' && byte != b']' && is_whitespace == byte.is_ascii_whitespace() + }) { + end_location = location; + bytes += 1; + } + + let value = &input[..bytes]; + input = &input[bytes..]; + Token::ComponentPart { + kind: if is_whitespace { + ComponentKind::Whitespace + } else { + ComponentKind::NotWhitespace + }, + value, + span: Span::start_end(start_location, end_location), + } + } + }) + }) +} diff --git a/vendor/time/src/format_description/parse/mod.rs b/vendor/time/src/format_description/parse/mod.rs new file mode 100644 index 000000000..c73a67449 --- /dev/null +++ b/vendor/time/src/format_description/parse/mod.rs @@ -0,0 +1,193 @@ +//! Parser for format descriptions. + +use alloc::vec::Vec; +use core::ops::{RangeFrom, RangeTo}; + +mod ast; +mod format_item; +mod lexer; + +/// Parse a sequence of items from the format description. +/// +/// The syntax for the format description can be found in [the +/// book](https://time-rs.github.io/book/api/format-description.html). +pub fn parse( + s: &str, +) -> Result<Vec<crate::format_description::FormatItem<'_>>, crate::error::InvalidFormatDescription> +{ + let lexed = lexer::lex(s.as_bytes()); + let ast = ast::parse(lexed); + let format_items = format_item::parse(ast); + Ok(format_items + .map(|res| res.map(Into::into)) + .collect::<Result<Vec<_>, _>>()?) +} + +/// Parse a sequence of items from the format description. +/// +/// The syntax for the format description can be found in [the +/// book](https://time-rs.github.io/book/api/format-description.html). +/// +/// Unlike [`parse`], this function returns [`OwnedFormatItem`], which owns its contents. This means +/// that there is no lifetime that needs to be handled. +/// +/// [`OwnedFormatItem`]: crate::format_description::OwnedFormatItem +pub fn parse_owned( + s: &str, +) -> Result<crate::format_description::OwnedFormatItem, crate::error::InvalidFormatDescription> { + let lexed = lexer::lex(s.as_bytes()); + let ast = ast::parse(lexed); + let format_items = format_item::parse(ast); + let items = format_items + .map(|res| res.map(Into::into)) + .collect::<Result<Vec<_>, _>>()? + .into_boxed_slice(); + Ok(crate::format_description::OwnedFormatItem::Compound(items)) +} + +/// A location within a string. +#[derive(Clone, Copy)] +struct Location { + /// The one-indexed line of the string. + line: usize, + /// The one-indexed column of the string. + column: usize, + /// The zero-indexed byte of the string. + byte: usize, +} + +impl Location { + /// Offset the location by the provided amount. + /// + /// Note that this assumes the resulting location is on the same line as the original location. + #[must_use = "this does not modify the original value"] + const fn offset(&self, offset: usize) -> Self { + Self { + line: self.line, + column: self.column + offset, + byte: self.byte + offset, + } + } + + /// Create an error with the provided message at this location. + const fn error(self, message: &'static str) -> ErrorInner { + ErrorInner { + _message: message, + _span: Span { + start: self, + end: self, + }, + } + } +} + +/// A start and end point within a string. +#[derive(Clone, Copy)] +struct Span { + #[allow(clippy::missing_docs_in_private_items)] + start: Location, + #[allow(clippy::missing_docs_in_private_items)] + end: Location, +} + +impl Span { + /// Create a new `Span` from the provided start and end locations. + const fn start_end(start: Location, end: Location) -> Self { + Self { start, end } + } + + /// Reduce this span to the provided range. + #[must_use = "this does not modify the original value"] + fn subspan(&self, range: impl Subspan) -> Self { + range.subspan(self) + } + + /// Obtain a `Span` pointing at the start of the pre-existing span. + #[must_use = "this does not modify the original value"] + const fn shrink_to_start(&self) -> Self { + Self { + start: self.start, + end: self.start, + } + } + + /// Obtain a `Span` pointing at the end of the pre-existing span. + #[must_use = "this does not modify the original value"] + const fn shrink_to_end(&self) -> Self { + Self { + start: self.end, + end: self.end, + } + } + + /// Create an error with the provided message at this span. + const fn error(self, message: &'static str) -> ErrorInner { + ErrorInner { + _message: message, + _span: self, + } + } + + /// Get the byte index that the span starts at. + const fn start_byte(&self) -> usize { + self.start.byte + } +} + +/// A trait for types that can be used to reduce a `Span`. +trait Subspan { + /// Reduce the provided `Span` to a new `Span`. + fn subspan(self, span: &Span) -> Span; +} + +impl Subspan for RangeFrom<usize> { + fn subspan(self, span: &Span) -> Span { + assert_eq!(span.start.line, span.end.line); + + Span { + start: Location { + line: span.start.line, + column: span.start.column + self.start, + byte: span.start.byte + self.start, + }, + end: span.end, + } + } +} + +impl Subspan for RangeTo<usize> { + fn subspan(self, span: &Span) -> Span { + assert_eq!(span.start.line, span.end.line); + + Span { + start: span.start, + end: Location { + line: span.start.line, + column: span.start.column + self.end - 1, + byte: span.start.byte + self.end - 1, + }, + } + } +} + +/// The internal error type. +struct ErrorInner { + /// The message displayed to the user. + _message: &'static str, + /// Where the error originated. + _span: Span, +} + +/// A complete error description. +struct Error { + /// The internal error. + _inner: ErrorInner, + /// The error needed for interoperability with the rest of `time`. + public: crate::error::InvalidFormatDescription, +} + +impl From<Error> for crate::error::InvalidFormatDescription { + fn from(error: Error) -> Self { + error.public + } +} |