servo_webvtt/

lib.rs

/* This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at https://mozilla.org/MPL/2.0/. */

use std::iter::Peekable;
use std::marker::PhantomData;
use std::mem;
use std::str::Chars;
use std::sync::LazyLock;

use html5ever::buffer_queue::{BufferQueue, SetResult};
use html5ever::tendril::StrTendril;
use markup5ever::small_char_set;
use regex::Regex;

/// <https://w3c.github.io/webvtt/#webvtt-cue-writing-direction>
#[derive(Debug, Default, PartialEq)]
pub enum WebVttWritingDirection {
    /// <https://w3c.github.io/webvtt/#webvtt-cue-horizontal-writing-direction>
    #[default]
    Horizontal,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-vertical-growing-left-writing-direction>
    VerticalGrowingLeft,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-vertical-growing-right-writing-direction>
    VerticalGrowingRight,
}

/// <https://w3c.github.io/webvtt/#webvtt-cue-text-alignment>
#[derive(Debug, Default, PartialEq)]
pub enum WebVttTextAlignment {
    /// <https://w3c.github.io/webvtt/#webvtt-cue-start-alignment>
    Start,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-center-alignment>
    #[default]
    Center,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-end-alignment>
    End,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-left-alignment>
    Left,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-right-alignment>
    Right,
}

/// <https://w3c.github.io/webvtt/#webvtt-cue-position-alignment>
#[derive(Debug, Default, PartialEq)]
pub enum WebVttPositionAlignment {
    /// <https://w3c.github.io/webvtt/#webvtt-cue-position-line-left-alignment>
    LineLeft,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-position-center-alignment>
    Center,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-position-line-right-alignment>
    LineRight,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-position-automatic-alignment>
    #[default]
    Auto,
}

/// <https://w3c.github.io/webvtt/#webvtt-cue-position>
#[derive(Clone, Debug, Default, PartialEq)]
pub enum WebVttLineAndPositionSetting {
    Double(f64),
    #[default]
    Auto,
}

/// <https://w3c.github.io/webvtt/#webvtt-cue-line-alignment>
#[derive(Debug, Default, PartialEq)]
pub enum WebVttLineAlignment {
    /// <https://w3c.github.io/webvtt/#webvtt-cue-line-start-alignment>
    #[default]
    Start,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-line-center-alignment>
    Center,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-line-end-alignment>
    End,
}

/// <https://w3c.github.io/webvtt/#webvtt-cue-snap-to-lines-flag>
/// This is an enum, since the default value is `true`
#[derive(Debug, Default, PartialEq)]
pub enum WebVttSnapToLines {
    #[default]
    Yes,
    No,
}

impl From<bool> for WebVttSnapToLines {
    fn from(boolean: bool) -> Self {
        match boolean {
            true => WebVttSnapToLines::Yes,
            false => WebVttSnapToLines::No,
        }
    }
}

/// <https://w3c.github.io/webvtt/#webvtt-cue-size>
/// This is a struct, since the default value is 100
#[derive(Debug, PartialEq)]
pub struct WebVttCueSize(pub f64);

impl Default for WebVttCueSize {
    fn default() -> Self {
        Self(100.)
    }
}

/// <https://w3c.github.io/webvtt/#webvtt-cue>
#[derive(Debug, Default, PartialEq)]
pub struct WebVttCue {
    /// <https://html.spec.whatwg.org/multipage/#text-track-cue-identifier>
    pub identifier: String,
    /// <https://html.spec.whatwg.org/multipage/#text-track-cue-start-time>
    pub start_time: f64,
    /// <https://html.spec.whatwg.org/multipage/#text-track-cue-end-time>
    pub end_time: f64,
    /// <https://w3c.github.io/webvtt/#cue-text>
    pub text: String,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-writing-direction>
    pub writing_direction: WebVttWritingDirection,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-text-alignment>
    pub text_alignment: WebVttTextAlignment,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-position-alignment>
    pub position_alignment: WebVttPositionAlignment,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-position>
    pub position: WebVttLineAndPositionSetting,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-line-alignment>
    pub line_alignment: WebVttLineAlignment,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-line>
    pub line: WebVttLineAndPositionSetting,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-snap-to-lines-flag>
    pub snap_to_lines: WebVttSnapToLines,
    /// <https://w3c.github.io/webvtt/#webvtt-cue-size>
    pub size: WebVttCueSize,
}

#[derive(Debug, PartialEq)]
pub enum WebVttParserError {
    InvalidHeader,
}

impl std::fmt::Display for WebVttParserError {
    fn fmt(&self, formatter: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
        match self {
            WebVttParserError::InvalidHeader => write!(formatter, "Invalid WebVTT header in file"),
        }
    }
}

#[derive(Clone, Copy, Default, PartialEq)]
enum MostSignificantUnits {
    #[default]
    Minutes,
    Hours,
}

#[derive(Clone, Copy, Default, PartialEq)]
enum ParserState {
    #[default]
    FileTag,
    WhitespaceAfterFileTag,
    BeforeNewlineAfterFileTag,
    BeforeHeader,
    InBlockLoop,
    AfterBlockLoop,
    Region,
    Finished,
}

pub trait WebVttParserSink<Context> {
    fn consume_cue(&self, cx: &mut Context, cue: WebVttCue);
}

#[derive(Default)]
pub struct IncrementalWebVTTParser<Context, Sink: WebVttParserSink<Context>> {
    phantom: PhantomData<Context>,
    pub sink: Sink,
    buffer: BufferQueue,

    // Checkpoint values
    seen_cue: bool,
    seen_eof: bool,
    seen_arrow: bool,

    // State values
    in_header: bool,
    line_count: u32,
    state: ParserState,

    // Storage values
    current_line_in_block: StrTendril,
    current_buffer_in_block: String,

    current_cue_in_block: Option<WebVttCue>,
}

pub type ParserUpdate = Result<(), WebVttParserError>;

impl<Context, Sink> IncrementalWebVTTParser<Context, Sink>
where
    Sink: WebVttParserSink<Context>,
{
    /// <https://w3c.github.io/webvtt/#webvtt-parser-algorithm>
    pub fn new(sink: Sink) -> Self {
        Self {
            sink,
            phantom: Default::default(),
            buffer: Default::default(),
            seen_cue: Default::default(),
            seen_eof: Default::default(),
            seen_arrow: Default::default(),
            in_header: Default::default(),
            line_count: Default::default(),
            state: Default::default(),
            current_line_in_block: Default::default(),
            current_buffer_in_block: Default::default(),
            current_cue_in_block: Default::default(),
        }
    }

    pub fn end(&mut self, cx: &mut Context) -> ParserUpdate {
        self.seen_eof = true;
        self.step(cx)
    }

    pub fn parse_sync(&mut self, cx: &mut Context, input: &str) -> ParserUpdate {
        self.seen_eof = true;
        self.parse(cx, input)
    }

    /// <https://w3c.github.io/webvtt/#webvtt-parser-algorithm>
    pub fn parse(&mut self, cx: &mut Context, input: &str) -> ParserUpdate {
        // Step 1. Let input be the string being parsed, after conversion to Unicode,
        // and with the following transformations applied:
        // > Replace all U+0000 NULL characters by U+FFFD REPLACEMENT CHARACTERs.
        // > Replace each U+000D CARRIAGE RETURN U+000A LINE FEED (CRLF) character pair
        // > by a single U+000A LINE FEED (LF) character.
        // > Replace all remaining U+000D CARRIAGE RETURN characters by U+000A LINE FEED (LF) characters.
        // TODO
        // Step 2. Let position be a pointer into input, initially pointing at the start of the string.
        // In an incremental WebVTT parser, when this algorithm (or further algorithms that it uses)
        // moves the position pointer, the user agent must wait until appropriate further characters
        // from the byte stream have been added to input before moving the pointer,
        // so that the algorithm never reads past the end of the input string.
        // Once the byte stream has ended, and all characters have been added to input,
        // then the position pointer may, when so instructed by the algorithms,
        // be moved past the end of input.
        self.buffer.push_back(StrTendril::from(input));
        self.step(cx)
    }

    fn step(&mut self, cx: &mut Context) -> ParserUpdate {
        loop {
            let current_state = self.state;
            match current_state {
                // https://w3c.github.io/webvtt/#webvtt-parser-algorithm
                ParserState::FileTag => {
                    // https://w3c.github.io/webvtt/#webvtt-file-body
                    // > An optional U+FEFF BYTE ORDER MARK (BOM) character.
                    if self.buffer.peek().is_some_and(|c| c == '\u{FEFF}') {
                        let _ = self.buffer.next();
                    }
                    let Some(input) = self.buffer.eat("WEBVTT", u8::eq) else {
                        // Step 4. If input is less than six characters long, then abort these steps.
                        // The file does not start with the correct WebVTT file signature
                        // and was therefore not successfully processed.
                        if self.seen_eof {
                            return Err(WebVttParserError::InvalidHeader);
                        }
                        return Ok(());
                    };
                    // Step 5. If input is exactly six characters long but does not exactly equal "WEBVTT",
                    // then abort these steps. The file does not start with the correct WebVTT
                    // file signature and was therefore not successfully processed.
                    // Step 6. If input is more than six characters long but the first six characters
                    // do not exactly equal "WEBVTT", or the seventh character is not a U+0020 SPACE character,
                    // a U+0009 CHARACTER TABULATION (tab) character, or a U+000A LINE FEED (LF) character,
                    // then abort these steps. The file does not start with the correct WebVTT file signature
                    // and was therefore not successfully processed.
                    //
                    // We check the first part of this step here
                    if !input {
                        return Err(WebVttParserError::InvalidHeader);
                    }
                    self.state = ParserState::WhitespaceAfterFileTag;
                    continue;
                },
                // https://w3c.github.io/webvtt/#webvtt-parser-algorithm
                ParserState::WhitespaceAfterFileTag => {
                    let Some(seventh) = self.buffer.peek() else {
                        // Input is exactly six characters and is a valid header
                        if self.seen_eof {
                            self.state = ParserState::Finished;
                            continue;
                        }
                        return Ok(());
                    };
                    // Step 6. If input is more than six characters long but the first six characters
                    // do not exactly equal "WEBVTT", or the seventh character is not a U+0020 SPACE character,
                    // a U+0009 CHARACTER TABULATION (tab) character, or a U+000A LINE FEED (LF) character,
                    // then abort these steps. The file does not start with the correct WebVTT file signature
                    // and was therefore not successfully processed.
                    //
                    // We check the second part of this step here
                    if !matches!(seventh, '\u{0020}' | '\u{0009}' | '\u{000A}') {
                        return Err(WebVttParserError::InvalidHeader);
                    }
                    self.state = ParserState::BeforeNewlineAfterFileTag;
                    continue;
                },
                // https://w3c.github.io/webvtt/#webvtt-parser-algorithm
                ParserState::BeforeNewlineAfterFileTag => {
                    // Step 7. collect a sequence of code points that are not U+000A LINE FEED (LF) characters.
                    let Some(current_char) = self.buffer.next() else {
                        // Step 8. If position is past the end of input, then abort these steps.
                        // The file was successfully processed, but it contains no useful data and so
                        // no WebVTT cues were added to output.
                        if self.seen_eof {
                            self.state = ParserState::Finished;
                            continue;
                        }
                        return Ok(());
                    };
                    // Step 9. The character indicated by position is a U+000A LINE FEED (LF) character.
                    // Advance position to the next character in input.
                    if current_char == '\u{000A}' {
                        self.state = ParserState::BeforeHeader;
                    }
                },
                // https://w3c.github.io/webvtt/#webvtt-parser-algorithm
                ParserState::BeforeHeader => {
                    let Some(current_char) = self.buffer.peek() else {
                        // Step 10. If position is past the end of input, then abort these steps.
                        // The file was successfully processed, but it contains no useful data and
                        // so no WebVTT cues were added to output.
                        if self.seen_eof {
                            self.state = ParserState::Finished;
                            continue;
                        }
                        return Ok(());
                    };
                    // Step 11. Header: If the character indicated by position is not
                    // a U+000A LINE FEED (LF) character,
                    // then collect a WebVTT block with the in header flag set.
                    // Otherwise, advance position to the next character in input.
                    if current_char != '\u{000A}' {
                        self.in_header = true;
                        self.start_collecting_webvtt_block();
                    } else {
                        self.buffer.next();
                        self.state = ParserState::Region;
                    }
                },
                // https://w3c.github.io/webvtt/#collect-a-webvtt-block
                ParserState::InBlockLoop => {
                    let Some(current_char) =
                        self.buffer.pop_except_from(small_char_set!('\u{000A}'))
                    else {
                        // Step 11.3. If position is past the end of input, let seen EOF be true.
                        // Otherwise, the character indicated by position is a U+000A LINE FEED (LF) character;
                        // advance position to the next character in input.
                        //
                        // We check the first part of this step here
                        // Step 11.7. If seen EOF is true, break out of loop.
                        if self.seen_eof {
                            // It depends when we see the EOF whether there is still content in the buffer or not.
                            // In the case that the EOF is at the end of a line (e.g. no `\n` in between), we
                            // should copy the current line to the buffer. The buffer is then populated with the
                            // line as usual, so that the last line of a file can still be the cue text.
                            if !self.current_line_in_block.is_empty() {
                                if !self.current_buffer_in_block.is_empty() {
                                    self.current_buffer_in_block.push('\n');
                                }
                                self.current_buffer_in_block
                                    .push_str(&self.current_line_in_block);
                            }
                            self.state = ParserState::AfterBlockLoop;
                            continue;
                        }
                        return Ok(());
                    };
                    match current_char {
                        // Step 11.3. If position is past the end of input, let seen EOF be true.
                        // Otherwise, the character indicated by position is a U+000A LINE FEED (LF) character;
                        // advance position to the next character in input.
                        //
                        // We check the second part of this step here
                        SetResult::FromSet('\u{000A}') => {
                            // Step 11.2. Increment line count by 1.
                            self.line_count += 1;
                            if self.in_header {
                                self.state = ParserState::AfterBlockLoop;
                            } else {
                                // Step 11.4. If line contains the three-character substring "-->"
                                // (U+002D HYPHEN-MINUS, U+002D HYPHEN-MINUS, U+003E GREATER-THAN SIGN),
                                // then run these substeps:
                                if self.current_line_in_block.contains("-->") {
                                    // Step 11.4.1. If in header is not set and at least
                                    // one of the following conditions are true:
                                    //
                                    // We already checked for the header set after step 11.2.
                                    if
                                    // line count is 1
                                    self.line_count == 1
                                        // line count is 2 and seen arrow is false
                                        || (self.line_count == 2 && !self.seen_arrow)
                                    {
                                        // Step 11.4.1.1. Let seen arrow be true.
                                        self.seen_arrow = true;
                                        // Step 11.4.1.2. Let previous position be position.
                                        // TODO
                                        // Step 11.4.1.3. Cue creation: Let cue be a new WebVTT cue and initialize it as follows:
                                        // Step 11.4.1.3.1. Let cue’s text track cue identifier be buffer.
                                        let identifier = self.current_buffer_in_block.clone();
                                        // Step 11.4.1.4. Collect WebVTT cue timings and settings from line using regions for cue.
                                        // If that fails, let cue be null.
                                        // Otherwise, let buffer be the empty string and let seen cue be true.
                                        let cue = collect_webvtt_cue_timings_and_settings(
                                            identifier,
                                            &self.current_line_in_block,
                                        );
                                        let has_cue = cue.is_some();
                                        self.current_cue_in_block = cue;
                                        if has_cue {
                                            self.current_buffer_in_block.clear();
                                            self.current_line_in_block.clear();
                                            self.seen_cue = true;
                                        }
                                    } else {
                                        // Otherwise, let position be previous position and break out of loop.
                                        self.state = ParserState::AfterBlockLoop;
                                    }
                                    continue;
                                } else if self.current_line_in_block.is_empty() {
                                    // Step 11.5. Otherwise, if line is the empty string, break out of loop.
                                    self.state = ParserState::AfterBlockLoop;
                                } else {
                                    // Step 11.6. Otherwise, run these substeps:
                                    // Step 11.6.1. If in header is not set and line count is 2, run these substeps:
                                    // TODO
                                    // Step 11.6.2. If buffer is not the empty string,
                                    // append a U+000A LINE FEED (LF) character to buffer.
                                    if !self.current_buffer_in_block.is_empty() {
                                        self.current_buffer_in_block.push('\u{000A}');
                                    }
                                    // Step 11.6.3. Append line to buffer.
                                    self.current_buffer_in_block
                                        .push_str(&self.current_line_in_block);
                                    // Step 11.6.4. Let previous position be position.
                                    self.current_line_in_block.clear();
                                }
                            }
                            continue;
                        },
                        // Step 11.1. collect a sequence of code points that are not U+000A LINE FEED (LF) characters.
                        // Let line be those characters, if any.
                        SetResult::NotFromSet(current_tendril) => {
                            if !self.in_header {
                                self.current_line_in_block.push_tendril(&current_tendril);
                            }
                        },
                        _ => {
                            unreachable!();
                        },
                    }
                },
                ParserState::AfterBlockLoop => {
                    // https://w3c.github.io/webvtt/#collect-a-webvtt-block
                    // Step 12. If cue is not null, let the cue text of cue be buffer, and return cue.
                    // https://w3c.github.io/webvtt/#webvtt-parser-algorithm
                    // Step 14.2. If block is a WebVTT cue, add block to the text track list of cues output.
                    if let Some(mut cue) = self.current_cue_in_block.take() {
                        cue.text = self.current_buffer_in_block.clone();
                        self.sink.consume_cue(cx, cue);
                    }
                    // Step 14.3. Otherwise, if block is a CSS style sheet, add block to stylesheets.
                    // TODO
                    // Step 14.4. Otherwise, if block is a WebVTT region object, add block to regions.
                    // TODO
                    // Step 14.5. collect a sequence of code points that are U+000A LINE FEED (LF) characters.
                    let Some(current_char) = self.buffer.peek() else {
                        if self.seen_eof {
                            self.state = ParserState::Finished;
                            continue;
                        }
                        return Ok(());
                    };
                    if current_char == '\u{000A}' {
                        // Since we don't change the state here, it means that if the next character is
                        // also a newline, we re-enter this block and consume it again. Therefore, we
                        // only consume one-by-one.
                        let _ = self.buffer.next();
                    } else {
                        // If we were in the header block, then we should proceed with the next step
                        // which is collecting a region in step 12. Otherwise, we are in the general loop of
                        // step 14.

                        if mem::take(&mut self.in_header) {
                            self.state = ParserState::Region;
                        } else {
                            self.start_collecting_webvtt_block();
                        }
                    }
                },
                // https://w3c.github.io/webvtt/#webvtt-parser-algorithm
                ParserState::Region => {
                    // Step 12. collect a sequence of code points that are U+000A LINE FEED (LF) characters.
                    // TODO
                    self.start_collecting_webvtt_block();
                },
                ParserState::Finished => {
                    // Step 15. End: The file has ended. Abort these steps. The WebVTT parser has finished.
                    // The file was successfully processed.
                    return Ok(());
                },
            }
        }
    }

    /// <https://w3c.github.io/webvtt/#collect-a-webvtt-block>
    fn start_collecting_webvtt_block(&mut self) {
        // Step 2. Let line count be zero.
        self.line_count = 0;
        // Step 4. Let line be the empty string.
        self.current_line_in_block.clear();
        // Step 5. Let buffer be the empty string.
        self.current_buffer_in_block.clear();
        // Step 7. Let seen arrow be false.
        self.seen_arrow = false;
        // Step 8. Let cue be null.
        self.current_cue_in_block = None;
        self.state = ParserState::InBlockLoop;
    }
}

/// <https://w3c.github.io/webvtt/#collect-webvtt-cue-timings-and-settings>
fn collect_webvtt_cue_timings_and_settings(identifier: String, input: &str) -> Option<WebVttCue> {
    // Step 1. Let input be the string being parsed.
    //
    // Passed in as argument

    // Step 2. Let position be a pointer into input,
    // initially pointing at the start of the string.
    let mut position = input.chars().peekable();
    // Step 3. Skip whitespace.
    skip_whitespace(&mut position);
    // Step 4. Collect a WebVTT timestamp. If that algorithm fails,
    // then abort these steps and return failure.
    // Otherwise, let cue’s text track cue start time be the collected time.
    let start_time = collect_webvtt_timestamp(position.by_ref())?;
    // Step 5. Skip whitespace.
    skip_whitespace(&mut position);
    // Step 6. If the character at position is not a U+002D HYPHEN-MINUS character (-)
    // then abort these steps and return failure.
    // Otherwise, move position forwards one character.
    let _ = position.next().filter(|c| *c == '\u{002D}')?;
    // Step 7. If the character at position is not a U+002D HYPHEN-MINUS character (-)
    // then abort these steps and return failure.
    // Otherwise, move position forwards one character.
    let _ = position.next().filter(|c| *c == '\u{002D}')?;
    // Step 8. If the character at position is not a U+003E GREATER-THAN SIGN character (>)
    // then abort these steps and return failure.
    // Otherwise, move position forwards one character.
    let _ = position.next().filter(|c| *c == '\u{003E}')?;
    // Step 9. Skip whitespace.
    skip_whitespace(&mut position);
    // Step 10. Collect a WebVTT timestamp. If that algorithm fails,
    // then abort these steps and return failure.
    // Otherwise, let cue’s text track cue end time be the collected time.
    let end_time = collect_webvtt_timestamp(position.by_ref())?;
    // Step 11. Let remainder be the trailing substring of input starting at position.
    let remainder = collect_for_closure(&mut position, |_| true);
    // Step 12. Parse the WebVTT cue settings from remainder using regions for cue.
    let cue = WebVttCue {
        identifier,
        start_time,
        end_time,
        ..Default::default()
    };
    Some(parse_the_webvtt_cue_settings(cue, remainder))
}

/// <https://w3c.github.io/webvtt/#collect-a-webvtt-timestamp>
fn collect_webvtt_timestamp(position: &mut Peekable<Chars<'_>>) -> Option<f64> {
    // Step 1. Let input and position be the same variables
    // as those of the same name in the algorithm that invoked these steps.
    //
    // Passed in as argument

    // Step 2. Let most significant units be minutes.
    let mut most_significant_units = MostSignificantUnits::Minutes;
    // Step 3. If position is past the end of input, return an error and abort these steps.
    // Step 4. If the character indicated by position is not an ASCII digit,
    // then return an error and abort these steps.
    if !position.peek()?.is_ascii_digit() {
        return None;
    }
    // Step 5. Collect a sequence of code points that are ASCII digits,
    // and let string be the collected substring.
    let string = collect_ascii_digits(position);
    // Step 6. Interpret string as a base-ten integer. Let value1 be that integer.
    let mut value_1 = string.parse::<f64>().ok()?;
    // Step 7. If string is not exactly two characters in length,
    // or if value1 is greater than 59, let most significant units be hours.
    if string.len() != 2 || value_1 > 59_f64 {
        most_significant_units = MostSignificantUnits::Hours;
    }
    // Step 8. If position is beyond the end of input or if the character at position is
    // not a U+003A COLON character (:), then return an error and abort these steps.
    // Otherwise, move position forwards one character.
    let _ = position.next().filter(|c| *c == '\u{003A}')?;
    // Step 9. Collect a sequence of code points that are ASCII digits,
    // and let string be the collected substring.
    let string = collect_ascii_digits(position);
    // Step 10. If string is not exactly two characters in length,
    // return an error and abort these steps.
    if string.len() != 2 {
        return None;
    }
    // Step 11. Interpret string as a base-ten integer. Let value2 be that integer.
    let mut value_2 = string.parse::<f64>().ok()?;
    // Step 12. If most significant units is hours,
    // or if position is not beyond the end of input and the character
    // at position is a U+003A COLON character (:), run these substeps:
    let value_3: f64;
    if most_significant_units == MostSignificantUnits::Hours ||
        position.peek().is_some_and(|c| *c == '\u{003A}')
    {
        // Step 12.1. If position is beyond the end of input or if
        // the character at position is not a U+003A COLON character (:),
        // then return an error and abort these steps.
        // Otherwise, move position forwards one character.
        position.next().filter(|c| *c == '\u{003A}')?;
        // Step 12.2. Collect a sequence of code points that are ASCII digits,
        // and let string be the collected substring.
        let string = collect_ascii_digits(position);
        // Step 12.3. If string is not exactly two characters in length,
        // return an error and abort these steps.
        if string.len() != 2 {
            return None;
        }
        // Step 12.4. Interpret string as a base-ten integer. Let value3 be that integer.
        value_3 = string.parse::<f64>().ok()?;
    } else {
        // Otherwise (if most significant units is not hours,
        // and either position is beyond the end of input,
        // or the character at position is not a U+003A COLON character (:)),
        // let value3 have the value of value2,
        // then value2 have the value of value1, then let value1 equal zero.
        value_3 = value_2;
        value_2 = value_1;
        value_1 = 0_f64;
    }
    // Step 13. If position is beyond the end of input or if the character at
    // position is not a U+002E FULL STOP character (.),
    // then return an error and abort these steps.
    // Otherwise, move position forwards one character.
    position.next().filter(|c| *c == '\u{002E}')?;
    // Step 14. Collect a sequence of code points that are ASCII digits,
    // and let string be the collected substring.
    let string = collect_ascii_digits(position);
    // Step 15. If string is not exactly three characters in length,
    // return an error and abort these steps.
    if string.len() != 3 {
        return None;
    }
    // Step 16. Interpret string as a base-ten integer. Let value4 be that integer.
    let value_4 = string.parse::<f64>().ok()?;
    // Step 17. If value2 is greater than 59 or if value3 is greater than 59,
    // return an error and abort these steps.
    if value_2 > 59_f64 || value_3 > 59_f64 {
        return None;
    }
    // Step 18. Let result be value1×60×60 + value2×60 + value3 + value4∕1000.
    // Step 19. Return result.
    Some(value_1 * 60_f64 * 60_f64 + value_2 * 60_f64 + value_3 + value_4 / 1000_f64)
}

/// <https://w3c.github.io/webvtt/#parse-the-webvtt-cue-settings>
fn parse_the_webvtt_cue_settings(mut cue: WebVttCue, input: String) -> WebVttCue {
    // Step 1. Let settings be the result of splitting input on spaces.
    let settings = input.split_ascii_whitespace();
    // Step 2. For each token setting in the list settings, run the following substeps:
    'next_setting: for setting in settings {
        // Step 2.2. Let name be the leading substring of setting up to
        // and excluding the first U+003A COLON character (:) in that string.
        // Step 2.3. Let value be the trailing substring of setting starting from the
        // character immediately after the first U+003A COLON character (:) in that string.
        let Some((name, value)) = setting.split_once('\u{003A}') else {
            // Step 2.1. If setting does not contain a U+003A COLON character (:),
            // or if the first U+003A COLON character (:) in setting is either
            // the first or last character of setting,
            // then jump to the step labeled next setting.
            //
            // We check the first part here
            continue 'next_setting;
        };
        // Step 2.1. If setting does not contain a U+003A COLON character (:),
        // or if the first U+003A COLON character (:) in setting is either
        // the first or last character of setting,
        // then jump to the step labeled next setting.
        //
        // We check the second part here
        if name.is_empty() || value.is_empty() {
            continue 'next_setting;
        }
        // Step 2.4. Run the appropriate substeps that apply for the value of name, as follows:
        match name {
            // > If name is a case-sensitive match for "vertical"
            "vertical" => {
                // Step 2.4."vertical".1. If value is a case-sensitive match for the string "rl",
                // then let cue’s WebVTT cue writing direction be vertical growing left.
                if value == "rl" {
                    cue.writing_direction = WebVttWritingDirection::VerticalGrowingLeft;
                }
                // Step 2.4."vertical".2. Otherwise, if value is a case-sensitive match for the string "lr",
                // then let cue’s WebVTT cue writing direction be vertical growing right.
                if value == "lr" {
                    cue.writing_direction = WebVttWritingDirection::VerticalGrowingRight;
                }
                // Step 2.4."vertical".3. If cue’s WebVTT cue writing direction is not horizontal,
                // let cue’s WebVTT cue region be null (there are no vertical regions).
                // TODO
            },
            // > If name is a case-sensitive match for "line"
            "line" => {
                // Step 2.4."line".1. If value contains a U+002C COMMA character (,),
                // then let linepos be the leading substring of value up to and excluding the
                // first U+002C COMMA character (,) in that string and let linealign be the
                // trailing substring of value starting from the character immediately after the
                // first U+002C COMMA character (,) in that string.
                // Step 2.4."line".2. Otherwise let linepos be the full value string and linealign be null.
                let (linepos, linealign) = value
                    .split_once('\u{002C}')
                    .map(|(linepos, linealign)| (linepos, Some(linealign)))
                    .unwrap_or((value, None));

                // Step 2.4."line".4. If the last character in linepos is a U+0025 PERCENT SIGN character (%)
                let last_char_is_percentage =
                    linepos.chars().last().is_some_and(|c| c == '\u{0025}');
                let number = if last_char_is_percentage {
                    // If parse a percentage string from linepos doesn’t fail,
                    // let number be the returned percentage, otherwise jump to the step labeled next setting.
                    let Some(number) = parse_a_percentage_string(linepos) else {
                        continue 'next_setting;
                    };
                    number
                } else {
                    let mut chars = linepos.chars().peekable();
                    let mut has_at_least_one_dot = false;
                    let mut last_char: Option<char> = None;
                    let mut at_least_one_digit = false;
                    while let Some(current_char) = chars.next() {
                        match current_char {
                            // Step 2.4."line".4.2. If any character in linepos other than the first character is
                            // a U+002D HYPHEN-MINUS character (-), then jump to the step labeled next setting.
                            '\u{002D}' => {
                                if last_char.is_some() {
                                    continue 'next_setting;
                                }
                            },
                            '\u{002E}' => {
                                // Step 2.4."line".4.3. If there are more than one U+002E DOT characters (.),
                                // then jump to the step labeled next setting.
                                if has_at_least_one_dot {
                                    continue 'next_setting;
                                }
                                has_at_least_one_dot = true;
                                // Step 2.4."line".4.4. If there is a U+002E DOT character (.)
                                // and the character before or the character after is not an ASCII digit,
                                // or if the U+002E DOT character (.) is the first or the last character,
                                // then jump to the step labeled next setting.
                                if last_char.is_none_or(|c| !c.is_ascii_digit()) ||
                                    chars.peek().is_none_or(|c| !c.is_ascii_digit())
                                {
                                    continue 'next_setting;
                                }
                            },
                            _ => {
                                // Step 2.4."line".4.1. If linepos contains any characters other than
                                // U+002D HYPHEN-MINUS characters (-), ASCII digits, and U+002E DOT character (.),
                                // then jump to the step labeled next setting.
                                if !current_char.is_ascii_digit() {
                                    continue 'next_setting;
                                }
                                at_least_one_digit = true;
                            },
                        }
                        last_char = Some(current_char);
                    }
                    // Step 2.4."line".3. If linepos does not contain at least one ASCII digit,
                    // then jump to the step labeled next setting.
                    if !at_least_one_digit {
                        continue 'next_setting;
                    }
                    // Step 2.4."line".4.5. Let number be the result of parsing linepos using the
                    // rules for parsing floating-point number values. [HTML]
                    let Some(number) = linepos.parse::<f64>().ok() else {
                        // Step 2.4."line".4.6. If number is an error,
                        // then jump to the step labeled next setting.
                        continue 'next_setting;
                    };
                    number
                };
                match linealign {
                    // Step 2.4."line".5. If linealign is a case-sensitive match for the string "start",
                    // then let cue’s WebVTT cue line alignment be start alignment.
                    Some("start") => {
                        cue.line_alignment = WebVttLineAlignment::Start;
                    },
                    // Step 2.4."line".6. If linealign is a case-sensitive match for the string "center",
                    // then let cue’s WebVTT cue line alignment be center alignment.
                    Some("center") => {
                        cue.line_alignment = WebVttLineAlignment::Center;
                    },
                    // Step 2.4."line".7. If linealign is a case-sensitive match for the string "end",
                    // then let cue’s WebVTT cue line alignment be end alignment.
                    Some("end") => {
                        cue.line_alignment = WebVttLineAlignment::End;
                    },
                    // Step 2.4."line".8. Otherwise, if linealign is not null,
                    // then jump to the step labeled next setting.
                    Some(_) => {
                        continue 'next_setting;
                    },
                    _ => {},
                }
                // Step 2.4."line".9. Let cue’s WebVTT cue line be number.
                cue.line = WebVttLineAndPositionSetting::Double(number);
                // Step 2.4."line".10. If the last character in linepos is a U+0025 PERCENT SIGN character (%),
                // then let cue’s WebVTT cue snap-to-lines flag be false. Otherwise, let it be true.
                cue.snap_to_lines = (!last_char_is_percentage).into();
                // If cue’s WebVTT cue line is not auto,
                // let cue’s WebVTT cue region be null
                // (the cue has been explicitly positioned with a line offset
                // and thus drops out of the region).
                // TODO
            },
            // > If name is a case-sensitive match for "position"
            "position" => {
                // Step 2.4."position".1. If value contains a U+002C COMMA character (,),
                // then let colpos be the leading substring of value up to and excluding the
                // first U+002C COMMA character (,) in that string and let colalign be the
                // trailing substring of value starting from the character immediately after the
                // first U+002C COMMA character (,) in that string.
                // Step 2.4."position".2. Otherwise let colpos be the full value string and colalign be null.
                let (colpos, colalign) = value
                    .split_once('\u{002C}')
                    .map(|(colpos, colalign)| (colpos, Some(colalign)))
                    .unwrap_or((value, None));
                // Step 2.4."position".3. If parse a percentage string from colpos doesn’t fail,
                // let number be the returned percentage,
                // otherwise jump to the step labeled next setting
                // (position’s value remains the special value auto).
                let Some(number) = parse_a_percentage_string(colpos) else {
                    continue 'next_setting;
                };
                match colalign {
                    // Step 2.4."position".4. If colalign is a case-sensitive match for the string "line-left",
                    // then let cue’s WebVTT cue position alignment be line-left alignment.
                    Some("line-left") => {
                        cue.position_alignment = WebVttPositionAlignment::LineLeft;
                    },
                    // Step 2.4."position".5. Otherwise, if colalign is a case-sensitive match for the string "center",
                    // then let cue’s WebVTT cue position alignment be center alignment.
                    Some("center") => {
                        cue.position_alignment = WebVttPositionAlignment::Center;
                    },
                    // Step 2.4."position".6. Otherwise, if colalign is a case-sensitive match for the string "line-right",
                    // then let cue’s WebVTT cue position alignment be line-right alignment.
                    Some("line-right") => {
                        cue.position_alignment = WebVttPositionAlignment::LineRight;
                    },
                    // Step 2.4."position".7. Otherwise, if colalign is not null,
                    // then jump to the step labeled next setting.
                    Some(_) => {
                        continue 'next_setting;
                    },
                    _ => {},
                }
                // Step 2.4."position".8. Let cue’s position be number.
                cue.position = WebVttLineAndPositionSetting::Double(number);
            },
            // > If name is a case-sensitive match for "size"
            "size" => {
                // Step 2.4."size".1. If parse a percentage string from value doesn’t fail,
                // let number be the returned percentage,
                // otherwise jump to the step labeled next setting.
                let Some(number) = parse_a_percentage_string(value) else {
                    continue 'next_setting;
                };
                // Step 2.4."size".2. Let cue’s WebVTT cue size be number.
                cue.size = WebVttCueSize(number);
                // Step 2.4."size".3. If cue’s WebVTT cue size is not 100,
                // let cue’s WebVTT cue region be null
                // (the cue has been explicitly sized and thus drops out of the region).
                // TODO
            },
            // > If name is a case-sensitive match for "align"
            "align" => {
                // Step 2.4."align".1. If value is a case-sensitive match for the string "start",
                // then let cue’s WebVTT cue text alignment be start alignment.
                if value == "start" {
                    cue.text_alignment = WebVttTextAlignment::Start;
                }
                // Step 2.4."align".2. If value is a case-sensitive match for the string "center",
                // then let cue’s WebVTT cue text alignment be center alignment.
                if value == "center" {
                    cue.text_alignment = WebVttTextAlignment::Center;
                }
                // Step 2.4."align".3. If value is a case-sensitive match for the string "end",
                // then let cue’s WebVTT cue text alignment be end alignment.
                if value == "end" {
                    cue.text_alignment = WebVttTextAlignment::End;
                }
                // Step 2.4."align".4. If value is a case-sensitive match for the string "left",
                // then let cue’s WebVTT cue text alignment be left alignment.
                if value == "left" {
                    cue.text_alignment = WebVttTextAlignment::Left;
                }
                // Step 2.4."align".5. If value is a case-sensitive match for the string "right",
                // then let cue’s WebVTT cue text alignment be right alignment.
                if value == "right" {
                    cue.text_alignment = WebVttTextAlignment::Right;
                }
            },
            _ => {},
        }
    }
    cue
}

fn collect_for_closure<F>(position: &mut Peekable<Chars<'_>>, f: F) -> String
where
    F: FnOnce(&char) -> bool + Copy,
{
    let mut string = String::new();
    while let Some(next) = position.next_if(f) {
        string.push(next);
    }
    string
}

fn collect_ascii_digits(position: &mut Peekable<Chars<'_>>) -> String {
    collect_for_closure(position, char::is_ascii_digit)
}

fn skip_whitespace(position: &mut Peekable<Chars<'_>>) {
    collect_for_closure(position, |c| matches!(c, '\r' | '\n' | '\t' | ' '));
}

/// <https://w3c.github.io/webvtt/#webvtt-percentage>
static WEB_VTT_PERCENTAGE_GRAMMAR: LazyLock<Regex> =
    LazyLock::new(|| Regex::new(r#"^(?P<number>[0-9]+(\.[0-9]+)?)%$"#).unwrap());

/// <https://w3c.github.io/webvtt/#parse-a-percentage-string>
fn parse_a_percentage_string(input: &str) -> Option<f64> {
    // Step 1. Let input be the string being parsed.
    //
    // Passed in as argument

    // Step 2. If input does not match the syntax for a WebVTT percentage, then fail.
    let captures = WEB_VTT_PERCENTAGE_GRAMMAR.captures(input)?;
    // Step 3. Remove the last character from input.
    let input = captures.name("number").expect("Must always have a capture");
    // Step 4. Let percentage be the result of parsing input using the rules for parsing floating-point number values. [HTML]
    // Step 5. If percentage is an error, is less than 0, or is greater than 100, then fail.
    // Step 6. Return percentage.
    input
        .as_str()
        .trim()
        .parse::<f64>()
        .ok()
        .filter(|percentage| *percentage >= 0. && *percentage <= 100.)
}

#[cfg(any(test, feature = "test-util"))]
pub mod shared_test_setup;

#[cfg(test)]
mod tests {
    use crate::WebVttParserError;
    use crate::shared_test_setup::{compute_result_in_seconds, parser_with_dummy_sink};

    #[test]
    fn test_header_in_two_chunks() {
        let mut parser = parser_with_dummy_sink();
        assert_eq!(parser.parse(&mut (), "WEB"), Ok(()));
        assert_eq!(parser.parse(&mut (), "VTT"), Ok(()));
        assert_eq!(parser.end(&mut ()), Ok(()));
    }

    #[test]
    fn test_invalid_header_in_two_chunks() {
        let mut parser = parser_with_dummy_sink();
        assert_eq!(parser.parse(&mut (), "WEB"), Ok(()));
        assert_eq!(
            parser.parse(&mut (), "NOT"),
            Err(WebVttParserError::InvalidHeader)
        );
    }

    #[test]
    fn test_valid_space_character_after_header() {
        let mut parser = parser_with_dummy_sink();
        assert_eq!(parser.parse_sync(&mut (), "WEBVTT "), Ok(()));
    }

    #[test]
    fn test_no_space_character_after_header_multiple_chunks() {
        let mut parser = parser_with_dummy_sink();
        assert_eq!(parser.parse(&mut (), "WEB"), Ok(()));
        assert_eq!(parser.parse(&mut (), "VTT"), Ok(()));
        assert_eq!(
            parser.parse(&mut (), "2"),
            Err(WebVttParserError::InvalidHeader)
        );
    }

    mod cue_settings {
        use crate::tests::compute_result_in_seconds;
        use crate::{WebVttCue, collect_webvtt_cue_timings_and_settings};

        #[test]
        fn test_parses_cue_correctly() {
            assert_eq!(
                collect_webvtt_cue_timings_and_settings(
                    Default::default(),
                    "01:10:03.000 --> 02:20:23.000"
                ),
                Some(WebVttCue {
                    start_time: compute_result_in_seconds(1., 10., 3., 0.),
                    end_time: compute_result_in_seconds(2., 20., 23., 0.),
                    ..Default::default()
                })
            );
        }

        #[test]
        fn test_does_not_require_whitespace_around_arrow() {
            assert_eq!(
                collect_webvtt_cue_timings_and_settings(
                    Default::default(),
                    "01:10:03.000-->02:20:23.000"
                ),
                Some(WebVttCue {
                    start_time: compute_result_in_seconds(1., 10., 3., 0.),
                    end_time: compute_result_in_seconds(2., 20., 23., 0.),
                    ..Default::default()
                })
            );
        }

        #[test]
        fn test_can_handle_tabs_around_arrow() {
            assert_eq!(
                collect_webvtt_cue_timings_and_settings(
                    Default::default(),
                    "01:10:03.000\t-->\t02:20:23.000"
                ),
                Some(WebVttCue {
                    start_time: compute_result_in_seconds(1., 10., 3., 0.),
                    end_time: compute_result_in_seconds(2., 20., 23., 0.),
                    ..Default::default()
                })
            );
        }

        #[test]
        fn test_arrow_too_short_is_invalid() {
            assert_eq!(
                collect_webvtt_cue_timings_and_settings(
                    Default::default(),
                    "01:10:03.000 -> t02:20:23.000"
                ),
                None
            );
        }

        #[test]
        fn test_arrow_too_long_is_invalid() {
            assert_eq!(
                collect_webvtt_cue_timings_and_settings(
                    Default::default(),
                    "01:10:03.000 ---> t02:20:23.000"
                ),
                None
            );
        }

        #[test]
        fn test_skips_whitespace_at_start() {
            assert_eq!(
                collect_webvtt_cue_timings_and_settings(
                    Default::default(),
                    "  01:10:03.000 --> 02:20:23.000"
                ),
                Some(WebVttCue {
                    start_time: compute_result_in_seconds(1., 10., 3., 0.),
                    end_time: compute_result_in_seconds(2., 20., 23., 0.),
                    ..Default::default()
                })
            );
        }
    }

    mod timestamp {
        use crate::collect_webvtt_timestamp;
        use crate::tests::compute_result_in_seconds;

        fn parse_timestamp(input: &str) -> Option<f64> {
            collect_webvtt_timestamp(&mut input.chars().peekable())
        }

        #[test]
        fn test_parses_start_timestamp_correctly() {
            assert_eq!(
                parse_timestamp("01:10:03.000"),
                Some(compute_result_in_seconds(1., 10., 3., 0.))
            );
        }

        #[test]
        fn test_parses_maximum_minute_timestamp() {
            assert_eq!(
                parse_timestamp("10:59:03.000"),
                Some(compute_result_in_seconds(10., 59., 3., 0.))
            );
        }

        #[test]
        fn test_parses_maximum_second_timestamp() {
            assert_eq!(
                parse_timestamp("10:04:59.000"),
                Some(compute_result_in_seconds(10., 4., 59., 0.))
            );
        }

        #[test]
        fn test_hours_more_than_59_ensures_three_values() {
            assert_eq!(parse_timestamp("60:04.000"), None);
        }

        #[test]
        fn test_first_value_below_60_implies_minutes() {
            assert_eq!(
                parse_timestamp("59:04.000"),
                Some(compute_result_in_seconds(0., 59., 4., 0.))
            );
        }

        #[test]
        fn test_seconds_cannot_exceed_59() {
            assert_eq!(parse_timestamp("05:60.000"), None);
        }

        #[test]
        fn test_minutes_cannot_exceed_59() {
            assert_eq!(parse_timestamp("05:60:35.000"), None);
        }

        #[test]
        fn test_hours_can_exceed_59() {
            assert_eq!(
                parse_timestamp("60:20:35.000"),
                Some(compute_result_in_seconds(60., 20., 35., 0.))
            );
        }
    }
}