#![doc = include_str!("../README.md")]

use std::borrow::Cow;
use std::collections::VecDeque;
use std::error::Error;
use std::fmt::Display;
use std::iter::Enumerate;
use std::mem::take;
use std::str::CharIndices;

const NEWLINE: char = '\u{000A}';

/// Parses the contents of a .wsv (whitespace separated value) file.
/// The result is either a 2 dimensional vec where the outer layer is
/// the line and the inner layer is the column or a WSVError. '-' values will be
/// converted to 'None' and all other values will be 'Some'
///
/// For example, given the wsv file:
/// ```wsv
/// 1 -
/// 3 4
/// ```
/// the returned value would be [[Some(1), None], [Some(3), Some(4)]]
///
/// The source text will be sanitized. That is to say:
/// 1. All `"/"` escape sequences within quoted strings will be replaced with
/// `\n` inside the string.
/// 2. All `""` (two double-quote character) escape sequences within strings
/// will be replaced with `"` (one double-quote character)
/// 3. Any wrapping quotes around a string will be removed. Ex. `"hello world!"`
/// will just be `hello world!` in the output.
pub fn parse(source_text: &str) -> Result<Vec<Vec<Option<Cow<'_, str>>>>, WSVError> {
    // Just use the vec default size of 0.
    parse_with_col_count(source_text, 0)
}

/// Same as parse (see the documentation there for behavior details),
/// but accepts an expected column count to avoid unnecessary reallocations
/// of the Vecs.
pub fn parse_with_col_count(
    source_text: &str,
    col_count: usize,
) -> Result<Vec<Vec<Option<Cow<'_, str>>>>, WSVError> {
    let mut result = Vec::new();
    result.push(Vec::with_capacity(col_count));
    let mut last_line_num = 0;

    for fallible_token in WSVTokenizer::new(source_text) {
        let token = fallible_token?;
        match token {
            WSVToken::LF => {
                result.push(Vec::with_capacity(col_count));
                last_line_num += 1;
            }
            WSVToken::Null => {
                result[last_line_num].push(None);
            }
            WSVToken::Value(value) => {
                result[last_line_num].push(Some(value));
            }
            WSVToken::Comment(_) => {}
        }
    }

    // We pushed extra vecs on eagerly every time we saw an
    // LF, so pop the last one if it was empty.
    if result[last_line_num].len() == 0 {
        result.pop();
    }

    Ok(result)
}

/// Same as parse, (see the documentation there for behavior details),
/// but parses lazily. The input will be read a single line at a time,
/// allowing for lazy loading of very large files to be pushed thorugh
/// this API without issues. If you need to be even lazier (loading the
/// file token-by-token), use WSVLazyTokenizer directly.
pub fn parse_lazy<Chars: IntoIterator<Item = char>>(source_text: Chars) -> WSVLineIterator<Chars> {
    WSVLineIterator::new(source_text)
}

/// An iterator over the lines of a WSV file. This is used to allow lazy
/// parsing of files that do not fit into memory.
pub struct WSVLineIterator<Chars>
where
    Chars: IntoIterator<Item = char>,
{
    tokenizer: WSVLazyTokenizer<Chars>,
    lookahead_error: Option<WSVError>,
    errored: bool,
    finished: bool,
}

impl<Chars> WSVLineIterator<Chars>
where
    Chars: IntoIterator<Item = char>,
{
    fn new(source_text: Chars) -> Self {
        Self {
            tokenizer: WSVLazyTokenizer::new(source_text),
            lookahead_error: None,
            errored: false,
            finished: false,
        }
    }
}

impl<Chars> Iterator for WSVLineIterator<Chars>
where
    Chars: IntoIterator<Item = char>,
{
    type Item = Result<Vec<Option<String>>, WSVError>;

    fn next(&mut self) -> Option<Self::Item> {
        if self.finished {
            return None;
        }

        if let Some(err) = take(&mut self.lookahead_error) {
            return Some(Err(err));
        }

        if self.errored {
            return None;
        }

        let mut line = Vec::new();
        loop {
            let token = self.tokenizer.next();
            match token {
                None => {
                    if line.is_empty() {
                        return None;
                    } else {
                        return Some(Ok(line));
                    }
                }
                Some(token) => match token {
                    Err(err) => {
                        self.errored = true;
                        if line.is_empty() {
                            return Some(Err(err));
                        } else {
                            self.lookahead_error = Some(err);
                            return Some(Ok(line));
                        }
                    }
                    Ok(token) => match token {
                        OwnedWSVToken::Comment(_) => {}
                        OwnedWSVToken::LF => return Some(Ok(line)),
                        OwnedWSVToken::Null => line.push(None),
                        OwnedWSVToken::Value(val) => line.push(Some(val)),
                    },
                },
            }
        }
    }
}

/// A struct for writing values to a .wsv file.
pub struct WSVWriter<OuterIter, InnerIter, BorrowStr>
where
    OuterIter: IntoIterator<Item = InnerIter>,
    InnerIter: IntoIterator<Item = Option<BorrowStr>>,
    BorrowStr: AsRef<str>,
{
    align_columns: ColumnAlignment,
    values: Enumerate<OuterIter::IntoIter>,
    current_inner: Option<InnerIter::IntoIter>,
    lookahead_chars: VecDeque<char>,
}

impl<OuterIter, InnerIter, BorrowStr> WSVWriter<OuterIter, InnerIter, BorrowStr>
where
    OuterIter: Iterator<Item = InnerIter>,
    InnerIter: IntoIterator<Item = Option<BorrowStr>>,
    BorrowStr: AsRef<str> + From<&'static str> + ToString,
{
    pub fn new<OuterInto>(values: OuterInto) -> Self
    where
        OuterInto: IntoIterator<Item = InnerIter, IntoIter = OuterIter>,
    {
        let outer_into = values.into_iter();

        Self {
            align_columns: ColumnAlignment::default(),
            values: outer_into.enumerate(),
            current_inner: None,
            lookahead_chars: VecDeque::new(),
        }
    }

    /// Sets the column alignment of this Writer.
    /// Note: Left and Right alignments cannot use lazy
    /// evaluation, so do not set this value if you need
    /// lazy evaluation.
    pub fn align_columns(mut self, alignment: ColumnAlignment) -> Self {
        self.align_columns = alignment;
        self
    }

    pub fn to_string(self) -> String {
        match self.align_columns {
            ColumnAlignment::Packed => self.collect::<String>(),
            ColumnAlignment::Left | ColumnAlignment::Right => {
                let mut max_col_widths = Vec::new();

                let vecs = self
                    .values
                    .map(|(line_num, inner)| {
                        (
                            line_num,
                            inner
                                .into_iter()
                                .enumerate()
                                .map(|(index, value)| {
                                    // Figure out 2 things while consuming the iterators:
                                    // 1. Whether or not the value needs quotes
                                    // 2. The length of the string we will be writing
                                    let mut needs_quotes = false;
                                    let mut value_len = 0;
                                    match value.as_ref() {
                                        None => value_len = 1,
                                        Some(val) => {
                                            for ch in val.as_ref().chars() {
                                                match ch {
                                                    // account for escape sequences.
                                                    '\n' => {
                                                        value_len += 3;
                                                        needs_quotes = true;
                                                    }
                                                    '"' => {
                                                        value_len += 2;
                                                        needs_quotes = true;
                                                    }
                                                    '#' => {
                                                        value_len += 1;
                                                        needs_quotes = true;
                                                    }
                                                    ch => {
                                                        value_len += 1;
                                                        needs_quotes |= ch == '#'
                                                            || WSVTokenizer::is_whitespace(ch);
                                                    }
                                                }
                                            }
                                        }
                                    }

                                    if needs_quotes {
                                        value_len += 2;
                                    }
                                    match max_col_widths.get_mut(index) {
                                        None => max_col_widths.push(value_len),
                                        Some(longest_len) => {
                                            if value_len > *longest_len {
                                                *longest_len = value_len
                                            }
                                        }
                                    }
                                    return (needs_quotes, value_len, value);
                                })
                                .collect::<Vec<_>>(),
                        )
                    })
                    .collect::<Vec<_>>();

                let mut result = String::new();
                for (line_num, line) in vecs {
                    if line_num != 0 {
                        result.push('\n');
                    }

                    for (i, col) in line.into_iter().enumerate() {
                        if i != 0 {
                            result.push(' ');
                        }

                        let value = match col.2.as_ref() {
                            None => "-",
                            Some(string) => string.as_ref(),
                        };

                        if let &ColumnAlignment::Right = &self.align_columns {
                            for _ in col.1..max_col_widths[i] {
                                result.push(' ');
                            }
                        }

                        if col.0 {
                            result.push('"');
                        }

                        for ch in value.chars() {
                            if ch == '\n' {
                                result.push('"');
                                result.push('/');
                                result.push('"');
                            } else if ch == '"' {
                                result.push('"');
                                result.push('"');
                            } else {
                                result.push(ch);
                            }
                        }

                        if col.0 {
                            result.push('"');
                        }

                        if let &ColumnAlignment::Left = &self.align_columns {
                            for _ in col.1..max_col_widths[i] {
                                result.push(' ');
                            }
                        }
                    }
                }

                result
            }
        }
    }
}

impl<OuterIter, InnerIter, BorrowStr> Iterator for WSVWriter<OuterIter, InnerIter, BorrowStr>
where
    OuterIter: Iterator<Item = InnerIter>,
    InnerIter: IntoIterator<Item = Option<BorrowStr>>,
    BorrowStr: AsRef<str> + From<&'static str> + ToString,
{
    type Item = char;
    fn next(&mut self) -> Option<Self::Item> {
        loop {
            if let Some(ch) = self.lookahead_chars.pop_front() {
                return Some(ch);
            }

            if let Some(inner_mut) = self.current_inner.as_mut() {
                match inner_mut.next() {
                    None => {
                        self.current_inner = None;
                    }
                    Some(next_string_like) => match next_string_like {
                        None => {
                            self.lookahead_chars.push_back(' ');
                            return Some('-');
                        }
                        Some(string_like) => {
                            let mut needs_quotes = false;
                            for ch in string_like.as_ref().chars() {
                                match ch {
                                    '\n' => {
                                        self.lookahead_chars.push_back('"');
                                        self.lookahead_chars.push_back('/');
                                        self.lookahead_chars.push_back('"');
                                        needs_quotes = true;
                                    }
                                    '"' => {
                                        self.lookahead_chars.push_back('"');
                                        self.lookahead_chars.push_back('"');
                                        needs_quotes = true;
                                    }
                                    ch => {
                                        self.lookahead_chars.push_back(ch);
                                        needs_quotes |=
                                            ch == '#' || WSVTokenizer::is_whitespace(ch);
                                    }
                                }
                            }
                            if needs_quotes {
                                self.lookahead_chars.push_front('"');
                                self.lookahead_chars.push_back('"');
                            }
                            self.lookahead_chars.push_back(' ');
                            continue;
                        }
                    },
                }
            }

            match self.values.next() {
                None => return None,
                Some((i, inner)) => {
                    self.current_inner = Some(inner.into_iter());
                    if i != 0 {
                        return Some('\n');
                    }
                }
            }
        }
    }
}
#[derive(Default)]
pub enum ColumnAlignment {
    Left,
    Right,
    #[default]
    Packed,
}

/// A tokenizer for the .wsv (whitespace separated value)
/// file format. This struct implements Iterator, so to
/// extract the tokens use your desired iterator method
/// or a standard for loop.
pub struct WSVTokenizer<'wsv> {
    source: &'wsv str,
    chars: CharIndices<'wsv>,
    peeked: Option<(usize, char)>,
    current_location: Location,
    lookahead_error: Option<WSVError>,
    errored: bool,
}

impl<'wsv> WSVTokenizer<'wsv> {
    /// Creates a .wsv tokenizer from .wsv source text.
    pub fn new(source_text: &'wsv str) -> Self {
        Self {
            source: source_text,
            chars: source_text.char_indices(),
            peeked: None,
            current_location: Location::default(),
            lookahead_error: None,
            errored: false,
        }
    }

    fn match_string(&mut self) -> Option<Result<WSVToken<'wsv>, WSVError>> {
        if self.match_char('"').is_none() {
            return None;
        }
        let mut chunks = Vec::with_capacity(1);
        let mut chunk_start = None;
        loop {
            if self.match_char('"').is_some() {
                if self.match_char('"').is_some() {
                    // a quote is ascii, so subtracting 1 bytes should always be safe.
                    let end_location = self.current_location.byte_index - 1;
                    chunks.push(&self.source[chunk_start.unwrap_or(end_location)..end_location]);
                    chunk_start = Some(self.current_location.byte_index);
                } else if self.match_char('/').is_some() {
                    if self.match_char('"').is_none() {
                        self.errored = true;
                        return Some(Err(WSVError {
                            err_type: WSVErrorType::InvalidStringLineBreak,
                            location: self.current_location.clone(),
                        }));
                    }
                    let end_index = self.current_location.byte_index - 2;
                    chunks.push(&self.source[chunk_start.unwrap_or(end_index)..end_index]);
                    chunks.push("\n");
                    chunk_start = Some(self.current_location.byte_index + 1);
                } else {
                    // a quote is ascii, so subtracting 1 bytes should always be safe.
                    chunks.push(
                        &self.source[chunk_start.unwrap_or(self.current_location.byte_index)
                            ..self.current_location.byte_index],
                    );
                    break;
                }
            } else if let Some(NEWLINE) = self.peek() {
                if let Some(NEWLINE) = self.peek() {
                    self.errored = true;
                    return Some(Err(WSVError {
                        err_type: WSVErrorType::StringNotClosed,
                        location: self.current_location.clone(),
                    }));
                }
            } else if let None = chunk_start {
                chunk_start = Some(match self.peek_location() {
                    None => self.source.len(),
                    Some(val) => val.byte_index,
                });
            } else if self.match_char_if(&mut |_| true).is_none() {
                return Some(Err(WSVError {
                    err_type: WSVErrorType::StringNotClosed,
                    location: self.peek_location().into_iter().next().unwrap_or_else(|| {
                        let mut loc = self.current_location.clone();
                        loc.byte_index = self.source.len();
                        return loc;
                    }),
                }));
            }
        }

        if chunks.len() == 1 {
            return Some(Ok(WSVToken::Value(Cow::Borrowed(chunks[0]))));
        } else {
            return Some(Ok(WSVToken::Value(Cow::Owned(
                chunks.into_iter().collect::<String>(),
            ))));
        }
    }

    fn match_char_while<F: FnMut(char) -> bool>(&mut self, mut predicate: F) -> Option<&'wsv str> {
        let mut start = None;
        loop {
            match self.match_char_if(&mut predicate) {
                None => break,
                Some((index, _)) => {
                    if let None = start {
                        start = Some(index);
                    }
                }
            }
        }

        let start_val = match start {
            None => return None,
            Some(val) => val,
        };

        // Just get the side effect of setting peeked
        self.peek();
        let end_val = match self.peeked.as_ref() {
            None => self.source.len(),
            Some((index, _)) => *index,
        };

        return Some(&self.source[start_val..end_val]);
    }

    fn match_char(&mut self, ch: char) -> Option<(usize, char)> {
        self.match_char_if(&mut |found_char| ch == found_char)
    }

    fn match_char_if<F: FnMut(char) -> bool>(
        &mut self,
        predicate: &mut F,
    ) -> Option<(usize, char)> {
        if let Some(found_char) = self.peek() {
            if predicate(found_char) {
                let consumed = take(&mut self.peeked);

                match consumed {
                    None => {
                        return None;
                    }
                    Some((i, ch)) => {
                        if ch == NEWLINE {
                            self.current_location.line += 1;
                            self.current_location.col = 1;
                        } else {
                            self.current_location.col += 1;
                        }
                        self.current_location.byte_index = i;
                    }
                }

                return consumed.clone();
            }
        }

        return None;
    }

    fn peek_location(&mut self) -> Option<Location> {
        self.peek_inner();
        match self.peeked.as_ref() {
            None => None,
            Some((i, _)) => {
                let mut peeked_pos = self.current_location.clone();
                peeked_pos.col += 1;
                peeked_pos.byte_index = *i;
                Some(peeked_pos)
            }
        }
    }

    fn peek(&mut self) -> Option<char> {
        match self.peek_inner() {
            None => None,
            Some(peeked) => Some(peeked.1),
        }
    }

    fn peek_inner(&mut self) -> Option<&(usize, char)> {
        if let None = self.peeked.as_ref() {
            self.peeked = self.chars.next();
        }
        self.peeked.as_ref()
    }

    fn is_whitespace(ch: char) -> bool {
        match ch {
            '\u{0009}' | '\u{000B}' | '\u{000C}' | '\u{000D}' | '\u{0020}' | '\u{0085}'
            | '\u{00A0}' | '\u{1680}' | '\u{2000}' | '\u{2001}' | '\u{2002}' | '\u{2003}'
            | '\u{2004}' | '\u{2005}' | '\u{2006}' | '\u{2007}' | '\u{2008}' | '\u{2009}'
            | '\u{200A}' | '\u{2028}' | '\u{2029}' | '\u{202F}' | '\u{205F}' | '\u{3000}' => true,
            _ => false,
        }
    }
}

impl<'wsv> Iterator for WSVTokenizer<'wsv> {
    type Item = Result<WSVToken<'wsv>, WSVError>;

    fn next(&mut self) -> Option<Self::Item> {
        if self.errored {
            return None;
        }
        if let Some(err) = take(&mut self.lookahead_error) {
            self.errored = true;
            return Some(Err(err));
        }
        self.match_char_while(|ch| Self::is_whitespace(ch));

        let str = self.match_string();
        if str.is_some() {
            let lookahead = self.peek().unwrap_or(' ');
            if lookahead != NEWLINE && lookahead != '#' && !Self::is_whitespace(lookahead) {
                self.lookahead_error = Some(WSVError {
                    location: self.current_location.clone(),
                    err_type: WSVErrorType::InvalidCharacterAfterString,
                });
            }
            return str;
        } else if self.match_char('#').is_some() {
            // Comment
            return Some(Ok(WSVToken::Comment(
                self.match_char_while(|ch| ch != NEWLINE).unwrap_or(""),
            )));
        } else if self.match_char(NEWLINE).is_some() {
            return Some(Ok(WSVToken::LF));
        } else {
            // Value
            match self.match_char_while(|ch| {
                if ch == NEWLINE {
                    return false;
                }
                if ch == '"' {
                    return false;
                }
                if ch == '#' {
                    return false;
                }
                if Self::is_whitespace(ch) {
                    return false;
                }
                return true;
            }) {
                Some(str) => {
                    if str == "-" {
                        return Some(Ok(WSVToken::Null));
                    }
                    if let Some('"') = self.peek() {
                        self.lookahead_error = Some(WSVError {
                            location: self.current_location.clone(),
                            err_type: WSVErrorType::InvalidDoubleQuoteAfterValue,
                        });
                    }
                    return Some(Ok(WSVToken::Value(Cow::Borrowed(str))));
                }
                None => None,
            }
        }
    }
}

/// A lazy tokenizer for the .wsv (whitespace separated
/// value) file format. This struct implements Iterator,
/// so to extract the tokens use your desired iterator
/// method or a standard for loop.
pub struct WSVLazyTokenizer<Chars: IntoIterator<Item = char>> {
    source: Chars::IntoIter,
    peeked: Option<char>,
    current_location: Location,
    lookahead_error: Option<WSVError>,
    errored: bool,
}

impl<Chars> WSVLazyTokenizer<Chars>
where
    Chars: IntoIterator<Item = char>,
{
    pub fn new(source_text: Chars) -> Self {
        Self {
            source: source_text.into_iter(),
            peeked: None,
            current_location: Location::default(),
            lookahead_error: None,
            errored: false,
        }
    }

    fn match_string(&mut self) -> Option<Result<OwnedWSVToken, WSVError>> {
        if self.match_char('"').is_none() {
            return None;
        }
        let mut result = String::new();
        loop {
            if self.match_char('"').is_some() {
                if self.match_char('"').is_some() {
                    // a quote is ascii, so subtracting 1 bytes should always be safe.
                    result.push('"');
                } else if self.match_char('/').is_some() {
                    if self.match_char('"').is_none() {
                        self.errored = true;
                        return Some(Err(WSVError {
                            err_type: WSVErrorType::InvalidStringLineBreak,
                            location: self.current_location.clone(),
                        }));
                    }
                    result.push('\n');
                } else {
                    return Some(Ok(OwnedWSVToken::Value(result)));
                }
            } else if let Some(NEWLINE) = self.peek() {
                if let Some(NEWLINE) = self.peek() {
                    self.errored = true;
                    return Some(Err(WSVError {
                        err_type: WSVErrorType::StringNotClosed,
                        location: self.current_location.clone(),
                    }));
                }
            } else if let Some(ch) = self.match_char_if(&mut |_| true) {
                result.push(ch);
            } else {
                return Some(Err(WSVError {
                    err_type: WSVErrorType::StringNotClosed,
                    location: self
                        .peek_location()
                        .into_iter()
                        .next()
                        .unwrap_or_else(|| self.current_location.clone()),
                }));
            }
        }
    }

    fn match_char_while<F: FnMut(char) -> bool>(&mut self, mut predicate: F) -> Option<String> {
        let mut str = String::new();
        loop {
            match self.match_char_if(&mut predicate) {
                None => break,
                Some(ch) => {
                    str.push(ch);
                }
            }
        }

        if str.len() == 0 {
            return None;
        } else {
            return Some(str);
        }
    }

    fn match_char(&mut self, ch: char) -> Option<char> {
        self.match_char_if(&mut |found_char| ch == found_char)
    }

    fn match_char_if<F: FnMut(char) -> bool>(&mut self, predicate: &mut F) -> Option<char> {
        if let Some(found_char) = self.peek() {
            if predicate(found_char) {
                let consumed = take(&mut self.peeked);

                match consumed {
                    None => {
                        return None;
                    }
                    Some(ch) => {
                        if ch == NEWLINE {
                            self.current_location.line += 1;
                            self.current_location.col = 1;
                        } else {
                            self.current_location.col += 1;
                        }
                        return Some(ch);
                    }
                }
            }
        }

        return None;
    }

    fn peek_location(&mut self) -> Option<Location> {
        self.peek_inner();
        match self.peeked.as_ref() {
            None => None,
            Some(_) => {
                let mut peeked_pos = self.current_location.clone();
                peeked_pos.col += 1;
                Some(peeked_pos)
            }
        }
    }

    fn peek(&mut self) -> Option<char> {
        match self.peek_inner() {
            None => None,
            Some(peeked) => Some(*peeked),
        }
    }

    fn peek_inner(&mut self) -> Option<&char> {
        if let None = self.peeked.as_ref() {
            self.peeked = self.source.next();
        }
        self.peeked.as_ref()
    }

    fn is_whitespace(ch: char) -> bool {
        match ch {
            '\u{0009}' | '\u{000B}' | '\u{000C}' | '\u{000D}' | '\u{0020}' | '\u{0085}'
            | '\u{00A0}' | '\u{1680}' | '\u{2000}' | '\u{2001}' | '\u{2002}' | '\u{2003}'
            | '\u{2004}' | '\u{2005}' | '\u{2006}' | '\u{2007}' | '\u{2008}' | '\u{2009}'
            | '\u{200A}' | '\u{2028}' | '\u{2029}' | '\u{202F}' | '\u{205F}' | '\u{3000}' => true,
            _ => false,
        }
    }
}

impl<Chars> Iterator for WSVLazyTokenizer<Chars>
where
    Chars: IntoIterator<Item = char>,
{
    type Item = Result<OwnedWSVToken, WSVError>;
    fn next(&mut self) -> Option<Self::Item> {
        if self.errored {
            return None;
        }
        if let Some(err) = take(&mut self.lookahead_error) {
            self.errored = true;
            return Some(Err(err));
        }
        self.match_char_while(|ch| Self::is_whitespace(ch));

        let str = self.match_string();
        if str.is_some() {
            let lookahead = self.peek().unwrap_or(' ');
            if lookahead != NEWLINE && lookahead != '#' && !Self::is_whitespace(lookahead) {
                self.lookahead_error = Some(WSVError {
                    location: self.current_location.clone(),
                    err_type: WSVErrorType::InvalidCharacterAfterString,
                });
            }
            return str;
        } else if self.match_char('#').is_some() {
            // Comment
            return Some(Ok(OwnedWSVToken::Comment(
                self.match_char_while(|ch| ch != NEWLINE)
                    .unwrap_or_else(|| "".to_string()),
            )));
        } else if self.match_char(NEWLINE).is_some() {
            return Some(Ok(OwnedWSVToken::LF));
        } else {
            // Value
            match self.match_char_while(|ch| {
                if ch == NEWLINE {
                    return false;
                }
                if ch == '"' {
                    return false;
                }
                if ch == '#' {
                    return false;
                }
                if Self::is_whitespace(ch) {
                    return false;
                }
                return true;
            }) {
                Some(str) => {
                    if str == "-" {
                        return Some(Ok(OwnedWSVToken::Null));
                    }
                    if let Some('"') = self.peek() {
                        self.lookahead_error = Some(WSVError {
                            location: self.current_location.clone(),
                            err_type: WSVErrorType::InvalidDoubleQuoteAfterValue,
                        });
                    }
                    return Some(Ok(OwnedWSVToken::Value(str)));
                }
                None => None,
            }
        }
    }
}

/// A collection of all token types in a WSV file.
#[derive(Debug, Clone)]
pub enum WSVToken<'wsv> {
    /// Represents a line feed character (ex. '\n')
    LF,
    /// Represents a null value in the input (ex. '-')
    Null,
    /// Represents a non-null value in the input (ex. 'value')
    Value(Cow<'wsv, str>),
    /// Represents a comment (ex. '# comment')
    Comment(&'wsv str),
}

/// A collection of all token types in a WSV file.
pub enum OwnedWSVToken {
    /// Represents a line feed character (ex. '\n')
    LF,
    /// Represents a null value in the input (ex. '-')
    Null,
    /// Represents a non-null value in the input (ex. 'value')
    Value(String),
    /// Represents a comment (ex. '# comment')
    Comment(String),
}

/// A struct to represent an error in a WSV file. This contains
/// both the type of error and location of the error in the source
/// text.
#[derive(Debug, Clone)]
pub struct WSVError {
    err_type: WSVErrorType,
    location: Location,
}

impl WSVError {
    pub fn err_type(&self) -> WSVErrorType {
        self.err_type
    }

    pub fn location(&self) -> Location {
        self.location.clone()
    }
}

impl Display for WSVError {
    fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
        let mut description = String::new();

        let location = self.location();
        description.push_str("(line: ");
        description.push_str(&location.line().to_string());
        description.push_str(", column: ");
        description.push_str(&location.col().to_string());
        description.push_str(") ");

        match self.err_type() {
            WSVErrorType::InvalidCharacterAfterString => {
                description.push_str("Invalid Character After String");
            }
            WSVErrorType::InvalidDoubleQuoteAfterValue => {
                description.push_str("Invalid Double Quote After Value");
            }
            WSVErrorType::InvalidStringLineBreak => {
                description.push_str("Invalid String Line Break");
            }
            WSVErrorType::StringNotClosed => {
                description.push_str("String Not Closed");
            }
        }

        write!(f, "{}", description)?;
        Ok(())
    }
}
impl Error for WSVError {}

/// For details on these error types, see the Parser Errors
/// section of [https://dev.stenway.com/WSV/Specification.html](https://dev.stenway.com/WSV/Specification.html)
#[derive(Clone, Copy, Debug, PartialEq, Eq)]
pub enum WSVErrorType {
    StringNotClosed,
    InvalidDoubleQuoteAfterValue,
    InvalidCharacterAfterString,
    InvalidStringLineBreak,
}

/// Represents a location in the source text
#[derive(Debug, Clone)]
pub struct Location {
    byte_index: usize,
    line: usize,
    col: usize,
}

impl Location {
    /// The line number in the source text.
    pub fn line(&self) -> usize {
        self.line
    }
    /// The column number in the source text.
    pub fn col(&self) -> usize {
        self.col
    }
}

impl Default for Location {
    fn default() -> Self {
        Self {
            byte_index: 0,
            line: 1,
            col: 1,
        }
    }
}

#[cfg(debug_assertions)]
mod tests {
    use crate::{
        parse_lazy, OwnedWSVToken, WSVError, WSVErrorType, WSVLazyTokenizer, WSVToken, WSVTokenizer,
    };

    use super::{parse, WSVWriter};
    use std::{borrow::Cow, fmt::write};

    #[test]
    fn read_and_write() {
        let str = include_str!("../tests/1_stenway.com");
        let result = parse(str).unwrap();

        let result_str = WSVWriter::new(result)
            .align_columns(super::ColumnAlignment::Packed)
            .to_string();

        println!("{}", result_str);
    }

    #[test]
    fn read_and_write_lazy() {
        let str = r#"a 	U+0061    61            0061        "Latin Small Letter A"
~ 	U+007E    7E            007E        Tilde
¥ 	U+00A5    C2_A5         00A5        "Yen Sign"
» 	U+00BB    C2_BB         00BB        "Right-Pointing Double Angle Quotation Mark"
½ 	U+00BD    C2_BD         00BD        "Vulgar Fraction One Half"
¿ 	U+00BF    C2_BF         00BF        "Inverted#Question Mark" # This is a comment
ß 	U+00DF    C3_9F         00DF        "Latin Small Letter Sharp S"
ä 	U+00E4    C3_A4         00E4        "Latin Small Letter A with Diaeresis"
ï 	U+00EF    C3_AF         00EF        "Latin Small Letter I with Diaeresis"
œ 	U+0153    C5_93         0153        "Latin Small Ligature Oe"
€ 	U+20AC    E2_82_AC      20AC        "Euro Sign"
東 	U+6771    E6_9D_B1      6771        "CJK Unified Ideograph-6771"
𝄞 	U+1D11E   F0_9D_84_9E   D834_DD1E   "Musical Symbol G Clef"
𠀇 	U+20007   F0_A0_80_87   D840_DC07   "CJK Unified Ideograph-20007"
-   hyphen    qwro-qweb     -dasbe      "A hyphen character - represents null""#;
        let result = parse_lazy(str.chars());

        let result = result.map(|line| {
            line.unwrap().into_iter().map(|value| {
                let mut prefix = "-".to_string();
                prefix.push_str(&value.unwrap_or("-".to_string()));
                Some(prefix)
            })
        });

        let result_str = WSVWriter::new(result)
            .align_columns(super::ColumnAlignment::Packed)
            .to_string();

        println!("{}", result_str);
    }

    #[test]
    fn e2e_test() {
        let str = include_str!("../tests/1_stenway.com");
        let result = parse(str);

        let assert_matches_expected =
            |result: Result<Vec<Vec<Option<Cow<'_, str>>>>, WSVError>| match result {
                Err(_) => panic!("Should not have error"),
                Ok(values) => {
                    let expected = vec![
                        vec![
                            "a",
                            "U+0061",
                            "61",
                            "0061",
                            "Latin Small Letter A",
                            "\n\"\"",
                        ],
                        vec!["~", "U+007E", "7E", "007E", "Tilde"],
                        vec!["¥", "U+00A5", "C2_A5", "00A5", "Yen Sign"],
                        vec![
                            "»",
                            "U+00BB",
                            "C2_BB",
                            "00BB",
                            "Right-Pointing Double Angle Quotation Mark",
                        ],
                        vec!["½", "U+00BD", "C2_BD", "00BD", "Vulgar Fraction One Half"],
                        vec!["¿", "U+00BF", "C2_BF", "00BF", "Inverted#Question Mark"],
                        vec!["ß", "U+00DF", "C3_9F", "00DF", "Latin Small Letter Sharp S"],
                        vec![
                            "ä",
                            "U+00E4",
                            "C3_A4",
                            "00E4",
                            "Latin Small Letter A with Diaeresis",
                        ],
                        vec![
                            "ï",
                            "U+00EF",
                            "C3_AF",
                            "00EF",
                            "Latin Small Letter I with Diaeresis",
                        ],
                        vec!["œ", "U+0153", "C5_93", "0153", "Latin Small Ligature Oe"],
                        vec!["€", "U+20AC", "E2_82_AC", "20AC", "Euro Sign"],
                        vec![
                            "東",
                            "U+6771",
                            "E6_9D_B1",
                            "6771",
                            "CJK Unified Ideograph-6771",
                        ],
                        vec![
                            "𝄞",
                            "U+1D11E",
                            "F0_9D_84_9E",
                            "D834_DD1E",
                            "Musical Symbol G Clef",
                        ],
                        vec![
                            "𠀇",
                            "U+20007",
                            "F0_A0_80_87",
                            "D840_DC07",
                            "CJK Unified Ideograph-20007",
                        ],
                        vec![
                            "-",
                            "hyphen",
                            "qwro-qweb",
                            "-dasbe",
                            "A hyphen character - represents null",
                        ],
                    ];

                    let mut expected_iter = expected.into_iter();
                    let mut acutal_iter = values.into_iter();

                    loop {
                        let expected_line = expected_iter.next();
                        let actual_line = acutal_iter.next();

                        assert_eq!(
                            expected_line.is_some(),
                            actual_line.is_some(),
                            "Line numbers should match"
                        );
                        if expected_line.is_none() || actual_line.is_none() {
                            break;
                        }

                        let mut expected_value_iter = expected_line.unwrap().into_iter();
                        let mut actual_value_iter = actual_line.unwrap().into_iter();
                        loop {
                            let expected_value = expected_value_iter.next();
                            let actual_value = actual_value_iter.next();

                            assert_eq!(
                                expected_value.is_some(),
                                expected_value.is_some(),
                                "Value counts should match"
                            );
                            if expected_value.is_none() || actual_value.is_none() {
                                break;
                            }

                            if expected_value.unwrap() == "-" {
                                assert_eq!(None, actual_value.unwrap(), "'-' should parse to None");
                            } else {
                                let actual_value = actual_value
                                .expect("Actual value to be populated at this poitn.")
                                .expect(
                                    "actual value should parse to Some() if expected is not '-'",
                                );
                                let expected = expected_value.as_ref().unwrap();
                                let actual = actual_value.as_ref();
                                if expected_value.unwrap().to_owned() != actual_value.to_owned() {
                                    println!("Mismatch: \nExpected: {expected}\nActual: {actual}");
                                    panic!();
                                }
                            }
                        }
                    }
                }
            };

        assert_matches_expected(result);

        let parsed = parse(str).unwrap();
        let written = WSVWriter::new(parsed).to_string();
        println!("Writer output: {}", written);
        let reparsed = parse(&written);
        println!("Reparsed: {:?}", reparsed);
        assert_matches_expected(reparsed);
    }

    #[test]
    fn e2e_test_lazy() {
        let str = include_str!("../tests/1_stenway.com");
        let result = parse_lazy(str.chars())
            .map(|line| line.unwrap())
            .collect::<Vec<_>>();

        let assert_matches_expected = |values: Vec<Vec<Option<String>>>| {
            let expected = vec![
                vec![
                    "a",
                    "U+0061",
                    "61",
                    "0061",
                    "Latin Small Letter A",
                    "\n\"\"",
                ],
                vec!["~", "U+007E", "7E", "007E", "Tilde"],
                vec!["¥", "U+00A5", "C2_A5", "00A5", "Yen Sign"],
                vec![
                    "»",
                    "U+00BB",
                    "C2_BB",
                    "00BB",
                    "Right-Pointing Double Angle Quotation Mark",
                ],
                vec!["½", "U+00BD", "C2_BD", "00BD", "Vulgar Fraction One Half"],
                vec!["¿", "U+00BF", "C2_BF", "00BF", "Inverted#Question Mark"],
                vec!["ß", "U+00DF", "C3_9F", "00DF", "Latin Small Letter Sharp S"],
                vec![
                    "ä",
                    "U+00E4",
                    "C3_A4",
                    "00E4",
                    "Latin Small Letter A with Diaeresis",
                ],
                vec![
                    "ï",
                    "U+00EF",
                    "C3_AF",
                    "00EF",
                    "Latin Small Letter I with Diaeresis",
                ],
                vec!["œ", "U+0153", "C5_93", "0153", "Latin Small Ligature Oe"],
                vec!["€", "U+20AC", "E2_82_AC", "20AC", "Euro Sign"],
                vec![
                    "東",
                    "U+6771",
                    "E6_9D_B1",
                    "6771",
                    "CJK Unified Ideograph-6771",
                ],
                vec![
                    "𝄞",
                    "U+1D11E",
                    "F0_9D_84_9E",
                    "D834_DD1E",
                    "Musical Symbol G Clef",
                ],
                vec![
                    "𠀇",
                    "U+20007",
                    "F0_A0_80_87",
                    "D840_DC07",
                    "CJK Unified Ideograph-20007",
                ],
                vec![
                    "-",
                    "hyphen",
                    "qwro-qweb",
                    "-dasbe",
                    "A hyphen character - represents null",
                ],
            ];

            let mut expected_iter = expected.into_iter();
            let mut acutal_iter = values.into_iter();

            loop {
                let expected_line = expected_iter.next();
                let actual_line = acutal_iter.next();

                assert_eq!(
                    expected_line.is_some(),
                    actual_line.is_some(),
                    "Line numbers should match"
                );
                if expected_line.is_none() || actual_line.is_none() {
                    break;
                }

                let mut expected_value_iter = expected_line.unwrap().into_iter();
                let mut actual_value_iter = actual_line.unwrap().into_iter();
                loop {
                    let expected_value = expected_value_iter.next();
                    let actual_value = actual_value_iter.next();

                    assert_eq!(
                        expected_value.is_some(),
                        expected_value.is_some(),
                        "Value counts should match"
                    );
                    if expected_value.is_none() || actual_value.is_none() {
                        break;
                    }

                    if expected_value.unwrap() == "-" {
                        assert_eq!(None, actual_value.unwrap(), "'-' should parse to None");
                    } else {
                        let actual_value = actual_value
                            .expect("Actual value to be populated at this poitn.")
                            .expect("actual value should parse to Some() if expected is not '-'");
                        assert_eq!(
                            expected_value.unwrap().to_owned(),
                            actual_value.to_owned(),
                            "string values should match"
                        );
                    }
                }
            }
        };

        assert_matches_expected(result);

        let parsed = parse(str).unwrap();
        let written = WSVWriter::new(parsed).to_string();
        let reparsed = parse_lazy(written.chars())
            .map(|line| line.unwrap())
            .collect();
        assert_matches_expected(reparsed);
    }

    #[test]
    fn readme_example_write() {
        use std::fs::File;
        use std::io::BufReader;
        // I recommend you pull in the utf8-chars crate as a dependency if
        // you need lazy parsing
        use crate::{parse_lazy, WSVWriter};
        use utf8_chars::BufReadCharsExt;

        let mut reader = BufReader::new(File::open("./my_very_large_file.txt").unwrap());

        let chars = reader.chars().map(|ch| ch.unwrap());

        let lines_lazy = parse_lazy(chars).map(|line| {
            // For this example we will assume we have valid WSV
            let sum = line
                .unwrap()
                .into_iter()
                // We're counting None as 0 in my case,
                // so flat_map the Nones out.
                .flat_map(|opt| opt)
                .map(|value| value.parse::<i32>().unwrap_or(0))
                .sum::<i32>();

            // The writer needs a 2D iterator of Option<String>,
            // so wrap the value in a Some and .to_string() it.
            // Also wrap in a Vec to make it a 2D iterator
            vec![Some(sum.to_string())]
        });
        // CAREFUL: Don't call .collect() here or we'll run out of memory!

        // The WSVWriter when using ColumnAlignment::Packed
        // (the default) is also lazy, so we can pass our
        // result in directly.
        for ch in WSVWriter::new(lines_lazy) {
            // Your code to dump the output to a file goes here.
            print!("{}", ch);
        }
    }

    #[test]
    fn in_and_out_with_cows() {
        let str = include_str!("../tests/1_stenway.com");

        let values = parse(str).unwrap_or_else(|err| panic!("{:?}", err));
        let output = WSVWriter::new(values)
            .align_columns(crate::ColumnAlignment::Right)
            .to_string();

        println!("{}", output);
    }

    #[test]
    fn writing_strings() {
        let values = vec![vec![None, Some("test".to_string())]];

        let output = WSVWriter::new(values)
            .align_columns(crate::ColumnAlignment::Packed)
            .to_string();

        println!("{}", output);
    }

    #[test]
    fn tokenizes_strings_correctly() {
        let input = "\"this is a string\"";
        let mut tokenizer = WSVTokenizer::new(input);
        assert!(are_equal(
            Ok(WSVToken::Value(Cow::Borrowed("this is a string"))),
            tokenizer.next().unwrap()
        ));
        assert!(tokenizer.next().is_none());
    }

    #[test]
    fn tokenizes_string_and_immediate_comment_correctly() {
        let input = "somekindofvalue#thenacomment";
        let mut tokenizer = WSVTokenizer::new(input);
        assert!(are_equal(
            Ok(WSVToken::Value(Cow::Borrowed("somekindofvalue"))),
            tokenizer.next().unwrap()
        ));
        assert!(are_equal(
            Ok(WSVToken::Comment("thenacomment")),
            tokenizer.next().unwrap()
        ));
    }

    #[test]
    fn tokenizes_string_and_immediate_comment_correctly_lazily() {
        let input = "somekindofvalue#thenacomment";
        let mut tokenizer = WSVLazyTokenizer::new(input.chars());
        assert!(owned_are_equal(
            Ok(OwnedWSVToken::Value("somekindofvalue".to_string())),
            tokenizer.next().unwrap()
        ));
        assert!(owned_are_equal(
            Ok(OwnedWSVToken::Comment("thenacomment".to_string())),
            tokenizer.next().unwrap()
        ));
    }

    #[test]
    fn catches_invalid_line_breaks() {
        let input = "\"this is a string with an invalid \"/ line break.\"";
        let mut tokenizer = WSVTokenizer::new(input);
        if let Err(err) = tokenizer.next().unwrap() {
            if let WSVErrorType::InvalidStringLineBreak = err.err_type() {
                assert!(tokenizer.next().is_none());
                return;
            }
        }
        panic!("Expected to find an InvalidStringLineBreak error");
    }

    #[test]
    fn doesnt_err_on_false_positive_line_breaks() {
        let input = "\"string \"\"/\"";
        let mut tokenizer = WSVTokenizer::new(input);
        let token = tokenizer.next().unwrap();
        assert!(are_equal(
            Ok(WSVToken::Value(Cow::Owned("string \"/".to_string()))),
            token
        ));
        assert!(tokenizer.next().is_none());
    }

    #[test]
    fn escapes_quotes_correctly() {
        let input = "\"\"\"\"\"\"\"\"";
        let mut tokenizer = WSVTokenizer::new(input);
        assert!(are_equal(
            Ok(WSVToken::Value(Cow::Owned("\"\"\"".to_string()))),
            tokenizer.next().unwrap()
        ));
        assert!(tokenizer.next().is_none());
    }

    #[test]
    fn escapes_new_lines_correctly() {
        let input = "\"\"/\"\"/\"\"/\"\"";
        let mut tokenizer = WSVTokenizer::new(input);
        let token = tokenizer.next().unwrap();
        println!("{:?}", token);
        assert!(are_equal(
            Ok(WSVToken::Value(Cow::Owned("\n\n\n".to_string()))),
            token
        ));
    }

    #[test]
    fn parses_quoted_string_and_immediate_comment_correctly() {
        let input = "\"somekindofvalue\"#thenacomment";
        let mut tokenizer = WSVTokenizer::new(input);
        assert!(are_equal(
            Ok(WSVToken::Value(Cow::Borrowed("somekindofvalue"))),
            tokenizer.next().unwrap()
        ));
        assert!(are_equal(
            Ok(WSVToken::Comment("thenacomment")),
            tokenizer.next().unwrap()
        ));
    }

    #[test]
    fn catches_unclosed_string() {
        let input = "\"this is an unclosed string";
        let mut tokenizer = WSVTokenizer::new(input);
        assert!(are_equal(
            Err(WSVError {
                location: crate::Location::default(),
                err_type: WSVErrorType::StringNotClosed
            }),
            tokenizer.next().unwrap()
        ));
        assert!(tokenizer.next().is_none());
    }

    #[test]
    fn atrocious_wsv() {
        let result = parse(include_str!("../tests/my_test.txt"));
        println!("{:?}", result.unwrap());
    }

    #[allow(dead_code)]
    fn are_equal(first: Result<WSVToken, WSVError>, second: Result<WSVToken, WSVError>) -> bool {
        match first {
            Ok(WSVToken::LF) => {
                if let Ok(WSVToken::LF) = second {
                    return true;
                } else {
                    return false;
                }
            }
            Ok(WSVToken::Null) => {
                if let Ok(WSVToken::Null) = second {
                    return true;
                } else {
                    return false;
                }
            }
            Ok(WSVToken::Comment(str1)) => {
                if let Ok(WSVToken::Comment(str2)) = second {
                    return str1 == str2;
                } else {
                    return false;
                }
            }
            Ok(WSVToken::Value(value1)) => {
                if let Ok(WSVToken::Value(value2)) = second {
                    return value1.as_ref() == value2.as_ref();
                } else {
                    return false;
                }
            }
            Err(err1) => {
                if let Err(err2) = second {
                    return err1.err_type() == err2.err_type();
                } else {
                    return false;
                }
            }
        }
    }

    #[allow(dead_code)]
    fn owned_are_equal(
        first: Result<OwnedWSVToken, WSVError>,
        second: Result<OwnedWSVToken, WSVError>,
    ) -> bool {
        match first {
            Ok(OwnedWSVToken::LF) => {
                if let Ok(OwnedWSVToken::LF) = second {
                    return true;
                } else {
                    return false;
                }
            }
            Ok(OwnedWSVToken::Null) => {
                if let Ok(OwnedWSVToken::Null) = second {
                    return true;
                } else {
                    return false;
                }
            }
            Ok(OwnedWSVToken::Comment(str1)) => {
                if let Ok(OwnedWSVToken::Comment(str2)) = second {
                    return str1 == str2;
                } else {
                    return false;
                }
            }
            Ok(OwnedWSVToken::Value(value1)) => {
                if let Ok(OwnedWSVToken::Value(value2)) = second {
                    return value1 == value2;
                } else {
                    return false;
                }
            }
            Err(err1) => {
                if let Err(err2) = second {
                    return err1.err_type() == err2.err_type();
                } else {
                    return false;
                }
            }
        }
    }

    #[test]
    fn write_really_large_file() {
        let values = (0..u32::MAX).map(|_| (0..10).into_iter().map(|val| Some(val.to_string())));
        for ch in WSVWriter::new(values) {
            print!("{}", ch);
            // This is so my computer doesn't fry when running unit tests.
            break;
        }
    }

    #[test]
    fn lazy_parse_write_example() {
        use crate::{parse_lazy, WSVWriter};

        // pretend that this input is some iterator over
        // all the characters in a 300 Gigabyte file.
        let input = String::new();
        let chars = input.chars();

        let lines = parse_lazy(chars).map(|line| {
            // You probably want to handle errors in your case
            // unless you are guaranteed to have valid WSV.
            let sum = line
                .unwrap()
                .into_iter()
                // We're counting None as 0, so flat_map them out.
                .flat_map(|opt| opt)
                .map(|value| value.parse::<i32>().unwrap_or(0))
                .sum::<i32>();

            vec![Some(sum.to_string())]
        });

        for ch in WSVWriter::new(lines) {
            // Your code to dump the output to a file goes here.
            print!("{}", ch)
        }
    }

    #[test]
    fn error_location_reporting_is_correct() {
        let input = r#"some values would go here
        and this is a second line,
        but the realy error happens
"here where the string is unclosed.
"#;

        for result in WSVLazyTokenizer::new(input.chars()) {
            match result {
                Ok(_) => {}
                Err(err) => {
                    assert_eq!(4, err.location().line());
                    assert_eq!(36, err.location().col());
                }
            }
        }
    }

    #[test]
    fn jagged_array_no_panic() {
        super::WSVWriter::new([vec![Some("1")], vec![Some("3"), None]])
            .align_columns(super::ColumnAlignment::Left)
            .to_string();
    }
}