@owenlamont/ryl 0.4.1

package/src/rules/commas.rs

@@ -0,0 +1,372 @@
+use std::ops::Range;
+
+use saphyr_parser::{Event, Marker, Parser, Span, SpannedEventReceiver};
+
+use crate::config::YamlLintConfig;
+use crate::rules::span_utils::span_char_index_to_byte;
+
+pub const ID: &str = "commas";
+const TOO_MANY_BEFORE: &str = "too many spaces before comma";
+const TOO_FEW_AFTER: &str = "too few spaces after comma";
+const TOO_MANY_AFTER: &str = "too many spaces after comma";
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub struct Config {
+    max_spaces_before: i64,
+    min_spaces_after: i64,
+    max_spaces_after: i64,
+}
+
+impl Config {
+    const DEFAULT_MAX_BEFORE: i64 = 0;
+    const DEFAULT_MIN_AFTER: i64 = 1;
+    const DEFAULT_MAX_AFTER: i64 = 1;
+
+    #[must_use]
+    pub fn resolve(cfg: &YamlLintConfig) -> Self {
+        let max_spaces_before = cfg
+            .rule_option(ID, "max-spaces-before")
+            .and_then(saphyr::YamlOwned::as_integer)
+            .unwrap_or(Self::DEFAULT_MAX_BEFORE);
+        let min_spaces_after = cfg
+            .rule_option(ID, "min-spaces-after")
+            .and_then(saphyr::YamlOwned::as_integer)
+            .unwrap_or(Self::DEFAULT_MIN_AFTER);
+        let max_spaces_after = cfg
+            .rule_option(ID, "max-spaces-after")
+            .and_then(saphyr::YamlOwned::as_integer)
+            .unwrap_or(Self::DEFAULT_MAX_AFTER);
+
+        Self {
+            max_spaces_before,
+            min_spaces_after,
+            max_spaces_after,
+        }
+    }
+
+    #[must_use]
+    pub const fn new_for_tests(
+        max_spaces_before: i64,
+        min_spaces_after: i64,
+        max_spaces_after: i64,
+    ) -> Self {
+        Self {
+            max_spaces_before,
+            min_spaces_after,
+            max_spaces_after,
+        }
+    }
+
+    #[must_use]
+    pub const fn max_spaces_before(&self) -> i64 {
+        self.max_spaces_before
+    }
+
+    #[must_use]
+    pub const fn min_spaces_after(&self) -> i64 {
+        self.min_spaces_after
+    }
+
+    #[must_use]
+    pub const fn max_spaces_after(&self) -> i64 {
+        self.max_spaces_after
+    }
+}
+
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct Violation {
+    pub line: usize,
+    pub column: usize,
+    pub message: String,
+}
+
+enum FlowKind {
+    Sequence,
+    Mapping,
+}
+
+struct ScalarRangeCollector {
+    ranges: Vec<Range<usize>>,
+}
+
+impl ScalarRangeCollector {
+    const fn new() -> Self {
+        Self { ranges: Vec::new() }
+    }
+
+    fn push_range(&mut self, span: Span) {
+        let start = span.start.index();
+        let end = span.end.index();
+        if start < end {
+            self.ranges.push(start..end);
+        }
+    }
+
+    fn into_sorted(self) -> Vec<Range<usize>> {
+        let mut ranges = self.ranges;
+        ranges.sort_by(|a, b| a.start.cmp(&b.start));
+        ranges
+    }
+}
+
+impl SpannedEventReceiver<'_> for ScalarRangeCollector {
+    fn on_event(&mut self, ev: Event<'_>, span: Span) {
+        if matches!(ev, Event::Scalar(..)) {
+            self.push_range(span);
+        }
+    }
+}
+
+enum BeforeResult {
+    SameLine { spaces: usize },
+    Ignored,
+}
+
+enum AfterResult {
+    SameLine { spaces: usize, next_char: usize },
+    Ignored,
+}
+
+#[must_use]
+pub fn check(buffer: &str, cfg: &Config) -> Vec<Violation> {
+    if buffer.is_empty() {
+        return Vec::new();
+    }
+
+    let mut parser = Parser::new_from_str(buffer);
+    let mut collector = ScalarRangeCollector::new();
+    let _ = parser.load(&mut collector, true);
+    let scalar_ranges = collector.into_sorted();
+
+    let chars: Vec<(usize, char)> = buffer.char_indices().collect();
+    let buffer_len = buffer.len();
+
+    let line_starts = build_line_starts(buffer);
+
+    let mut violations = Vec::new();
+    let mut contexts: Vec<FlowKind> = Vec::new();
+    let mut i = 0usize;
+    let mut range_idx = 0usize;
+
+    while i < chars.len() {
+        let (byte_idx, ch) = chars[i];
+
+        while range_idx < scalar_ranges.len()
+            && span_char_index_to_byte(&chars, scalar_ranges[range_idx].end, buffer_len)
+                <= byte_idx
+        {
+            range_idx += 1;
+        }
+
+        if let Some(range) = scalar_ranges.get(range_idx) {
+            let start_byte = span_char_index_to_byte(&chars, range.start, buffer_len);
+            let end_byte = span_char_index_to_byte(&chars, range.end, buffer_len);
+            if byte_idx >= start_byte && byte_idx < end_byte {
+                i = range.end;
+                continue;
+            }
+        }
+
+        match ch {
+            '[' => contexts.push(FlowKind::Sequence),
+            '{' => contexts.push(FlowKind::Mapping),
+            ']' | '}' => {
+                contexts.pop();
+            }
+            '#' => {
+                i = skip_comment(&chars, i);
+                continue;
+            }
+            ',' => {
+                if !contexts.is_empty() {
+                    evaluate_comma(cfg, &mut violations, &chars, i, &line_starts);
+                }
+            }
+            _ => {}
+        }
+
+        i += 1;
+    }
+
+    violations
+}
+
+fn skip_comment(chars: &[(usize, char)], mut idx: usize) -> usize {
+    idx += 1;
+    while idx < chars.len() {
+        let ch = chars[idx].1;
+        if ch == '\n' {
+            break;
+        }
+        if ch == '\r' {
+            if idx + 1 < chars.len() && chars[idx + 1].1 == '\n' {
+                idx += 1;
+            }
+            break;
+        }
+        idx += 1;
+    }
+    idx
+}
+
+fn evaluate_comma(
+    cfg: &Config,
+    violations: &mut Vec<Violation>,
+    chars: &[(usize, char)],
+    comma_idx: usize,
+    line_starts: &[usize],
+) {
+    if let BeforeResult::SameLine { spaces } = compute_spaces_before(chars, comma_idx)
+        && cfg.max_spaces_before >= 0
+    {
+        let spaces_i64 = i64::try_from(spaces).unwrap_or(i64::MAX);
+        if spaces_i64 > cfg.max_spaces_before {
+            let comma_byte = chars[comma_idx].0;
+            let (line, column) = line_and_column(line_starts, comma_byte);
+            let highlight_column = column.saturating_sub(1).max(1);
+            violations.push(Violation {
+                line,
+                column: highlight_column,
+                message: TOO_MANY_BEFORE.to_string(),
+            });
+        }
+    }
+
+    if let AfterResult::SameLine { spaces, next_char } =
+        compute_spaces_after(chars, comma_idx)
+    {
+        let spaces_i64 = i64::try_from(spaces).unwrap_or(i64::MAX);
+        let next_byte = chars[next_char].0;
+        let (line, column) = line_and_column(line_starts, next_byte);
+        if cfg.max_spaces_after >= 0 && spaces_i64 > cfg.max_spaces_after {
+            let highlight_column = column.saturating_sub(1).max(1);
+            violations.push(Violation {
+                line,
+                column: highlight_column,
+                message: TOO_MANY_AFTER.to_string(),
+            });
+        }
+        if cfg.min_spaces_after >= 0 && spaces_i64 < cfg.min_spaces_after {
+            violations.push(Violation {
+                line,
+                column,
+                message: TOO_FEW_AFTER.to_string(),
+            });
+        }
+    }
+}
+
+fn compute_spaces_before(chars: &[(usize, char)], comma_idx: usize) -> BeforeResult {
+    let mut spaces = 0usize;
+    let mut idx = comma_idx;
+
+    loop {
+        let Some(prev) = idx.checked_sub(1) else {
+            return BeforeResult::SameLine { spaces };
+        };
+
+        let ch = chars[prev].1;
+        if matches!(ch, ' ' | '\t') {
+            spaces += 1;
+            idx = prev;
+            continue;
+        }
+        if matches!(ch, '\n' | '\r') {
+            return BeforeResult::Ignored;
+        }
+        return BeforeResult::SameLine { spaces };
+    }
+}
+
+fn compute_spaces_after(chars: &[(usize, char)], comma_idx: usize) -> AfterResult {
+    let mut spaces = 0usize;
+    let mut idx = comma_idx + 1;
+    while idx < chars.len() {
+        match chars[idx].1 {
+            ' ' | '\t' => {
+                spaces += 1;
+                idx += 1;
+            }
+            '\n' | '\r' | '#' => return AfterResult::Ignored,
+            _ => {
+                return AfterResult::SameLine {
+                    spaces,
+                    next_char: idx,
+                };
+            }
+        }
+    }
+    AfterResult::Ignored
+}
+
+fn build_line_starts(buffer: &str) -> Vec<usize> {
+    let mut starts = Vec::new();
+    starts.push(0);
+    let bytes = buffer.as_bytes();
+    let mut idx = 0usize;
+    while idx < bytes.len() {
+        match bytes[idx] {
+            b'\n' => {
+                starts.push(idx + 1);
+                idx += 1;
+            }
+            b'\r' => {
+                if idx + 1 < bytes.len() && bytes[idx + 1] == b'\n' {
+                    starts.push(idx + 2);
+                    idx += 2;
+                } else {
+                    starts.push(idx + 1);
+                    idx += 1;
+                }
+            }
+            _ => idx += 1,
+        }
+    }
+    starts
+}
+
+fn line_and_column(line_starts: &[usize], byte_idx: usize) -> (usize, usize) {
+    let mut left = 0usize;
+    let mut right = line_starts.len();
+    while left + 1 < right {
+        let mid = usize::midpoint(left, right);
+        if line_starts[mid] <= byte_idx {
+            left = mid;
+        } else {
+            right = mid;
+        }
+    }
+    let line_start = line_starts[left];
+    (left + 1, byte_idx - line_start + 1)
+}
+
+#[doc(hidden)]
+#[must_use]
+pub fn coverage_compute_spaces_before(buffer: &str, comma_idx: usize) -> Option<usize> {
+    let chars: Vec<(usize, char)> = buffer.char_indices().collect();
+    debug_assert!(comma_idx < chars.len());
+    match compute_spaces_before(&chars, comma_idx) {
+        BeforeResult::SameLine { spaces } => Some(spaces),
+        BeforeResult::Ignored => None,
+    }
+}
+
+#[doc(hidden)]
+#[must_use]
+pub fn coverage_skip_zero_length_span() -> usize {
+    let mut collector = ScalarRangeCollector::new();
+    collector.push_range(Span::empty(Marker::default()));
+    collector.into_sorted().len()
+}
+
+#[doc(hidden)]
+#[must_use]
+pub fn coverage_skip_comment_crlf() -> (usize, usize) {
+    let chars_crlf: Vec<(usize, char)> = "#\r\n".char_indices().collect();
+    let idx_crlf = skip_comment(&chars_crlf, 0);
+
+    let chars_cr: Vec<(usize, char)> = "#\r".char_indices().collect();
+    let idx_cr = skip_comment(&chars_cr, 0);
+
+    (idx_crlf, idx_cr)
+}

package/src/rules/comments.rs

@@ -0,0 +1,299 @@
+use saphyr::YamlOwned;
+
+use crate::config::YamlLintConfig;
+
+pub const ID: &str = "comments";
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
+pub struct Config {
+    require_starting_space: bool,
+    ignore_shebangs: bool,
+    min_spaces_from_content: Option<usize>,
+}
+
+impl Config {
+    #[must_use]
+    pub fn resolve(cfg: &YamlLintConfig) -> Self {
+        let require_starting_space = cfg
+            .rule_option(ID, "require-starting-space")
+            .and_then(YamlOwned::as_bool)
+            .unwrap_or(true);
+
+        let ignore_shebangs = cfg
+            .rule_option(ID, "ignore-shebangs")
+            .and_then(YamlOwned::as_bool)
+            .unwrap_or(true);
+
+        let min_spaces_value = cfg
+            .rule_option(ID, "min-spaces-from-content")
+            .and_then(YamlOwned::as_integer)
+            .unwrap_or(2);
+
+        let min_spaces_from_content = if min_spaces_value < 0 {
+            None
+        } else {
+            Some(usize::try_from(min_spaces_value).unwrap_or(usize::MAX))
+        };
+
+        Self {
+            require_starting_space,
+            ignore_shebangs,
+            min_spaces_from_content,
+        }
+    }
+
+    const fn require_starting_space(&self) -> bool {
+        self.require_starting_space
+    }
+
+    const fn ignore_shebangs(&self) -> bool {
+        self.ignore_shebangs
+    }
+
+    const fn min_spaces_from_content(&self) -> Option<usize> {
+        self.min_spaces_from_content
+    }
+}
+
+#[derive(Debug, Clone, PartialEq, Eq)]
+pub struct Violation {
+    pub line: usize,
+    pub column: usize,
+    pub message: String,
+}
+
+#[must_use]
+pub fn check(buffer: &str, cfg: &Config) -> Vec<Violation> {
+    let mut violations = Vec::new();
+    let mut quote_state = QuoteState::default();
+    let mut block_tracker = BlockScalarTracker::default();
+
+    for (line_idx, line) in buffer.lines().enumerate() {
+        let indent = leading_indent_width(line);
+        let content = &line[indent..];
+
+        if block_tracker.consume_line(indent, content) {
+            continue;
+        }
+
+        let Some(comment_start) = find_comment_start(line, &mut quote_state) else {
+            block_tracker.observe_indicator(indent, content);
+            continue;
+        };
+
+        if let Some(required) = cfg.min_spaces_from_content()
+            && is_inline_comment(line, comment_start)
+            && inline_spacing_width(line, comment_start) < required
+        {
+            violations.push(Violation {
+                line: line_idx + 1,
+                column: column_at(line, comment_start),
+                message: format!("too few spaces before comment: expected {required}"),
+            });
+        }
+
+        if !cfg.require_starting_space() {
+            continue;
+        }
+
+        let after_hash_idx = comment_start + skip_hashes(&line[comment_start..]);
+        if after_hash_idx >= line.len() {
+            continue;
+        }
+
+        let next_char = line[after_hash_idx..].chars().next().unwrap_or(' ');
+
+        if cfg.ignore_shebangs()
+            && line_idx == 0
+            && comment_start == 0
+            && next_char == '!'
+        {
+            continue;
+        }
+
+        if next_char != ' ' {
+            violations.push(Violation {
+                line: line_idx + 1,
+                column: column_at(line, after_hash_idx),
+                message: "missing starting space in comment".to_string(),
+            });
+        }
+
+        block_tracker.observe_indicator(indent, content);
+    }
+
+    violations
+}
+
+#[derive(Debug, Default)]
+struct BlockScalarTracker {
+    state: Option<BlockScalarState>,
+}
+
+#[derive(Debug)]
+struct BlockScalarState {
+    indicator_indent: usize,
+    content_indent: Option<usize>,
+}
+
+impl BlockScalarTracker {
+    fn consume_line(&mut self, indent: usize, content: &str) -> bool {
+        let Some(state) = self.state.as_mut() else {
+            return false;
+        };
+
+        if content.trim().is_empty() {
+            return true;
+        }
+
+        if let Some(content_indent) = state.content_indent {
+            if indent >= content_indent {
+                return true;
+            }
+
+            if indent <= state.indicator_indent {
+                self.state = None;
+                return false;
+            }
+
+            state.content_indent = Some(content_indent.min(indent));
+            return true;
+        }
+
+        if indent > state.indicator_indent {
+            state.content_indent = Some(indent);
+            return true;
+        }
+
+        self.state = None;
+        false
+    }
+
+    fn observe_indicator(&mut self, indent: usize, content: &str) {
+        let candidate = strip_trailing_comment_for_block(content).trim_end();
+        if is_block_scalar_indicator(candidate) {
+            self.state = Some(BlockScalarState {
+                indicator_indent: indent,
+                content_indent: None,
+            });
+        }
+    }
+}
+
+#[derive(Debug, Default, Clone, Copy)]
+struct QuoteState {
+    in_single: bool,
+    in_double: bool,
+    escaped: bool,
+}
+
+fn find_comment_start(line: &str, state: &mut QuoteState) -> Option<usize> {
+    for (idx, ch) in line.char_indices() {
+        if ch == '\\' && !state.in_single {
+            state.escaped = !state.escaped;
+            continue;
+        }
+
+        if state.escaped {
+            state.escaped = false;
+            continue;
+        }
+
+        match ch {
+            '\'' if !state.in_double => {
+                state.in_single = !state.in_single;
+            }
+            '"' if !state.in_single => {
+                state.in_double = !state.in_double;
+            }
+            '#' if !state.in_single && !state.in_double => {
+                if is_comment_position(line, idx) {
+                    return Some(idx);
+                }
+            }
+            _ => {}
+        }
+    }
+
+    state.escaped = false;
+    None
+}
+
+fn is_inline_comment(line: &str, comment_start: usize) -> bool {
+    !line[..comment_start].trim().is_empty()
+}
+
+fn inline_spacing_width(line: &str, comment_start: usize) -> usize {
+    line[..comment_start]
+        .chars()
+        .rev()
+        .take_while(|ch| ch.is_whitespace())
+        .count()
+}
+
+fn skip_hashes(slice: &str) -> usize {
+    slice
+        .chars()
+        .take_while(|ch| *ch == '#')
+        .map(char::len_utf8)
+        .sum()
+}
+
+fn column_at(line: &str, byte_idx: usize) -> usize {
+    line[..byte_idx].chars().count() + 1
+}
+
+fn leading_indent_width(line: &str) -> usize {
+    line.chars()
+        .take_while(|ch| matches!(ch, ' ' | '\t'))
+        .count()
+}
+
+fn is_comment_position(line: &str, idx: usize) -> bool {
+    line[..idx].chars().last().is_none_or(char::is_whitespace)
+}
+
+fn strip_trailing_comment_for_block(content: &str) -> &str {
+    let mut in_single = false;
+    let mut in_double = false;
+    let mut escaped = false;
+    for (idx, ch) in content.char_indices() {
+        if ch == '\\' && !in_single {
+            escaped = !escaped;
+            continue;
+        }
+
+        if escaped {
+            escaped = false;
+            continue;
+        }
+
+        match ch {
+            '\'' if !in_double => {
+                in_single = !in_single;
+            }
+            '"' if !in_single => {
+                in_double = !in_double;
+            }
+            '#' if !in_single && !in_double => {
+                return content[..idx].trim_end();
+            }
+            _ => {}
+        }
+    }
+    content.trim_end()
+}
+
+fn is_block_scalar_indicator(content: &str) -> bool {
+    if content.is_empty() {
+        return false;
+    }
+
+    let trimmed = content.trim_end_matches(|ch: char| ch.is_whitespace());
+    trimmed.ends_with("|-")
+        || trimmed.ends_with("|+")
+        || trimmed.ends_with('|')
+        || trimmed.ends_with(">-")
+        || trimmed.ends_with(">+")
+        || trimmed.ends_with('>')
+}