poexam 0.0.10

Blazingly fast PO linter.
// SPDX-FileCopyrightText: 2026 Sébastien Helleu <flashcode@flashtux.org>
//
// SPDX-License-Identifier: GPL-3.0-or-later

//! Implementation of the `double-quotes` rule: check missing/extra double quotes.

use crate::checker::Checker;
use crate::diagnostic::{Diagnostic, Severity};
use crate::po::entry::Entry;
use crate::po::message::Message;
use crate::rules::rule::RuleChecker;

const DOUBLE_QUOTES: [char; 8] = [
    '"',  // U+0022: quotation mark
    '«',  // U+00AB: left pointing double angle quotation mark
    '»',  // U+00BB: right pointing double angle quotation mark
    '',  // U+201C: left double quotation mark
    '',  // U+201D: right double quotation mark
    '',  // U+201E: double low quotation mark
    '',  // U+201F: double high-reversed-9 quotation mark
    '', // U+FF02: fullwidth quotation mark
];

pub struct DoubleQuotesRule;

impl RuleChecker for DoubleQuotesRule {
    fn name(&self) -> &'static str {
        "double-quotes"
    }

    fn description(&self) -> &'static str {
        "Check for missing or extra double quotes in translation."
    }

    fn is_default(&self) -> bool {
        true
    }

    fn is_check(&self) -> bool {
        true
    }

    /// Check for missing or extra double quotes in the translation.
    ///
    /// The following quotes are considered:
    /// - quotation mark: '"' (U+0022)
    /// - left pointing double angle quotation mark: '«' (U+00AB)
    /// - right pointing double angle quotation mark: '»' (U+00BB)
    /// - left double quotation mark: '“' (U+201C)
    /// - right double quotation mark: '”' (U+201D)
    /// - double low quotation mark: '„' (U+201E)
    /// - double high-reversed-9 quotation mark: '‟' (U+201F)
    /// - fullwidth quotation mark: '"' (U+FF02)
    ///
    /// Wrong entry:
    /// ```text
    /// msgid "this is a \"test\""
    /// msgstr "ceci est un test"
    /// ```
    ///
    /// Correct entry:
    /// ```text
    /// msgid "this is a \"test\""
    /// msgstr "ceci est un \"test\""
    /// ```
    ///
    /// Diagnostics reported:
    /// - [`info`](Severity::Info): `missing double quotes (# / #)`
    /// - [`info`](Severity::Info): `extra double quotes (# / #)`
    fn check_msg(
        &self,
        checker: &Checker,
        _entry: &Entry,
        msgid: &Message,
        msgstr: &Message,
    ) -> Vec<Diagnostic> {
        let id_count = msgid.value.matches(DOUBLE_QUOTES).count();
        let str_count = msgstr.value.matches(DOUBLE_QUOTES).count();
        let msg = match id_count.cmp(&str_count) {
            std::cmp::Ordering::Equal => return vec![],
            std::cmp::Ordering::Greater => {
                format!("missing double quotes ({id_count} / {str_count})")
            }
            std::cmp::Ordering::Less => {
                format!("extra double quotes ({id_count} / {str_count})")
            }
        };
        self.new_diag(checker, Severity::Info, msg)
            .map(|d| {
                d.with_msgs_hl(
                    msgid,
                    msgid
                        .value
                        .match_indices(DOUBLE_QUOTES)
                        .map(|(idx, value)| (idx, idx + value.len())),
                    msgstr,
                    msgstr
                        .value
                        .match_indices(DOUBLE_QUOTES)
                        .map(|(idx, value)| (idx, idx + value.len())),
                )
            })
            .into_iter()
            .collect()
    }
}

/// Trim one pair of quotes from both sides of the email, if any.
///
/// The quote skipped at the beginning may be different from the quote at the end.
pub(crate) fn trim_quotes(s: &str) -> &str {
    if s.starts_with(DOUBLE_QUOTES) && s.ends_with(DOUBLE_QUOTES) {
        // Return the string without the first and last UTF-8 char.
        let start = s.chars().next().unwrap().len_utf8();
        let end = s.char_indices().next_back().unwrap().0;
        return &s[start..end];
    }
    s
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::{diagnostic::Diagnostic, rules::rule::Rules};

    fn check_double_quotes(content: &str) -> Vec<Diagnostic> {
        let mut checker = Checker::new(content.as_bytes());
        let rules = Rules::new(vec![Box::new(DoubleQuotesRule {})]);
        checker.do_all_checks(&rules);
        checker.diagnostics
    }

    #[test]
    fn test_no_double_quotes() {
        let diags = check_double_quotes(
            r#"
msgid "tested"
msgstr "testé"
"#,
        );
        assert!(diags.is_empty());
    }

    #[test]
    fn test_double_quotes_ok() {
        let diags = check_double_quotes(
            r#"
msgid "this is a \"test\""
msgstr "ceci est un « test »"
"#,
        );
        assert!(diags.is_empty());
    }

    #[test]
    fn test_double_quotes_error_noqa() {
        let diags = check_double_quotes(
            r#"
#, noqa:double-quotes
msgid "this is a \"test\""
msgstr "ceci est un test"
"#,
        );
        assert!(diags.is_empty());
    }

    #[test]
    fn test_double_quotes_error() {
        let diags = check_double_quotes(
            r#"
msgid "this is a \"test\""
msgstr "ceci est un test"

msgid "this is a test"
msgstr "ceci est un \"test\""
"#,
        );
        assert_eq!(diags.len(), 2);
        let diag = &diags[0];
        assert_eq!(diag.severity, Severity::Info);
        assert_eq!(diag.message, "missing double quotes (2 / 0)");
        let diag = &diags[1];
        assert_eq!(diag.severity, Severity::Info);
        assert_eq!(diag.message, "extra double quotes (0 / 2)");
    }
}