Curator/src/core/legacy_meta.rs

//! Readers for Curator's *pre-index* metadata layouts.
//!
//! Before the unified pool index, Tagger / early Curator stored a movie's
//! metadata in a per-file sidecar `.{filename}.!tag` (JSON) next to the video.
//! When importing a file from such a pool we can recover its ČSFD link (and
//! title / year) from that sidecar — or from an old `.Curator.!index` found by
//! walking up from the file. Read-only recovery for the import dialog.

use std::fs;
use std::path::Path;

use serde_json::Value;

use super::paths;
use super::pool_index::INDEX_FILENAME;

pub const SIDECAR_SUFFIX: &str = ".!tag";
const MAX_INDEX_WALK_UP: usize = 4;

/// What an old sidecar / index record can still tell us about a movie.
#[derive(Debug, Clone, Default, PartialEq)]
pub struct LegacyMeta {
    pub csfd_url: Option<String>,
    pub title: Option<String>,
    pub year: Option<String>,
    /// "sidecar" | "index".
    pub source: String,
}

fn nonempty_str(v: Option<&Value>) -> Option<String> {
    v.and_then(Value::as_str)
        .filter(|s| !s.is_empty())
        .map(String::from)
}

/// Pull the useful fields out of a stored movie record (any schema version).
fn record_to_meta(record: &Value, source: &str) -> Option<LegacyMeta> {
    let rec = record.as_object()?;
    let csfd = nonempty_str(rec.get("csfd_url")).or_else(|| nonempty_str(rec.get("csfd_link")));
    let title = nonempty_str(rec.get("title"));
    let from = |key: &str| {
        rec.get(key)
            .and_then(Value::as_object)
            .and_then(|o| o.get("year"))
            .cloned()
    };
    let raw_year = from("meta")
        .filter(super::pool_index::truthy)
        .or_else(|| from("csfd_cache"))
        .filter(|v| !v.is_null());
    let year = raw_year
        .map(|v| {
            v.as_str()
                .map_or_else(|| v.to_string(), String::from)
                .trim()
                .to_string()
        })
        .filter(|y| !y.is_empty() && y.chars().all(|c| c.is_ascii_digit()));
    if csfd.is_none() && title.is_none() && year.is_none() {
        return None;
    }
    Some(LegacyMeta {
        csfd_url: csfd,
        title,
        year,
        source: source.into(),
    })
}

fn read_json(path: &Path) -> Option<Value> {
    serde_json::from_str(&fs::read_to_string(path).ok()?).ok()
}

fn from_sidecar(video_path: &Path) -> Option<LegacyMeta> {
    let sidecar = video_path
        .parent()?
        .join(format!(".{}{SIDECAR_SUFFIX}", paths::name(video_path)));
    if !sidecar.is_file() {
        return None;
    }
    record_to_meta(&read_json(&sidecar)?, "sidecar")
}

/// The record for `video_path` in an index's `movies` map — exact pool-relative
/// key first, then a unique basename match.
fn lookup_in_index<'a>(
    movies: &'a serde_json::Map<String, Value>,
    video_path: &Path,
    index_dir: &Path,
) -> Option<&'a Value> {
    if video_path.starts_with(index_dir)
        && let Some(rec) = movies.get(&paths::posix_relative(video_path, index_dir))
    {
        return Some(rec);
    }
    let name = paths::name(video_path);
    let matches: Vec<&Value> = movies
        .iter()
        .filter(|(k, _)| k.rsplit('/').next() == Some(name.as_str()))
        .map(|(_, v)| v)
        .collect();
    (matches.len() == 1).then(|| matches[0])
}

fn from_legacy_index(video_path: &Path) -> Option<LegacyMeta> {
    let mut directory = video_path.parent()?.to_path_buf();
    for _ in 0..=MAX_INDEX_WALK_UP {
        let index_path = directory.join(INDEX_FILENAME);
        if index_path.is_file() {
            let data = read_json(&index_path)?;
            let obj = data.as_object()?;
            let movies = match obj.get("movies") {
                Some(m) => m.as_object()?,
                None => obj,
            };
            return record_to_meta(lookup_in_index(movies, video_path, &directory)?, "index");
        }
        match directory.parent() {
            Some(parent) if parent != directory => directory = parent.to_path_buf(),
            _ => break,
        }
    }
    None
}

/// Recover ČSFD link / title / year from a pre-index layout: a `.{name}.!tag`
/// sidecar next to the video, then a `.Curator.!index` up to 4 levels up.
pub fn read_legacy_meta(video_path: &Path) -> Option<LegacyMeta> {
    from_sidecar(video_path).or_else(|| from_legacy_index(video_path))
}

#[cfg(test)]
mod tests {
    use super::*;
    use serde_json::json;

    fn video(dir: &Path, rel: &str) -> std::path::PathBuf {
        let p = dir.join(rel);
        fs::create_dir_all(p.parent().unwrap()).unwrap();
        fs::write(&p, "x").unwrap();
        p
    }

    #[test]
    fn test_reads_tagger_era_sidecar() {
        let dir = tempfile::tempdir().unwrap();
        let v = video(dir.path(), "Matrix.mkv");
        fs::write(dir.path().join(".Matrix.mkv.!tag"),
            json!({"csfd_link": "https://www.csfd.cz/film/9499/", "title": "Matrix", "csfd_cache": {"year": 1999}}).to_string()).unwrap();
        let meta = read_legacy_meta(&v).unwrap();
        assert_eq!(
            meta.csfd_url.as_deref(),
            Some("https://www.csfd.cz/film/9499/")
        );
        assert_eq!(meta.year.as_deref(), Some("1999"));
        assert_eq!(meta.source, "sidecar");
    }

    #[test]
    fn test_reads_old_index_by_key_and_by_basename() {
        let dir = tempfile::tempdir().unwrap();
        let v = video(dir.path(), "Filmy/sub/Matrix.mkv");
        fs::write(dir.path().join(INDEX_FILENAME),
            json!({"movies": {"Filmy/sub/Matrix.mkv": {"csfd_url": "u1"}, "Old/Other.mkv": {"title": "Other"}}}).to_string()).unwrap();
        assert_eq!(
            read_legacy_meta(&v).unwrap().csfd_url.as_deref(),
            Some("u1")
        );
        let moved = video(dir.path(), "Elsewhere/Other.mkv");
        assert_eq!(
            read_legacy_meta(&moved).unwrap().title.as_deref(),
            Some("Other")
        );
    }

    #[test]
    fn test_sidecar_wins_over_index() {
        let dir = tempfile::tempdir().unwrap();
        let v = video(dir.path(), "Matrix.mkv");
        fs::write(
            dir.path().join(".Matrix.mkv.!tag"),
            json!({"csfd_url": "side"}).to_string(),
        )
        .unwrap();
        fs::write(
            dir.path().join(INDEX_FILENAME),
            json!({"movies": {"Matrix.mkv": {"csfd_url": "idx"}}}).to_string(),
        )
        .unwrap();
        assert_eq!(
            read_legacy_meta(&v).unwrap().csfd_url.as_deref(),
            Some("side")
        );
    }

    #[test]
    fn test_none_cases() {
        let dir = tempfile::tempdir().unwrap();
        let v = video(dir.path(), "Matrix.mkv");
        assert!(read_legacy_meta(&v).is_none());
        fs::write(dir.path().join(".Matrix.mkv.!tag"), "{broken").unwrap();
        assert!(read_legacy_meta(&v).is_none());
        fs::write(
            dir.path().join(".Matrix.mkv.!tag"),
            json!({"tags": []}).to_string(),
        )
        .unwrap();
        assert!(read_legacy_meta(&v).is_none());
    }
}