opendata-log 0.1.1

//! Key listing for log streams.
//!
//! This module provides efficient key enumeration via per-segment listing records.
//! When entries are appended, listing records track which keys are present in each
//! segment. This enables key discovery without scanning all log entries.
//!
//! # Components
//!
//! - [`ListingReader`]: Read-only access to listing entries
//! - [`ListingCache`]: In-memory cache for write-path delta building
//! - [`ListingDelta`]: Delta representing listing state changes
//! - [`LogKeyIterator`]: Iterator over keys from listing entries

use std::collections::BTreeSet;
use std::ops::Range;
use std::sync::Arc;

use bytes::Bytes;
use common::{Record, StorageRead};

use crate::error::{Error, Result};
use crate::model::SegmentId;
use crate::segment::SegmentDelta;
use crate::serde::{ListingEntryKey, ListingEntryValue};

/// A key returned from the listing iterator.
#[derive(Debug, Clone, PartialEq, Eq)]
pub struct LogKey {
    /// The user-provided key
    pub key: Bytes,
}

impl LogKey {
    /// Creates a new log key.
    pub fn new(key: Bytes) -> Self {
        Self { key }
    }
}

/// Iterator over keys from listing entries.
///
/// Iterates over distinct keys present in the specified segment range.
/// Keys are deduplicated and returned in lexicographic order.
pub struct LogKeyIterator {
    /// Iterator over keys.
    keys: std::collections::btree_set::IntoIter<Bytes>,
}

impl LogKeyIterator {
    /// Creates a new key iterator by scanning listing entries.
    ///
    /// This loads all keys from the segment range into memory for deduplication.
    /// The async API is preserved for future streaming implementations.
    pub(crate) async fn open(
        storage: Arc<dyn StorageRead>,
        segment_range: Range<SegmentId>,
    ) -> Result<Self> {
        // Empty range means no keys
        if segment_range.start >= segment_range.end {
            return Ok(Self {
                keys: BTreeSet::new().into_iter(),
            });
        }

        // Scan all listing entries in the segment range
        let scan_range = ListingEntryKey::scan_range(segment_range);
        let mut iter = storage
            .scan_iter(scan_range)
            .await
            .map_err(|e| Error::Storage(e.to_string()))?;

        // Collect keys into BTreeSet (deduplicates and sorts)
        let mut keys = BTreeSet::new();
        while let Some(record) = iter
            .next()
            .await
            .map_err(|e| Error::Storage(e.to_string()))?
        {
            let entry_key = ListingEntryKey::deserialize(&record.key)?;
            keys.insert(entry_key.key);
        }

        Ok(Self {
            keys: keys.into_iter(),
        })
    }

    /// Returns the next key, or `None` if exhausted.
    pub async fn next(&mut self) -> Result<Option<LogKey>> {
        Ok(self.keys.next().map(LogKey::new))
    }
}

/// Delta representing listing state changes.
///
/// Produced by [`ListingCache::build_delta`] and consumed by [`ListingCache::apply_delta`].
#[derive(Debug, Clone)]
pub(crate) struct ListingDelta {
    /// The segment ID these entries belong to.
    segment_id: SegmentId,
    /// Keys that are new to this segment.
    new_keys: BTreeSet<Bytes>,
}

/// In-memory cache of keys seen in the current segment.
///
/// Used to build deltas for listing entries during ingestion, avoiding
/// duplicate writes for the same key within a segment.
///
/// # Usage
///
/// ```ignore
/// let mut records = Vec::new();
///
/// // Build delta and add listing records
/// let listing_delta = listing_cache.build_delta(&seg_delta, &keys, &mut records);
///
/// // ... write records to storage ...
///
/// // Apply delta to cache
/// listing_cache.apply_delta(listing_delta);
/// ```
pub(crate) struct ListingCache {
    /// The segment ID this cache is tracking.
    current_segment_id: Option<SegmentId>,
    /// Keys seen in the current segment.
    keys: BTreeSet<Bytes>,
}

impl ListingCache {
    /// Creates a new empty cache.
    pub(crate) fn new() -> Self {
        Self {
            current_segment_id: None,
            keys: BTreeSet::new(),
        }
    }

    /// Builds a delta for the given keys.
    ///
    /// For each key that is new to the segment (not in cache and not seen
    /// earlier in this batch), adds:
    /// - The key to the returned delta
    /// - A listing entry record to the records vec
    ///
    /// Does NOT update the cache - call `apply_delta()` after the storage
    /// write succeeds.
    pub(crate) fn build_delta(
        &self,
        seg_delta: &SegmentDelta,
        keys: &[Bytes],
        records: &mut Vec<Record>,
    ) -> ListingDelta {
        let segment_id = seg_delta.segment().id();
        let mut new_keys = BTreeSet::new();
        let value = ListingEntryValue::new().serialize();

        for key in keys {
            // Skip if already seen in this batch or cached
            if new_keys.contains(key) || !self.is_new(segment_id, key) {
                continue;
            }

            // Add listing entry record
            let storage_key = ListingEntryKey::new(segment_id, key.clone()).serialize();
            records.push(Record::new(storage_key, value.clone()));
            new_keys.insert(key.clone());
        }

        ListingDelta {
            segment_id,
            new_keys,
        }
    }

    /// Applies a delta to the cache, recording all new keys as seen.
    ///
    /// Call this after the storage write succeeds to update the cache
    /// with the newly written keys.
    ///
    /// If the delta's segment differs from the cached segment, the cache
    /// is reset before applying.
    pub(crate) fn apply_delta(&mut self, delta: ListingDelta) {
        if self.current_segment_id != Some(delta.segment_id) {
            self.keys = delta.new_keys;
            self.current_segment_id = Some(delta.segment_id);
        } else {
            self.keys.extend(delta.new_keys);
        }
    }

    /// Checks if a key is new for the given segment.
    fn is_new(&self, segment_id: SegmentId, key: &Bytes) -> bool {
        if self.current_segment_id != Some(segment_id) {
            return true;
        }
        !self.keys.contains(key)
    }
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::segment::LogSegment;
    use crate::serde::SegmentMeta;
    use crate::storage::LogStorage;

    fn test_segment(id: u32) -> LogSegment {
        LogSegment::new(id, SegmentMeta::new(0, 0))
    }

    fn test_seg_delta(segment_id: u32) -> SegmentDelta {
        SegmentDelta::new(test_segment(segment_id), false)
    }

    mod log_key_iterator {
        use super::*;

        async fn write_listing_entry(storage: &LogStorage, segment_id: u32, key: &[u8]) {
            let storage_key =
                ListingEntryKey::new(segment_id, Bytes::copy_from_slice(key)).serialize();
            let value = ListingEntryValue::new().serialize();
            storage
                .put_with_options(
                    vec![common::Record::new(storage_key, value)],
                    common::WriteOptions::default(),
                )
                .await
                .unwrap();
        }

        #[tokio::test]
        async fn should_return_empty_for_empty_range() {
            // given
            let storage = LogStorage::in_memory();

            // when
            let mut iter = storage.as_read().list_keys(0..0).await.unwrap();

            // then
            assert!(iter.next().await.unwrap().is_none());
        }

        #[tokio::test]
        async fn should_return_empty_when_no_listing_entries() {
            // given
            let storage = LogStorage::in_memory();

            // when
            let mut iter = storage.as_read().list_keys(0..10).await.unwrap();

            // then
            assert!(iter.next().await.unwrap().is_none());
        }

        #[tokio::test]
        async fn should_iterate_keys_in_single_segment() {
            // given
            let storage = LogStorage::in_memory();
            write_listing_entry(&storage, 0, b"key-a").await;
            write_listing_entry(&storage, 0, b"key-b").await;
            write_listing_entry(&storage, 0, b"key-c").await;

            // when
            let mut iter = storage.as_read().list_keys(0..1).await.unwrap();

            // then - keys returned in lexicographic order
            let mut keys = Vec::new();
            while let Some(key) = iter.next().await.unwrap() {
                keys.push(key.key);
            }
            assert_eq!(keys.len(), 3);
            assert_eq!(keys[0], Bytes::from("key-a"));
            assert_eq!(keys[1], Bytes::from("key-b"));
            assert_eq!(keys[2], Bytes::from("key-c"));
        }

        #[tokio::test]
        async fn should_iterate_keys_across_multiple_segments() {
            // given
            let storage = LogStorage::in_memory();
            write_listing_entry(&storage, 0, b"key-a").await;
            write_listing_entry(&storage, 1, b"key-b").await;
            write_listing_entry(&storage, 2, b"key-c").await;

            // when
            let mut iter = storage.as_read().list_keys(0..3).await.unwrap();

            // then
            let mut keys = Vec::new();
            while let Some(key) = iter.next().await.unwrap() {
                keys.push(key.key);
            }
            assert_eq!(keys.len(), 3);
        }

        #[tokio::test]
        async fn should_deduplicate_keys_across_segments() {
            // given - same key in multiple segments
            let storage = LogStorage::in_memory();
            write_listing_entry(&storage, 0, b"shared-key").await;
            write_listing_entry(&storage, 1, b"shared-key").await;
            write_listing_entry(&storage, 2, b"shared-key").await;

            // when
            let mut iter = storage.as_read().list_keys(0..3).await.unwrap();

            // then - only one instance of the key
            let mut keys = Vec::new();
            while let Some(key) = iter.next().await.unwrap() {
                keys.push(key.key);
            }
            assert_eq!(keys.len(), 1);
            assert_eq!(keys[0], Bytes::from("shared-key"));
        }

        #[tokio::test]
        async fn should_respect_segment_range() {
            // given
            let storage = LogStorage::in_memory();
            write_listing_entry(&storage, 0, b"key-0").await;
            write_listing_entry(&storage, 1, b"key-1").await;
            write_listing_entry(&storage, 2, b"key-2").await;
            write_listing_entry(&storage, 3, b"key-3").await;

            // when - only query segments 1..3
            let mut iter = storage.as_read().list_keys(1..3).await.unwrap();

            // then - only keys from segments 1 and 2
            let mut keys = Vec::new();
            while let Some(key) = iter.next().await.unwrap() {
                keys.push(key.key);
            }
            assert_eq!(keys.len(), 2);
            assert_eq!(keys[0], Bytes::from("key-1"));
            assert_eq!(keys[1], Bytes::from("key-2"));
        }

        #[tokio::test]
        async fn should_return_keys_in_lexicographic_order() {
            // given - keys inserted out of order
            let storage = LogStorage::in_memory();
            write_listing_entry(&storage, 0, b"zebra").await;
            write_listing_entry(&storage, 0, b"apple").await;
            write_listing_entry(&storage, 0, b"mango").await;

            // when
            let mut iter = storage.as_read().list_keys(0..1).await.unwrap();

            // then - keys returned in lexicographic order
            let mut keys = Vec::new();
            while let Some(key) = iter.next().await.unwrap() {
                keys.push(key.key);
            }
            assert_eq!(keys[0], Bytes::from("apple"));
            assert_eq!(keys[1], Bytes::from("mango"));
            assert_eq!(keys[2], Bytes::from("zebra"));
        }
    }

    mod listing_cache {
        use super::*;

        #[test]
        fn should_build_delta_with_new_keys() {
            // given
            let cache = ListingCache::new();
            let seg_delta = test_seg_delta(0);
            let keys = vec![Bytes::from("key1"), Bytes::from("key2")];
            let mut records = Vec::new();

            // when
            let delta = cache.build_delta(&seg_delta, &keys, &mut records);

            // then
            assert_eq!(delta.new_keys.len(), 2);
            assert_eq!(records.len(), 2);
        }

        #[test]
        fn should_exclude_cached_keys_from_delta() {
            // given
            let mut cache = ListingCache::new();
            let seg_delta = test_seg_delta(0);
            let keys1 = vec![Bytes::from("key1"), Bytes::from("key2")];
            let mut records1 = Vec::new();

            // First batch
            let delta1 = cache.build_delta(&seg_delta, &keys1, &mut records1);
            cache.apply_delta(delta1);

            // when - second batch with overlap
            let keys2 = vec![Bytes::from("key2"), Bytes::from("key3")];
            let mut records2 = Vec::new();
            let delta2 = cache.build_delta(&seg_delta, &keys2, &mut records2);

            // then - only key3 is new
            assert_eq!(delta2.new_keys.len(), 1);
            assert!(delta2.new_keys.contains(&Bytes::from("key3")));
            assert_eq!(records2.len(), 1);
        }

        #[test]
        fn should_dedupe_keys_within_batch() {
            // given
            let cache = ListingCache::new();
            let seg_delta = test_seg_delta(0);
            let keys = vec![
                Bytes::from("key1"),
                Bytes::from("key2"),
                Bytes::from("key1"), // duplicate
            ];
            let mut records = Vec::new();

            // when
            let delta = cache.build_delta(&seg_delta, &keys, &mut records);

            // then
            assert_eq!(delta.new_keys.len(), 2);
            assert_eq!(records.len(), 2);
        }

        #[test]
        fn should_include_all_keys_after_segment_change() {
            // given
            let mut cache = ListingCache::new();
            let keys = vec![Bytes::from("key1"), Bytes::from("key2")];

            // First segment
            let seg_delta0 = test_seg_delta(0);
            let mut records0 = Vec::new();
            let delta0 = cache.build_delta(&seg_delta0, &keys, &mut records0);
            cache.apply_delta(delta0);

            // when - new segment with same keys
            let seg_delta1 = test_seg_delta(1);
            let mut records1 = Vec::new();
            let delta1 = cache.build_delta(&seg_delta1, &keys, &mut records1);

            // then - all keys are new in new segment
            assert_eq!(delta1.new_keys.len(), 2);
            assert_eq!(records1.len(), 2);
        }

        #[test]
        fn should_clear_cache_when_applying_delta_for_new_segment() {
            // given
            let mut cache = ListingCache::new();
            let seg_delta0 = test_seg_delta(0);
            let mut records0 = Vec::new();
            let delta0 = cache.build_delta(&seg_delta0, &[Bytes::from("key1")], &mut records0);
            cache.apply_delta(delta0);

            // when - apply delta for different segment
            let seg_delta1 = test_seg_delta(1);
            let mut records1 = Vec::new();
            let delta1 = cache.build_delta(&seg_delta1, &[Bytes::from("key2")], &mut records1);
            cache.apply_delta(delta1);

            // then - key1 should be new again (cache cleared)
            assert!(cache.is_new(1, &Bytes::from("key1")));
            // but key2 should not be new
            assert!(!cache.is_new(1, &Bytes::from("key2")));
        }

        #[test]
        fn should_create_correct_listing_entry_records() {
            // given
            let cache = ListingCache::new();
            let seg_delta = test_seg_delta(42);
            let keys = vec![Bytes::from("mykey")];
            let mut records = Vec::new();

            // when
            cache.build_delta(&seg_delta, &keys, &mut records);

            // then
            assert_eq!(records.len(), 1);
            let expected_key = ListingEntryKey::new(42, Bytes::from("mykey")).serialize();
            let expected_value = ListingEntryValue::new().serialize();
            assert_eq!(records[0].key, expected_key);
            assert_eq!(records[0].value, expected_value);
        }
    }
}