test store

Cargo.toml

@@ -47,6 +47,7 @@ murmurhash32 = "0.2"
 chrono = "0.4"
 smallvec = "1"
 rayon = "1"
+env_logger = "0.8"
 lru = "0.6"
 
 [target.'cfg(windows)'.dependencies]

src/core/index.rs

@@ -35,12 +35,18 @@ fn load_metas(
     inventory: &SegmentMetaInventory,
 ) -> crate::Result<IndexMeta> {
     let meta_data = directory.atomic_read(&META_FILEPATH)?;
-    let meta_string = String::from_utf8_lossy(&meta_data);
+    let meta_string = String::from_utf8(meta_data)
+        .map_err(|utf8_err| {
+            DataCorruption::new(
+                META_FILEPATH.to_path_buf(),
+                format!("Meta file is not valid utf-8. {:?}", utf8_err)
+            )
+        })?;
     IndexMeta::deserialize(&meta_string, &inventory)
         .map_err(|e| {
             DataCorruption::new(
                 META_FILEPATH.to_path_buf(),
-                format!("Meta file cannot be deserialized. {:?}.", e),
+                format!("Meta file cannot be deserialized. {:?}. content = {}", e, meta_string),
             )
         })
         .map_err(From::from)

src/functional_test.rs

@@ -20,6 +20,7 @@ fn check_index_content(searcher: &Searcher, vals: &[u64]) -> crate::Result<()> {
 #[test]
 #[ignore]
 fn test_functional_store() -> crate::Result<()> {
+    env_logger::init();
     let mut schema_builder = Schema::builder();
 
     let id_field = schema_builder.add_u64_field("id", INDEXED | STORED);
@@ -35,8 +36,7 @@ fn test_functional_store() -> crate::Result<()> {
     let mut doc_set: Vec<u64> = Vec::new();
 
     let mut doc_id = 0u64;
-    for iteration in 0..500 {
+    for iteration in 0.. {
-        dbg!(iteration);
         let num_docs: usize = rng.gen_range(0..4);
         if doc_set.len() >= 1 {
             let doc_to_remove_id = rng.gen_range(0..doc_set.len());
@@ -51,6 +51,7 @@ fn test_functional_store() -> crate::Result<()> {
         index_writer.commit()?;
         reader.reload()?;
         let searcher = reader.searcher();
+        println!("#{} - {}", iteration, searcher.segment_readers().len());
         check_index_content(&searcher, &doc_set)?;
     }
     Ok(())

src/store/index/mod.rs

@@ -1,4 +1,4 @@
-const CHECKPOINT_PERIOD: usize = 8;
+const CHECKPOINT_PERIOD: usize = 2;
 
 use std::fmt;
 mod block;

src/store/tests_store.rs

@@ -0,0 +1,50 @@
+use std::path::Path;
+
+use crate::HasLen;
+use crate::directory::{Directory, ManagedDirectory, MmapDirectory, RAMDirectory};
+use crate::fastfield::DeleteBitSet;
+
+use super::{StoreReader, StoreWriter};
+
+#[test]
+fn test_toto2() -> crate::Result<()> {
+    let directory = ManagedDirectory::wrap(MmapDirectory::open("src/store/broken_seg")?)?;
+    let path = Path::new("b6029ade1b954ea1acad15b432eaacb9.store");
+    assert!(directory.validate_checksum(path)?);
+    let store_file = directory.open_read(path)?;
+    let store = StoreReader::open(store_file)?;
+    let documents = store.documents();
+    // for doc in documents {
+    //     println!("{:?}", doc);
+    // }
+    let doc= store.get(15_086)?;
+    Ok(())
+}
+
+#[test]
+fn test_toto() -> crate::Result<()> {
+    let directory = ManagedDirectory::wrap(MmapDirectory::open("src/store/broken_seg")?)?;
+    assert!(directory.validate_checksum(Path::new("e6ece22e5bca4e0dbe7ce3e4dcbd5bbf.store"))?);
+    let store_file = directory.open_read(Path::new("e6ece22e5bca4e0dbe7ce3e4dcbd5bbf.store.patched"))?;
+    let store = StoreReader::open(store_file)?;
+    let doc= store.get(53)?;
+    println!("{:?}", doc);
+    // let documents = store.documents();
+    // let ram_directory = RAMDirectory::create();
+    // let path = Path::new("store");
+
+    // let store_wrt = ram_directory.open_write(path)?;
+    // let mut store_writer = StoreWriter::new(store_wrt);
+    // for doc in &documents {
+    //     store_writer.store(doc)?;
+    // }
+    // store_writer.close()?;
+    // let store_data = ram_directory.open_read(path)?;
+    // let new_store = StoreReader::open(store_data)?;
+    // for doc in 0..59 {
+    //     println!("{}", doc);
+    //     let doc = new_store.get(doc)?;
+    //     println!("{:?}", doc);
+    // }
+    Ok(())
+}

src/store/writer.rs

@@ -10,7 +10,7 @@ use crate::store::index::Checkpoint;
 use crate::DocId;
 use std::io::{self, Write};
 
-const BLOCK_SIZE: usize = 16_384;
+const BLOCK_SIZE: usize = 30;
 
 /// Write tantivy's [`Store`](./index.html)
 ///