OXYGENHUB
# 252e8cb23fbc oxygen / crates / oxy-core / src / store.rs
26.5 KiB · Rust · last changed Garbage collection: operation-log retention, pruning with grace period, pack rewriting; oxy gc --now 1 month ago source blame history raw
1
//! Object store (spec §13.1, §13.3, §13.4): loose objects, one file per
2
//! object under `objects/loose/<hh>/<rest>`, and packs, `pack-<id>.pack`
3
//! with a sorted `pack-<id>.idx`, holding zstd-compressed objects. Reads
4
//! check loose objects first, then every pack; every read re-hashes.
5
//!
6
//! Pack file: `OXYPACK1` then entries of
7
//! `type u8 | flags u8 (bit 0: zstd) | stored_len u32 | raw_len u32 | ref 32 | body`.
8
//! Index file: `OXYIDX01` then `count u32` then sorted entries of
9
//! `ref 32 | offset u64 | stored_len u32 | raw_len u32 | type u8 | flags u8`.
10
11
use crate::attest::Attest;
12
use crate::hash::{hash_object, header, ObjectRef, ObjectType};
13
use crate::identity::{Identity, Signature};
14
use crate::objects::{Change, ObjectError, StateObject};
15
use rand::RngCore;
16
use std::cell::RefCell;
17
use std::fs;
18
use std::io::{self, Read, Seek, SeekFrom, Write};
19
use std::path::{Path, PathBuf};
20
21
const PACK_MAGIC: &[u8; 8] = b"OXYPACK1";
22
const IDX_MAGIC: &[u8; 8] = b"OXYIDX01";
23
const IDX_ENTRY: usize = 32 + 8 + 4 + 4 + 1 + 1;
24
const FLAG_ZSTD: u8 = 1;
25
const ZSTD_LEVEL: i32 = 3;
26
27
#[derive(Clone, Copy, Debug)]
28
struct PackEntry {
29
    offset: u64,
30
    stored_len: u32,
31
    raw_len: u32,
32
    ty: ObjectType,
33
    flags: u8,
34
}
35
36
/// One pack's index, fully in memory and sorted by ObjectRef.
37
#[derive(Debug)]
38
struct PackIndex {
39
    path: PathBuf,
40
    refs: Vec<ObjectRef>,
41
    entries: Vec<PackEntry>,
42
}
43
44
impl PackIndex {
45
    fn load(idx_path: &Path, pack_path: &Path) -> Result<PackIndex> {
46
        let data = fs::read(idx_path)?;
47
        let bad = |m: &str| StoreError::Corrupt(ObjectRef::ZERO, format!("{}: {m}", idx_path.display()));
48
        if data.len() < 12 || &data[..8] != IDX_MAGIC {
49
            return Err(bad("bad index header"));
50
        }
51
        let count = u32::from_le_bytes(data[8..12].try_into().unwrap()) as usize;
52
        if data.len() != 12 + count * IDX_ENTRY {
53
            return Err(bad("index length mismatch"));
54
        }
55
        let mut refs = Vec::with_capacity(count);
56
        let mut entries = Vec::with_capacity(count);
57
        for i in 0..count {
58
            let e = &data[12 + i * IDX_ENTRY..12 + (i + 1) * IDX_ENTRY];
59
            let r = ObjectRef::from_slice(&e[..32]).unwrap();
60
            if let Some(prev) = refs.last() {
61
                if r <= *prev {
62
                    return Err(bad("index not sorted"));
63
                }
64
            }
65
            refs.push(r);
66
            entries.push(PackEntry {
67
                offset: u64::from_le_bytes(e[32..40].try_into().unwrap()),
68
                stored_len: u32::from_le_bytes(e[40..44].try_into().unwrap()),
69
                raw_len: u32::from_le_bytes(e[44..48].try_into().unwrap()),
70
                ty: ObjectType::ALL.get(e[48] as usize).copied().ok_or_else(|| bad("unknown object type"))?,
71
                flags: e[49],
72
            });
73
        }
74
        Ok(PackIndex { path: pack_path.to_path_buf(), refs, entries })
75
    }
76
77
    fn find(&self, r: &ObjectRef) -> Option<PackEntry> {
78
        self.refs.binary_search(r).ok().map(|i| self.entries[i])
79
    }
80
81
    fn read(&self, r: &ObjectRef, e: PackEntry) -> Result<(ObjectType, Vec<u8>)> {
82
        let mut f = fs::File::open(&self.path)?;
83
        f.seek(SeekFrom::Start(e.offset))?;
84
        let mut stored = vec![0u8; e.stored_len as usize];
85
        f.read_exact(&mut stored)?;
86
        let body = if e.flags & FLAG_ZSTD != 0 {
87
            zstd::bulk::decompress(&stored, e.raw_len as usize).map_err(|err| StoreError::Corrupt(*r, format!("zstd: {err}")))?
88
        } else {
89
            stored
90
        };
91
        if body.len() != e.raw_len as usize || hash_object(e.ty, &body) != *r {
92
            return Err(StoreError::Corrupt(*r, "packed object does not match its reference".into()));
93
        }
94
        Ok((e.ty, body))
95
    }
96
}
97
98
fn type_index(ty: ObjectType) -> u8 {
99
    ObjectType::ALL.iter().position(|t| *t == ty).unwrap() as u8
100
}
101
102
/// What `pack_loose` did.
103
#[derive(Debug, Clone, Default, PartialEq, Eq)]
104
pub struct PackReport {
105
    pub objects: usize,
106
    pub raw_bytes: u64,
107
    pub packed_bytes: u64,
108
}
109
110
/// What `prune` did.
111
#[derive(Debug, Clone, Default, PartialEq, Eq)]
112
pub struct PruneReport {
113
    pub objects: usize,
114
    pub bytes: u64,
115
    /// Unreachable objects left alone because they are younger than the grace period.
116
    pub deferred: usize,
117
}
118
119
#[derive(Debug, thiserror::Error)]
120
pub enum StoreError {
121
    #[error("io: {0}")]
122
    Io(#[from] io::Error),
123
    #[error("corrupt object {0}: {1}")]
124
    Corrupt(ObjectRef, String),
125
    #[error("object {0}")]
126
    Object(#[from] ObjectError),
127
    #[error("object {0} is a {1}, expected {2}")]
128
    WrongType(ObjectRef, &'static str, &'static str),
129
    #[error("object {0} not found")]
130
    NotFound(ObjectRef),
131
}
132
133
pub type Result<T> = std::result::Result<T, StoreError>;
134
135
#[derive(Debug)]
136
pub struct Store {
137
    dir: PathBuf,
138
    packs: RefCell<Vec<PackIndex>>,
139
}
140
141
impl Store {
142
    pub fn open(dir: impl Into<PathBuf>) -> io::Result<Store> {
143
        let dir = dir.into();
144
        fs::create_dir_all(dir.join("loose"))?;
145
        let store = Store { dir, packs: RefCell::new(Vec::new()) };
146
        store.reload_packs().map_err(|e| io::Error::other(e.to_string()))?;
147
        Ok(store)
148
    }
149
150
    fn reload_packs(&self) -> Result<()> {
151
        let mut packs = Vec::new();
152
        let mut names: Vec<PathBuf> = fs::read_dir(&self.dir)?
153
            .filter_map(|e| e.ok().map(|e| e.path()))
154
            .filter(|p| p.extension().is_some_and(|x| x == "idx") && p.file_name().is_some_and(|n| n.to_string_lossy().starts_with("pack-")))
155
            .collect();
156
        names.sort();
157
        for idx in names {
158
            let pack = idx.with_extension("pack");
159
            if pack.is_file() {
160
                packs.push(PackIndex::load(&idx, &pack)?);
161
            }
162
        }
163
        *self.packs.borrow_mut() = packs;
164
        Ok(())
165
    }
166
167
    pub fn pack_count(&self) -> usize {
168
        self.packs.borrow().len()
169
    }
170
171
    fn find_packed(&self, r: &ObjectRef) -> Option<(usize, PackEntry)> {
172
        self.packs.borrow().iter().enumerate().find_map(|(i, p)| p.find(r).map(|e| (i, e)))
173
    }
174
175
    /// Loose objects on disk, with their paths.
176
    fn loose_objects(&self) -> io::Result<Vec<(ObjectRef, PathBuf)>> {
177
        let mut out = Vec::new();
178
        let loose = self.dir.join("loose");
179
        for d in fs::read_dir(&loose)? {
180
            let d = d?;
181
            if !d.file_type()?.is_dir() {
182
                continue;
183
            }
184
            let prefix = d.file_name().to_string_lossy().to_string();
185
            for f in fs::read_dir(d.path())? {
186
                let f = f?;
187
                let name = f.file_name().to_string_lossy().to_string();
188
                if let Some(r) = ObjectRef::from_hex(&format!("{prefix}{name}")) {
189
                    out.push((r, f.path()));
190
                }
191
            }
192
        }
193
        out.sort();
194
        Ok(out)
195
    }
196
197
    pub fn loose_count(&self) -> io::Result<usize> {
198
        Ok(self.loose_objects()?.len())
199
    }
200
201
    /// Move every loose object into a new pack (spec §13.4). Loose files are
202
    /// deleted only after the pack and its index are durable.
203
    pub fn pack_loose(&self) -> Result<PackReport> {
204
        let loose = self.loose_objects()?;
205
        self.write_pack(&loose.iter().map(|(r, _)| *r).collect::<Vec<_>>(), &loose.iter().map(|(_, p)| p.clone()).collect::<Vec<_>>(), &[])
206
    }
207
208
    /// Delete every object not in `keep` (spec §21). Loose objects and packs
209
    /// younger than `grace` are left alone, since a concurrent writer may not
210
    /// have recorded its references yet. Packs holding unreachable objects
211
    /// are rewritten without them. Signature sidecars and seal-index entries
212
    /// of deleted objects go too.
213
    pub fn prune(&self, keep: &std::collections::HashSet<ObjectRef>, grace: std::time::Duration) -> Result<PruneReport> {
214
        let now = std::time::SystemTime::now();
215
        let old_enough = |m: &fs::Metadata| m.modified().map(|t| now.duration_since(t).unwrap_or_default() >= grace).unwrap_or(true);
216
        let mut report = PruneReport::default();
217
        // Loose objects.
218
        for (r, path) in self.loose_objects()? {
219
            if keep.contains(&r) {
220
                continue;
221
            }
222
            let meta = fs::metadata(&path)?;
223
            if !old_enough(&meta) {
224
                report.deferred += 1;
225
                continue;
226
            }
227
            report.bytes += meta.len();
228
            report.objects += 1;
229
            fs::remove_file(&path)?;
230
            let _ = fs::remove_dir(path.parent().unwrap());
231
            self.remove_sidecars(&r)?;
232
        }
233
        // Packs: rewrite those that hold unreachable objects.
234
        let mut keep_refs: Vec<ObjectRef> = Vec::new();
235
        let mut old_paths: Vec<PathBuf> = Vec::new();
236
        let mut rewrite = false;
237
        for p in self.packs.borrow().iter() {
238
            let meta = fs::metadata(&p.path)?;
239
            let unreachable = p.refs.iter().filter(|r| !keep.contains(r)).count();
240
            if unreachable == 0 {
241
                continue;
242
            }
243
            if !old_enough(&meta) {
244
                report.deferred += unreachable;
245
                continue;
246
            }
247
            rewrite = true;
248
            old_paths.push(p.path.clone());
249
            for (r, e) in p.refs.iter().zip(p.entries.iter()) {
250
                if keep.contains(r) {
251
                    keep_refs.push(*r);
252
                } else {
253
                    report.objects += 1;
254
                    report.bytes += e.stored_len as u64;
255
                }
256
            }
257
        }
258
        if rewrite {
259
            let dropped: Vec<ObjectRef> = {
260
                let packs = self.packs.borrow();
261
                packs.iter().filter(|p| old_paths.contains(&p.path)).flat_map(|p| p.refs.iter().copied()).filter(|r| !keep.contains(r)).collect()
262
            };
263
            self.write_pack(&keep_refs, &[], &old_paths)?;
264
            for r in dropped {
265
                self.remove_sidecars(&r)?;
266
            }
267
        }
268
        Ok(report)
269
    }
270
271
    fn remove_sidecars(&self, r: &ObjectRef) -> Result<()> {
272
        for path in [self.sig_path(r), self.seal_index_path(r)] {
273
            match fs::remove_file(&path) {
274
                Ok(()) => {
275
                    let _ = fs::remove_dir(path.parent().unwrap());
276
                }
277
                Err(e) if e.kind() == io::ErrorKind::NotFound => {}
278
                Err(e) => return Err(e.into()),
279
            }
280
        }
281
        Ok(())
282
    }
283
284
    /// Write `refs` into a new pack, then delete `loose_paths` and `old_packs`.
285
    fn write_pack(&self, refs: &[ObjectRef], loose_paths: &[PathBuf], old_packs: &[PathBuf]) -> Result<PackReport> {
286
        if refs.is_empty() {
287
            for p in old_packs {
288
                fs::remove_file(p)?;
289
                fs::remove_file(p.with_extension("idx"))?;
290
            }
291
            self.reload_packs()?;
292
            return Ok(PackReport::default());
293
        }
294
        let mut refs: Vec<ObjectRef> = refs.to_vec();
295
        refs.sort();
296
        refs.dedup();
297
        let mut nonce = [0u8; 8];
298
        rand::rngs::OsRng.fill_bytes(&mut nonce);
299
        let id = hex::encode(nonce);
300
        let pack_path = self.dir.join(format!("pack-{id}.pack"));
301
        let idx_path = self.dir.join(format!("pack-{id}.idx"));
302
        let tmp_pack = self.dir.join(format!("tmp-{id}.pack"));
303
        let tmp_idx = self.dir.join(format!("tmp-{id}.idx"));
304
        let mut report = PackReport::default();
305
        let mut index: Vec<(ObjectRef, PackEntry)> = Vec::with_capacity(refs.len());
306
        {
307
            let mut f = io::BufWriter::new(fs::File::create(&tmp_pack)?);
308
            f.write_all(PACK_MAGIC)?;
309
            let mut offset = PACK_MAGIC.len() as u64;
310
            for r in &refs {
311
                let (ty, body) = self.get(r)?;
312
                let compressed = zstd::bulk::compress(&body, ZSTD_LEVEL).map_err(|e| StoreError::Corrupt(*r, format!("zstd: {e}")))?;
313
                let (flags, stored) = if compressed.len() < body.len() { (FLAG_ZSTD, compressed) } else { (0, body.clone()) };
314
                let header_len = 1 + 1 + 4 + 4 + 32;
315
                f.write_all(&[type_index(ty), flags])?;
316
                f.write_all(&(stored.len() as u32).to_le_bytes())?;
317
                f.write_all(&(body.len() as u32).to_le_bytes())?;
318
                f.write_all(&r.0)?;
319
                f.write_all(&stored)?;
320
                index.push((*r, PackEntry { offset: offset + header_len, stored_len: stored.len() as u32, raw_len: body.len() as u32, ty, flags }));
321
                offset += header_len + stored.len() as u64;
322
                report.objects += 1;
323
                report.raw_bytes += body.len() as u64;
324
                report.packed_bytes += stored.len() as u64;
325
            }
326
            f.flush()?;
327
            f.get_ref().sync_all()?;
328
        }
329
        {
330
            let mut f = io::BufWriter::new(fs::File::create(&tmp_idx)?);
331
            f.write_all(IDX_MAGIC)?;
332
            f.write_all(&(index.len() as u32).to_le_bytes())?;
333
            for (r, e) in &index {
334
                f.write_all(&r.0)?;
335
                f.write_all(&e.offset.to_le_bytes())?;
336
                f.write_all(&e.stored_len.to_le_bytes())?;
337
                f.write_all(&e.raw_len.to_le_bytes())?;
338
                f.write_all(&[type_index(e.ty), e.flags])?;
339
            }
340
            f.flush()?;
341
            f.get_ref().sync_all()?;
342
        }
343
        fs::rename(&tmp_pack, &pack_path)?;
344
        fs::rename(&tmp_idx, &idx_path)?;
345
        for p in old_packs {
346
            fs::remove_file(p)?;
347
            fs::remove_file(p.with_extension("idx"))?;
348
        }
349
        self.reload_packs()?;
350
        for path in loose_paths {
351
            fs::remove_file(path)?;
352
            if let Some(dir) = path.parent() {
353
                let _ = fs::remove_dir(dir);
354
            }
355
        }
356
        Ok(report)
357
    }
358
359
    pub fn dir(&self) -> &Path {
360
        &self.dir
361
    }
362
363
    fn path_for(&self, r: &ObjectRef) -> PathBuf {
364
        let h = r.to_hex();
365
        self.dir.join("loose").join(&h[..2]).join(&h[2..])
366
    }
367
368
    pub fn has(&self, r: &ObjectRef) -> bool {
369
        self.path_for(r).is_file() || self.find_packed(r).is_some()
370
    }
371
372
    /// Store `body` as an object of type `ty`; idempotent.
373
    pub fn put(&self, ty: ObjectType, body: &[u8]) -> Result<ObjectRef> {
374
        let r = hash_object(ty, body);
375
        let path = self.path_for(&r);
376
        if path.is_file() || self.find_packed(&r).is_some() {
377
            return Ok(r);
378
        }
379
        fs::create_dir_all(path.parent().unwrap())?;
380
        let mut nonce = [0u8; 8];
381
        rand::rngs::OsRng.fill_bytes(&mut nonce);
382
        let tmp = self.dir.join("loose").join(format!("tmp-{}", hex::encode(nonce)));
383
        {
384
            let mut f = fs::File::create(&tmp)?;
385
            f.write_all(&header(ty, body.len()))?;
386
            f.write_all(body)?;
387
            f.sync_all()?;
388
        }
389
        fs::rename(&tmp, &path)?;
390
        Ok(r)
391
    }
392
393
    /// Read and verify an object, loose or packed.
394
    pub fn get(&self, r: &ObjectRef) -> Result<(ObjectType, Vec<u8>)> {
395
        let path = self.path_for(r);
396
        let data = match fs::read(&path) {
397
            Ok(d) => d,
398
            Err(e) if e.kind() == io::ErrorKind::NotFound => {
399
                return match self.find_packed(r) {
400
                    Some((i, e)) => self.packs.borrow()[i].read(r, e),
401
                    None => Err(StoreError::NotFound(*r)),
402
                };
403
            }
404
            Err(e) => return Err(e.into()),
405
        };
406
        let nul = data.iter().position(|&b| b == 0).ok_or_else(|| StoreError::Corrupt(*r, "no header".into()))?;
407
        let head = std::str::from_utf8(&data[..nul]).map_err(|_| StoreError::Corrupt(*r, "header not UTF-8".into()))?;
408
        let mut parts = head.split(' ');
409
        let (Some("oxy1"), Some(ty), Some(len), None) = (parts.next(), parts.next(), parts.next(), parts.next()) else {
410
            return Err(StoreError::Corrupt(*r, format!("bad header {head:?}")));
411
        };
412
        let ty = ObjectType::from_name(ty).ok_or_else(|| StoreError::Corrupt(*r, format!("unknown type {ty}")))?;
413
        let len: usize = len.parse().map_err(|_| StoreError::Corrupt(*r, "bad length".into()))?;
414
        let body = &data[nul + 1..];
415
        if body.len() != len {
416
            return Err(StoreError::Corrupt(*r, format!("length {} != header {len}", body.len())));
417
        }
418
        if hash_object(ty, body) != *r {
419
            return Err(StoreError::Corrupt(*r, "hash mismatch".into()));
420
        }
421
        Ok((ty, body.to_vec()))
422
    }
423
424
    pub fn get_typed(&self, r: &ObjectRef, want: ObjectType) -> Result<Vec<u8>> {
425
        let (ty, body) = self.get(r)?;
426
        if ty != want {
427
            return Err(StoreError::WrongType(*r, ty.name(), want.name()));
428
        }
429
        Ok(body)
430
    }
431
432
    pub fn put_change(&self, c: &Change) -> Result<ObjectRef> {
433
        self.put(ObjectType::Change, &c.encode())
434
    }
435
436
    pub fn get_change(&self, r: &ObjectRef) -> Result<Change> {
437
        Ok(Change::decode(&self.get_typed(r, ObjectType::Change)?)?)
438
    }
439
440
    pub fn put_chunk(&self, bytes: &[u8]) -> Result<ObjectRef> {
441
        self.put(ObjectType::Chunk, bytes)
442
    }
443
444
    pub fn put_state(&self, st: &StateObject) -> Result<ObjectRef> {
445
        self.put(ObjectType::State, &st.encode())
446
    }
447
448
    pub fn get_state(&self, r: &ObjectRef) -> Result<StateObject> {
449
        Ok(StateObject::decode(&self.get_typed(r, ObjectType::State)?)?)
450
    }
451
452
    pub fn put_attest(&self, a: &Attest) -> Result<ObjectRef> {
453
        self.put(ObjectType::Attest, &a.encode())
454
    }
455
456
    pub fn get_attest(&self, r: &ObjectRef) -> Result<Attest> {
457
        Ok(Attest::decode(&self.get_typed(r, ObjectType::Attest)?)?)
458
    }
459
460
    fn seal_index_path(&self, change: &ObjectRef) -> PathBuf {
461
        let h = change.to_hex();
462
        self.dir.join("seals").join(&h[..2]).join(&h[2..])
463
    }
464
465
    /// Record that seal attestation `attest` covers `change`.
466
    pub fn index_seal(&self, change: &ObjectRef, attest: &ObjectRef) -> Result<()> {
467
        let mut seals = self.seals_for(change)?;
468
        if seals.contains(attest) {
469
            return Ok(());
470
        }
471
        seals.push(*attest);
472
        seals.sort();
473
        let path = self.seal_index_path(change);
474
        fs::create_dir_all(path.parent().unwrap())?;
475
        let text: String = seals.iter().map(|s| format!("{}\n", s.to_hex())).collect();
476
        let tmp = path.with_extension("tmp");
477
        fs::write(&tmp, text)?;
478
        fs::rename(&tmp, &path)?;
479
        Ok(())
480
    }
481
482
    /// Seal attestations known to cover `change`.
483
    pub fn seals_for(&self, change: &ObjectRef) -> Result<Vec<ObjectRef>> {
484
        match fs::read_to_string(self.seal_index_path(change)) {
485
            Ok(t) => Ok(t.lines().filter_map(ObjectRef::from_hex).collect()),
486
            Err(e) if e.kind() == io::ErrorKind::NotFound => Ok(vec![]),
487
            Err(e) => Err(e.into()),
488
        }
489
    }
490
491
    pub fn put_identity(&self, id: &Identity) -> Result<ObjectRef> {
492
        self.put(ObjectType::Identity, &id.encode())
493
    }
494
495
    pub fn get_identity(&self, r: &ObjectRef) -> Result<Identity> {
496
        Ok(Identity::decode(&self.get_typed(r, ObjectType::Identity)?)?)
497
    }
498
499
    fn sig_path(&self, r: &ObjectRef) -> PathBuf {
500
        let h = r.to_hex();
501
        self.dir.join("sigs").join(&h[..2]).join(&h[2..])
502
    }
503
504
    /// Signatures stored alongside object `r` (spec §5.4). Empty if none.
505
    pub fn signatures(&self, r: &ObjectRef) -> Result<Vec<Signature>> {
506
        match fs::read(self.sig_path(r)) {
507
            Ok(b) => Ok(Signature::decode_list(&b)?),
508
            Err(e) if e.kind() == io::ErrorKind::NotFound => Ok(vec![]),
509
            Err(e) => Err(e.into()),
510
        }
511
    }
512
513
    /// Attach a signature to object `r`, replacing any earlier one by the same key.
514
    pub fn put_signature(&self, r: &ObjectRef, sig: &Signature) -> Result<()> {
515
        let mut sigs = self.signatures(r)?;
516
        sigs.retain(|s| s.key != sig.key);
517
        sigs.push(sig.clone());
518
        sigs.sort_by_key(|s| s.key);
519
        let path = self.sig_path(r);
520
        fs::create_dir_all(path.parent().unwrap())?;
521
        let mut nonce = [0u8; 8];
522
        rand::rngs::OsRng.fill_bytes(&mut nonce);
523
        let tmp = self.dir.join(format!("sigs-tmp-{}", hex::encode(nonce)));
524
        {
525
            let mut f = fs::File::create(&tmp)?;
526
            f.write_all(&Signature::encode_list(&sigs))?;
527
            f.sync_all()?;
528
        }
529
        fs::rename(&tmp, &path)?;
530
        Ok(())
531
    }
532
533
    /// Every object reference in the store, loose and packed (slow; for gc and tests).
534
    pub fn list(&self) -> io::Result<Vec<ObjectRef>> {
535
        let mut out: Vec<ObjectRef> = self.loose_objects()?.into_iter().map(|(r, _)| r).collect();
536
        for p in self.packs.borrow().iter() {
537
            out.extend(p.refs.iter().copied());
538
        }
539
        out.sort();
540
        out.dedup();
541
        Ok(out)
542
    }
543
}
544
545
#[cfg(test)]
546
mod tests {
547
    use super::*;
548
549
    #[test]
550
    fn put_get_verify() {
551
        let dir = tempfile::tempdir().unwrap();
552
        let s = Store::open(dir.path().join("objects")).unwrap();
553
        let r = s.put(ObjectType::Chunk, b"hello").unwrap();
554
        assert!(s.has(&r));
555
        assert_eq!(s.get(&r).unwrap(), (ObjectType::Chunk, b"hello".to_vec()));
556
        assert_eq!(s.put(ObjectType::Chunk, b"hello").unwrap(), r);
557
        assert_eq!(s.list().unwrap(), vec![r]);
558
        assert!(matches!(s.get_typed(&r, ObjectType::Blob), Err(StoreError::WrongType(..))));
559
        assert!(matches!(s.get(&ObjectRef([9; 32])), Err(StoreError::NotFound(_))));
560
561
        // Tamper: flip a byte, read must fail.
562
        let path = s.path_for(&r);
563
        let mut data = fs::read(&path).unwrap();
564
        let last = data.len() - 1;
565
        data[last] ^= 1;
566
        fs::write(&path, data).unwrap();
567
        assert!(matches!(s.get(&r), Err(StoreError::Corrupt(..))));
568
    }
569
570
    #[test]
571
    fn packing_keeps_every_object_readable() {
572
        let dir = tempfile::tempdir().unwrap();
573
        let s = Store::open(dir.path()).unwrap();
574
        let mut refs = Vec::new();
575
        for i in 0..300u32 {
576
            let body = format!("object {i} ").repeat(1 + (i % 7) as usize);
577
            refs.push((s.put(ObjectType::Chunk, body.as_bytes()).unwrap(), body));
578
        }
579
        let c = crate::objects::tests::sample_change();
580
        let cr = s.put_change(&c).unwrap();
581
        assert_eq!(s.loose_count().unwrap(), 301);
582
        let report = s.pack_loose().unwrap();
583
        assert_eq!(report.objects, 301);
584
        assert!(report.packed_bytes < report.raw_bytes, "{report:?}");
585
        assert_eq!(s.loose_count().unwrap(), 0);
586
        assert_eq!(s.pack_count(), 1);
587
        for (r, body) in &refs {
588
            assert!(s.has(r));
589
            assert_eq!(s.get(r).unwrap(), (ObjectType::Chunk, body.as_bytes().to_vec()));
590
        }
591
        assert_eq!(s.get_change(&cr).unwrap(), c);
592
        assert_eq!(s.list().unwrap().len(), 301);
593
        // Re-putting a packed object is a no-op; new objects go loose and a
594
        // second pack coexists with the first.
595
        assert_eq!(s.put(ObjectType::Chunk, refs[0].1.as_bytes()).unwrap(), refs[0].0);
596
        assert_eq!(s.loose_count().unwrap(), 0);
597
        let extra = s.put(ObjectType::Chunk, b"after packing").unwrap();
598
        s.pack_loose().unwrap();
599
        assert_eq!(s.pack_count(), 2);
600
        assert_eq!(s.get(&extra).unwrap().1, b"after packing");
601
        // A reopened store sees the packs.
602
        let s2 = Store::open(dir.path()).unwrap();
603
        assert_eq!(s2.pack_count(), 2);
604
        assert_eq!(s2.get(&refs[5].0).unwrap().1, refs[5].1.as_bytes());
605
        assert!(matches!(s2.get(&ObjectRef([9; 32])), Err(StoreError::NotFound(_))));
606
        // Corrupting an object's body inside the pack is detected on read.
607
        let target = refs[5].0;
608
        let (pack_path, offset) = {
609
            let packs = s2.packs.borrow();
610
            let (i, e) = s2.find_packed(&target).unwrap();
611
            (packs[i].path.clone(), e.offset as usize)
612
        };
613
        let mut data = fs::read(&pack_path).unwrap();
614
        data[offset] ^= 0xff;
615
        fs::write(&pack_path, data).unwrap();
616
        let s3 = Store::open(dir.path()).unwrap();
617
        assert!(matches!(s3.get(&target), Err(StoreError::Corrupt(..))));
618
        assert_eq!(s3.get(&refs[6].0).unwrap().1, refs[6].1.as_bytes(), "other objects are unaffected");
619
    }
620
621
    #[test]
622
    fn prune_removes_unreachable_loose_and_packed_objects() {
623
        use std::collections::HashSet;
624
        let dir = tempfile::tempdir().unwrap();
625
        let s = Store::open(dir.path()).unwrap();
626
        let keep_a = s.put(ObjectType::Chunk, b"keep a").unwrap();
627
        let drop_a = s.put(ObjectType::Chunk, b"drop a").unwrap();
628
        s.pack_loose().unwrap();
629
        let keep_b = s.put(ObjectType::Chunk, b"keep b").unwrap();
630
        let drop_b = s.put(ObjectType::Chunk, b"drop b").unwrap();
631
        let sk = crate::identity::SecretKey { ed_seed: [1; 32], ml_seed: [2; 32] };
632
        let id = sk.identity("a", "b", 0);
633
        let key = s.put_identity(&id).unwrap();
634
        s.put_signature(&drop_b, &sk.sign(crate::identity::Domain::Change, &key, &drop_b)).unwrap();
635
        s.index_seal(&drop_a, &ObjectRef([3; 32])).unwrap();
636
        let keep: HashSet<ObjectRef> = [keep_a, keep_b, key].into_iter().collect();
637
        // Everything is too young under a grace period.
638
        let r = s.prune(&keep, std::time::Duration::from_secs(3600)).unwrap();
639
        assert_eq!((r.objects, r.deferred), (0, 2));
640
        let r = s.prune(&keep, std::time::Duration::ZERO).unwrap();
641
        assert_eq!(r.objects, 2, "{r:?}");
642
        assert!(s.has(&keep_a) && s.has(&keep_b) && s.has(&key));
643
        assert!(!s.has(&drop_a) && !s.has(&drop_b));
644
        assert!(s.signatures(&drop_b).unwrap().is_empty());
645
        assert!(s.seals_for(&drop_a).unwrap().is_empty());
646
        assert_eq!(s.pack_count(), 1);
647
        assert_eq!(s.get(&keep_a).unwrap().1, b"keep a");
648
        assert_eq!(s.list().unwrap().len(), 3);
649
    }
650
651
    #[test]
652
    fn signatures_are_stored_beside_objects() {
653
        use crate::identity::{Domain, SecretKey};
654
        let dir = tempfile::tempdir().unwrap();
655
        let s = Store::open(dir.path()).unwrap();
656
        let r = s.put(ObjectType::Chunk, b"x").unwrap();
657
        assert!(s.signatures(&r).unwrap().is_empty());
658
        let sk = SecretKey { ed_seed: [1; 32], ml_seed: [2; 32] };
659
        let id = sk.identity("a", "b", 0);
660
        let key = s.put_identity(&id).unwrap();
661
        assert_eq!(s.get_identity(&key).unwrap(), id);
662
        let sig = sk.sign(Domain::Change, &key, &r);
663
        s.put_signature(&r, &sig).unwrap();
664
        s.put_signature(&r, &sig).unwrap();
665
        assert_eq!(s.signatures(&r).unwrap(), vec![sig]);
666
        // The object itself is untouched by its signatures.
667
        assert_eq!(s.get(&r).unwrap().1, b"x");
668
    }
669
670
    #[test]
671
    fn change_round_trip_through_store() {
672
        let dir = tempfile::tempdir().unwrap();
673
        let s = Store::open(dir.path()).unwrap();
674
        let c = crate::objects::tests::sample_change();
675
        let r = s.put_change(&c).unwrap();
676
        assert_eq!(r, c.hash());
677
        assert_eq!(s.get_change(&r).unwrap(), c);
678
    }
679
}