1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364
//! A [`crate::plain::Cache`] stores all cached file in a single
//! directory (there may also be a `.kismet_temp` subdirectory for
//! temporary files), and periodically scans for evictions with a
//! second chance strategy. This implementation does not scale up to
//! more than a few hundred files per cache directory (a
//! [`crate::sharded::Cache`] can go higher), but interoperates
//! seamlessly with other file-based programs that store cache files
//! in flat directories.
//!
//! This module is useful for lower level usage; in most cases, the
//! [`crate::Cache`] is more convenient and just as efficient. In
//! particular, a `crate::plain::Cache` *does not* invoke
//! [`std::fs::File::sync_all`] or [`std::fs::File::sync_data`]: the
//! caller should sync files before letting Kismet persist them in a
//! directory, if necessary.
//!
//! The cache's contents will grow past its stated capacity, but
//! should rarely reach more than twice that capacity.
use std::borrow::Cow;
use std::fs::File;
use std::io::Result;
use std::path::Path;
use std::path::PathBuf;
use crate::cache_dir::CacheDir;
use crate::trigger::PeriodicTrigger;
use crate::KISMET_TEMPORARY_SUBDIRECTORY as TEMP_SUBDIR;
/// How many times we want to trigger maintenance per "capacity"
/// inserts. For example, `MAINTENANCE_SCALE = 3` means we will
/// expect to trigger maintenance after inserting or updating
/// ~capacity / 3 files in the cache.
const MAINTENANCE_SCALE: usize = 3;
/// A "plain" cache is a single directory of files. Given a capacity
/// of `k` files, we will trigger a second chance maintance roughly
/// every `k / 3` (`k / 6` in the long run, given the way
/// `PeriodicTrigger` is implemented) insertions.
#[derive(Clone, Debug)]
pub struct Cache {
// The cached files are siblings of this directory for temporary
// files.
temp_dir: PathBuf,
// Initialised to trigger a second chance maintenance roughly
// every `capacity / MAINTENANCE_SCALE` cache writes.
trigger: PeriodicTrigger,
// The directory has a capacity of roughly this many files;
// between maintenance, the actual file count may temporarily
// exceed that capacity.
capacity: usize,
}
impl CacheDir for Cache {
#[inline]
fn temp_dir(&self) -> Cow<Path> {
Cow::from(&self.temp_dir)
}
#[inline]
fn base_dir(&self) -> Cow<Path> {
Cow::from(self.temp_dir.parent().unwrap_or(&self.temp_dir))
}
#[inline]
fn trigger(&self) -> &PeriodicTrigger {
&self.trigger
}
#[inline]
fn capacity(&self) -> usize {
self.capacity
}
}
impl Cache {
/// Returns a new cache for approximately `capacity` files in
/// `base_dir`.
pub fn new(base_dir: PathBuf, capacity: usize) -> Cache {
let mut temp_dir = base_dir;
temp_dir.push(TEMP_SUBDIR);
Cache {
temp_dir,
trigger: PeriodicTrigger::new((capacity / MAINTENANCE_SCALE) as u64),
capacity,
}
}
/// Returns a read-only file for `name` in the cache directory if
/// it exists, or None if there is no such file. Fails with
/// `ErrorKind::InvalidInput` if `name` is invalid (empty, or
/// starts with a dot or a forward or back slash).
///
///
/// Implicitly "touches" the cached file `name` if it exists.
pub fn get(&self, name: &str) -> Result<Option<File>> {
CacheDir::get(self, name)
}
/// Returns a temporary directory suitable for temporary files
/// that will be published to the cache directory.
pub fn temp_dir(&self) -> Result<Cow<Path>> {
CacheDir::ensure_temp_dir(self)
}
/// Inserts or overwrites the file at `value` as `name` in the
/// cache directory. Fails with `ErrorKind::InvalidInput` if
/// `name` is invalid (empty, or starts with a dot or a forward
/// or back slash).
///
/// Always consumes the file at `value` on success; may consume it
/// on error.
pub fn set(&self, name: &str, value: &Path) -> Result<()> {
CacheDir::set(self, name, value)?;
Ok(())
}
/// Inserts the file at `value` as `name` in the cache directory
/// if there is no such cached entry already, or touches the
/// cached file if it already exists. Fails with
/// `ErrorKind::InvalidInput` if `name` is invalid (empty, or
/// starts with a dot or a forward or back slash).
///
/// Always consumes the file at `value` on success; may consume it
/// on error.
pub fn put(&self, name: &str, value: &Path) -> Result<()> {
CacheDir::put(self, name, value)?;
Ok(())
}
/// Marks the cached file `name` as newly used, if it exists.
/// Fails with `ErrorKind::InvalidInput` if `name` is invalid
/// (empty, or starts with a dot or a forward or back slash).
///
/// Returns whether `name` exists.
pub fn touch(&self, name: &str) -> Result<bool> {
CacheDir::touch(self, name)
}
}
/// Put 20 files in a 10-file cache. We should find at least 10, but
/// fewer than 20, and their contents should match.
#[test]
fn smoke_test() {
use tempfile::NamedTempFile;
use test_dir::{DirBuilder, FileType, TestDir};
// The payload for file `i` is `PAYLOAD_MULTIPLIER * i`.
const PAYLOAD_MULTIPLIER: usize = 13;
// Also leave a file in the temporary subdirectory; we'll check
// that it gets cleaned up before leaving this function..
let temp = TestDir::temp()
.create(TEMP_SUBDIR, FileType::Dir)
.create(&format!("{}/garbage", TEMP_SUBDIR), FileType::ZeroFile(10));
// The garbage file must exist.
assert!(std::fs::metadata(temp.path(&format!("{}/garbage", TEMP_SUBDIR))).is_ok());
// Make sure the garbage file is old enough to be deleted.
std::thread::sleep(std::time::Duration::from_secs_f64(2.5));
let cache = Cache::new(temp.path("."), 10);
for i in 0..20 {
let name = format!("{}", i);
let tmp = NamedTempFile::new_in(cache.temp_dir().expect("temp_dir must succeed"))
.expect("new temp file must succeed");
std::fs::write(tmp.path(), format!("{}", PAYLOAD_MULTIPLIER * i))
.expect("write must succeed");
cache.put(&name, tmp.path()).expect("put must succeed");
}
let present: usize = (0..20)
.map(|i| {
let name = format!("{}", i);
match cache.get(&name).expect("get must succeed") {
Some(mut file) => {
use std::io::Read;
let mut buf = Vec::new();
file.read_to_end(&mut buf).expect("read must succeed");
assert_eq!(buf, format!("{}", PAYLOAD_MULTIPLIER * i).into_bytes());
1
}
None => 0,
}
})
.sum();
assert!(present >= 10);
assert!(present < 20);
// The temporary garbage file must have been deleted by now.
assert!(
matches!(std::fs::metadata(temp.path(&format!("{}/garbage", TEMP_SUBDIR))),
Err(e) if e.kind() == std::io::ErrorKind::NotFound)
);
}
/// Publish a file, make sure we can read it, then overwrite, and
/// confirm that the new contents are visible.
#[test]
fn test_set() {
use std::io::{Read, Write};
use tempfile::NamedTempFile;
use test_dir::{DirBuilder, TestDir};
let temp = TestDir::temp();
let cache = Cache::new(temp.path("."), 1);
{
let tmp = NamedTempFile::new_in(cache.temp_dir().expect("temp_dir must succeed"))
.expect("new temp file must succeed");
tmp.as_file().write_all(b"v1").expect("write must succeed");
cache
.set("entry", tmp.path())
.expect("initial set must succeed");
}
{
let mut cached = cache
.get("entry")
.expect("must succeed")
.expect("must be found");
let mut dst = Vec::new();
cached.read_to_end(&mut dst).expect("read must succeed");
assert_eq!(&dst, b"v1");
}
// Now overwrite; it should take.
{
let tmp = NamedTempFile::new_in(cache.temp_dir().expect("temp_dir must succeed"))
.expect("new temp file must succeed");
tmp.as_file().write_all(b"v2").expect("write must succeed");
cache
.set("entry", tmp.path())
.expect("overwrite must succeed");
}
{
let mut cached = cache
.get("entry")
.expect("must succeed")
.expect("must be found");
let mut dst = Vec::new();
cached.read_to_end(&mut dst).expect("read must succeed");
assert_eq!(&dst, b"v2");
}
}
/// Publish a file, make sure we can read it, and make sure that a
/// second put does not update its contents.
#[test]
fn test_put() {
use std::io::{Read, Write};
use tempfile::NamedTempFile;
use test_dir::{DirBuilder, TestDir};
let temp = TestDir::temp();
let cache = Cache::new(temp.path("."), 1);
{
let tmp = NamedTempFile::new_in(cache.temp_dir().expect("temp_dir must succeed"))
.expect("new temp file must succeed");
tmp.as_file().write_all(b"v1").expect("write must succeed");
cache
.put("entry", tmp.path())
.expect("initial set must succeed");
}
{
let mut cached = cache
.get("entry")
.expect("must succeed")
.expect("must be found");
let mut dst = Vec::new();
cached.read_to_end(&mut dst).expect("read must succeed");
assert_eq!(&dst, b"v1");
}
// Now put again; it shouldn't overwrite.
{
let tmp = NamedTempFile::new_in(cache.temp_dir().expect("temp_dir must succeed"))
.expect("new temp file must succeed");
tmp.as_file().write_all(b"v2").expect("write must succeed");
cache
.put("entry", tmp.path())
.expect("overwrite must succeed");
}
{
let mut cached = cache
.get("entry")
.expect("must succeed")
.expect("must be found");
let mut dst = Vec::new();
cached.read_to_end(&mut dst).expect("read must succeed");
assert_eq!(&dst, b"v1");
}
}
/// Keep publishing new files, but also always touch the first.
/// That first file should never be deleted.
#[test]
fn test_touch() {
use tempfile::NamedTempFile;
use test_dir::{DirBuilder, TestDir};
let temp = TestDir::temp();
let cache = Cache::new(temp.path("."), 5);
for i in 0..15 {
let name = format!("{}", i);
// After the first write, touch should find our file.
assert_eq!(cache.touch("0").expect("touch must not fail"), i > 0);
let tmp = NamedTempFile::new_in(cache.temp_dir().expect("temp_dir must succeed"))
.expect("new temp file must succeed");
cache.put(&name, tmp.path()).expect("put must succeed");
// Make sure enough time elapses for the first file to get
// an older timestamp than the rest.
if i == 0 {
std::thread::sleep(std::time::Duration::from_secs_f64(1.5));
}
}
// We should still find "0": it's the oldest, but we also keep
// touching it.
cache.get("0").expect("must succed").expect("must be found");
}
/// Trigger a cleanup while a very recent file is still in the
/// temporary subdirectory. It should remain there.
#[test]
fn test_recent_temp_file() {
use tempfile::NamedTempFile;
use test_dir::{DirBuilder, FileType, TestDir};
// Also leave a file in the temporary subdirectory; we'll check
// that it gets cleaned up before leaving this function..
let temp = TestDir::temp()
.create(TEMP_SUBDIR, FileType::Dir)
.create(&format!("{}/garbage", TEMP_SUBDIR), FileType::ZeroFile(10));
// The garbage file must exist.
assert!(std::fs::metadata(temp.path(&format!("{}/garbage", TEMP_SUBDIR))).is_ok());
let cache = Cache::new(temp.path("."), 1);
for i in 0..2 {
let tmp = NamedTempFile::new_in(cache.temp_dir().expect("temp_dir must succeed"))
.expect("new temp file must succeed");
cache
.put(&format!("{}", i), tmp.path())
.expect("put must succeed");
}
// The garbage file must still exist.
assert!(std::fs::metadata(temp.path(&format!("{}/garbage", TEMP_SUBDIR))).is_ok());
}