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
//! rkyv (and generic binary) structural inspector.
//!
//! rkyv archives are not self-describing — the format stores no field names or
//! type tags (<https://rkyv.org/format.html>), so a generic reader cannot recover
//! the schema. What it CAN do without the Rust type is show the raw structure:
//! a hex/ascii dump and the runs of printable text embedded in the archive
//! (strings, keys, interned identifiers). Typed CRUD would require a supplied
//! schema descriptor and is deferred.
use anyhow::{Context, Result};
use std::path::{Path, PathBuf};
use std::sync::Arc;
pub struct RkyvStore {
pub path: PathBuf,
/// The whole archive. Shared rather than owned: every edit path and the
/// background decoder need the bytes at the same time, and a 382 MB shard
/// copied per operation is 382 MB per keystroke.
pub bytes: Arc<[u8]>,
}
/// Most printable runs collected from one archive. Past this the list is no
/// longer something a person reads, and every entry costs a heap allocation.
const MAX_STRING_HITS: usize = 20_000;
/// Bytes scanned for printable runs. A 382 MB shard took 4.2 s to scan in full,
/// which is a stall the picker's file open cannot afford.
const STRINGS_SCAN_CAP: usize = 64 * 1024 * 1024;
/// The result of a bounded string extraction.
#[derive(Debug, Default)]
pub struct Strings {
pub hits: Vec<StringHit>,
/// A bound was hit, so this is not every run in the file.
pub truncated: bool,
/// How many bytes were scanned.
pub scanned: usize,
}
/// A run of printable ASCII found in the archive, with its byte offset.
#[derive(Debug)]
pub struct StringHit {
pub offset: usize,
pub text: String,
}
impl RkyvStore {
pub fn open(path: &Path) -> Result<Self> {
let bytes = std::fs::read(path).with_context(|| format!("read {}", path.display()))?;
Ok(Self {
path: path.to_path_buf(),
bytes: bytes.into(),
})
}
pub fn len(&self) -> usize {
self.bytes.len()
}
/// Extract runs of printable ASCII of at least `min_len` bytes, bounded so a
/// huge archive cannot stall the UI: at most [`MAX_STRING_HITS`] runs from the
/// first [`STRINGS_SCAN_CAP`] bytes. `truncated` says whether either bound
/// was hit, which the Strings view reports.
pub fn strings(&self, min_len: usize) -> Strings {
let scan_end = self.bytes.len().min(STRINGS_SCAN_CAP);
let mut hits = Vec::new();
let mut start: Option<usize> = None;
for (i, &b) in self.bytes[..scan_end].iter().enumerate() {
if hits.len() >= MAX_STRING_HITS {
break;
}
let printable = (0x20..0x7f).contains(&b);
match (printable, start) {
(true, None) => start = Some(i),
(false, Some(s)) => {
if i - s >= min_len {
hits.push(StringHit {
offset: s,
text: String::from_utf8_lossy(&self.bytes[s..i]).into_owned(),
});
}
start = None;
}
_ => {}
}
}
if let Some(s) = start {
if hits.len() < MAX_STRING_HITS && scan_end - s >= min_len {
hits.push(StringHit {
offset: s,
text: String::from_utf8_lossy(&self.bytes[s..scan_end]).into_owned(),
});
}
}
let truncated = hits.len() >= MAX_STRING_HITS || scan_end < self.bytes.len();
Strings {
hits,
truncated,
scanned: scan_end,
}
}
/// One 16-byte `offset hex bytes |ascii|` line, `xxd` style. Formatting
/// lives in `crate::hexedit` so every hex view in the app shares one layout.
pub fn hex_row(&self, offset: usize) -> String {
let end = (offset + 16).min(self.bytes.len());
let chunk = &self.bytes[offset.min(self.bytes.len())..end];
crate::hexedit::hex_dump_line(offset, chunk)
}
}
#[cfg(test)]
mod tests {
use super::{RkyvStore, MAX_STRING_HITS};
use std::sync::Arc;
fn store(bytes: &[u8]) -> RkyvStore {
RkyvStore {
path: std::path::PathBuf::from("/tmp/in-memory"),
bytes: Arc::from(bytes.to_vec()),
}
}
#[test]
fn runs_shorter_than_the_minimum_are_not_hits() {
// "ab" (2) then "hello" (5), separated by bytes that are not printable.
let s = store(b"\x00ab\x00hello\x00");
let hits = s.strings(3).hits;
assert_eq!(hits.len(), 1, "only the run that is long enough");
assert_eq!(hits[0].text, "hello");
assert_eq!(hits[0].offset, 4, "the offset is where the run starts");
// Lower the bar and both runs count.
assert_eq!(s.strings(2).hits.len(), 2);
}
#[test]
fn a_run_that_reaches_the_end_of_the_file_is_still_a_hit() {
// The scan closes a run when it meets a non-printable byte; a file that
// ends mid-run has to be closed by the end of the file instead.
let s = store(b"\x00trailing");
let hits = s.strings(4).hits;
assert_eq!(hits.len(), 1);
assert_eq!(hits[0].text, "trailing");
assert_eq!(hits[0].offset, 1);
}
#[test]
fn printable_is_space_through_tilde() {
// The boundaries either side: 0x1f and 0x7f are not text, 0x20 and 0x7e
// are. A tab and a newline are separators here, not characters.
let s = store(b"\x1f\x20\x7e\x7f\x00a\tb\nc");
let hits = s.strings(1).hits;
let texts: Vec<&str> = hits.iter().map(|h| h.text.as_str()).collect();
assert_eq!(texts, vec![" ~", "a", "b", "c"]);
}
#[test]
fn too_many_runs_stops_the_scan_and_says_so() {
// One more run than the list holds, each two bytes with a separator.
let mut bytes = Vec::new();
for _ in 0..MAX_STRING_HITS + 1 {
bytes.extend_from_slice(b"ab\x00");
}
let s = store(&bytes);
let got = s.strings(2);
assert_eq!(got.hits.len(), MAX_STRING_HITS, "capped");
assert!(got.truncated, "and the view is told the list is partial");
// A file whose runs all fit is not reported as truncated.
let small = store(b"\x00hello\x00world\x00");
let got = small.strings(3);
assert_eq!(got.hits.len(), 2);
assert!(!got.truncated);
assert_eq!(got.scanned, small.len());
}
#[test]
fn a_hex_row_stops_at_the_end_of_the_file() {
let s = store(b"0123456789");
// A full row's worth is asked for; only what exists is rendered.
let row = s.hex_row(0);
assert!(row.contains("30 31 32 33"), "{row}");
assert!(row.contains("0123456789"), "{row}");
// Past the end: no panic, and nothing invented.
let past = s.hex_row(64);
assert!(!past.contains("30"), "{past}");
}
}