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
//! `run_workspace_grep` helper — kept in its own file so `helpers.rs` stays under the
//! 1000-line cap as the MCP surface grows.
use std::sync::atomic::Ordering;
use rmcp::ErrorData as McpError;
use rmcp::model::CallToolResult;
use super::ServerState;
use super::helpers::{SEARCH_LIMIT_DEFAULT, SEARCH_LIMIT_MAX};
use super::types::{GrepHit, WorkspaceGrepParams, WorkspaceGrepResponse};
/// Body of the `workspace_grep` MCP tool.
///
/// Iterates over indexed files (bounded by `scan_cap = limit * 8`), reads each as UTF-8, and
/// applies the compiled regex. Non-UTF-8 files are silently skipped. Returns up to `limit` hits
/// with optional 1-line context. Supports in-memory pagination via `cursor` / `next_cursor`
/// using the same `encode_in_memory(offset, generation)` scheme as `list_files`.
pub(super) fn run_workspace_grep(
state: &ServerState,
params: WorkspaceGrepParams,
) -> Result<CallToolResult, McpError> {
let format = super::toon::ResponseFormat::parse(params.format.as_deref());
let limit = params
.limit
.unwrap_or(SEARCH_LIMIT_DEFAULT)
.min(SEARCH_LIMIT_MAX) as usize;
let scan_cap = limit.saturating_mul(8).max(2_000);
let generation = state.cache_generation.load(Ordering::Relaxed);
// Decode cursor and check snapshot id. Stale cursor → bail with empty page +
// cursor_invalidated=true so the caller can restart.
let skip: usize = match params.cursor.as_ref() {
Some(c) => {
let (offset, snapshot_id) = c.decode_in_memory()?;
if snapshot_id != generation {
return super::toon::format_result(
&WorkspaceGrepResponse {
pattern: params.pattern,
total_files_matched: 0,
total_matches: 0,
truncated: false,
budgeted: false,
hits: Vec::new(),
next_cursor: None,
cursor_invalidated: true,
},
format,
);
}
offset as usize
}
None => 0,
};
let re = regex::Regex::new(¶ms.pattern)
.map_err(|e| McpError::invalid_params(format!("invalid regex: {e}"), None))?;
let path_finder = params
.path_contains
.as_deref()
.map(|n| memchr::memmem::Finder::new(n.as_bytes()));
let lang_filter = params.language.as_deref();
let cache = state.cache.load_full();
let mut hits: Vec<GrepHit> = Vec::with_capacity(limit.min(64));
// Parallel to `hits`: the 1-based `files_seen` index of the file that produced each hit.
// Lets a token budget re-anchor the cursor to the file of the last KEPT hit so no hit is
// permanently lost when the budget cuts mid-file (that boundary file is re-scanned).
let mut hit_file_idx: Vec<usize> = Vec::with_capacity(limit.min(64));
let mut total_matches: u32 = 0;
let mut total_files_matched: usize = 0;
let mut truncated = false;
let mut files_visited: usize = 0;
// Track how many files passed filters so far (for cursor offset).
let mut files_seen: usize = 0;
'files: for (path, entry) in &cache.by_path {
// Honour scan_cap: stop iterating files once we've visited enough.
if files_visited >= scan_cap {
truncated = true;
break;
}
// Path filter (memchr).
let path_ok = path_finder
.as_ref()
.is_none_or(|f| f.find(path.as_bytes()).is_some());
if !path_ok {
continue;
}
// Language filter.
let lang_ok = lang_filter.is_none_or(|l| entry.language == l);
if !lang_ok {
continue;
}
// Skip files that were already returned on prior pages.
if files_seen < skip {
files_seen += 1;
continue;
}
files_seen += 1;
files_visited += 1;
// Read the file from the working tree; skip non-UTF-8 silently.
let abs = state.root.join(path.to_path_buf());
let source = match std::fs::read_to_string(&abs) {
Ok(s) => s,
Err(e) => {
tracing::debug!(path = %abs.display(), error = %e, "workspace_grep: skipping unreadable file");
continue;
}
};
// Pre-build line offset table so match→line/col is O(hits), not O(file).
// `line_starts[i]` = byte offset of the start of line `i` (0-based).
let line_starts: Vec<usize> = std::iter::once(0)
.chain(memchr::memchr_iter(b'\n', source.as_bytes()).map(|pos| pos + 1))
.collect();
let mut file_had_match = false;
for mat in re.find_iter(&source) {
total_matches = total_matches.saturating_add(1);
file_had_match = true;
if hits.len() >= limit {
// Keep counting total_matches, but stop materialising hits.
// Once limit is saturated, set truncated and stop processing files.
truncated = true;
break 'files;
}
// Binary search for the line that contains the match start.
let match_start = mat.start();
let line_idx = line_starts
.partition_point(|&ls| ls <= match_start)
.saturating_sub(1);
let line_start_byte = line_starts[line_idx];
let line_num = (line_idx as u32) + 1; // 1-based
let column = (match_start - line_start_byte) as u32; // 0-based byte col
// Context: extract line-string helper (strip trailing '\n'/'\r\n').
let context_before = if params.include_context && line_idx > 0 {
Some(extract_line(&source, &line_starts, line_idx - 1))
} else {
None
};
let context_after = if params.include_context && line_idx + 1 < line_starts.len() {
Some(extract_line(&source, &line_starts, line_idx + 1))
} else {
None
};
hits.push(GrepHit {
path: path.clone(),
line_num,
column,
matched_text: mat.as_str().to_owned(),
context_before,
context_after,
});
hit_file_idx.push(files_seen);
}
if file_had_match {
total_files_matched += 1;
}
}
// Emit a cursor when hits are saturated and there may be more files to visit.
// `files_seen` is the position past the last file we processed; the next page
// skips all files before that index.
let next_cursor = if truncated {
Some(super::cursor::Cursor::encode_in_memory(
files_seen as u64,
generation,
))
} else {
None
};
// Apply the token budget over the materialised hits. When it drops trailing hits,
// re-anchor the cursor to RE-SCAN the file of the last kept hit (offset = that file's
// index minus one) so no hit is permanently lost — the boundary file may re-emit a few
// already-returned hits, which is the safe trade-off for a file-granular cursor.
let budget = super::budget::apply_budget(hits, params.max_tokens);
let (hits, budgeted, next_cursor) = if budget.budgeted {
let kept = budget.items.len();
let resume_offset = hit_file_idx[kept - 1].saturating_sub(1);
(
budget.items,
true,
Some(super::cursor::Cursor::encode_in_memory(
resume_offset as u64,
generation,
)),
)
} else {
(budget.items, false, next_cursor)
};
super::toon::format_result(
&WorkspaceGrepResponse {
pattern: params.pattern,
total_files_matched,
total_matches,
truncated,
budgeted,
hits,
next_cursor,
cursor_invalidated: false,
},
format,
)
}
/// Extract the content of line `line_idx` from `source`, stripping the trailing
/// `\n` / `\r\n`. Returns an empty string when the line is empty or the index is
/// out of range.
fn extract_line(source: &str, line_starts: &[usize], line_idx: usize) -> String {
let start = line_starts[line_idx];
let end = line_starts
.get(line_idx + 1)
.copied()
.unwrap_or(source.len());
// `end` points past the '\n'; trim trailing CR+LF.
let raw = &source[start..end];
raw.trim_end_matches('\n').trim_end_matches('\r').to_owned()
}