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
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
//! Tree-sitter based Markdown source code extractor.
//!
//! Uses `tree-sitter-grammars/tree-sitter-markdown` (a split grammar):
//!
//! * `block::LANGUAGE` parses block-level structure (sections, headings,
//! paragraphs, lists). Inline content is left as opaque `(inline)` nodes.
//! * `inline::LANGUAGE` is run over each `(inline)` node's byte range to
//! produce links, emphasis, etc. We use `Parser::set_included_ranges`
//! so the inline tree's byte/row positions stay in the original source.
//!
//! `atx_heading` / `setext_heading` nodes become `Module` nodes; `inline_link`
//! nodes whose destination is a project-local source file emit `Uses` edges.
//! Frontmatter (`(minus_metadata)`, `(plus_metadata)`) is skipped — the
//! grammar makes it opaque, so we don't recurse into it.
use std::time::{Instant, SystemTime, UNIX_EPOCH};
use tree_sitter::{Node as TsNode, Parser, Range, Tree};
use crate::types::{
generate_node_id, Edge, EdgeKind, ExtractionResult, Node, NodeKind, Visibility,
};
pub struct MarkdownExtractor;
struct ExtractionState {
nodes: Vec<Node>,
edges: Vec<Edge>,
file_path: String,
source: Vec<u8>,
timestamp: u64,
/// (heading title, node id, level) — heading levels strictly increase
/// going *down* the stack. Headings of equal or shallower level pop
/// the stack so we always parent to the nearest ancestor heading.
node_stack: Vec<(String, String, usize)>,
/// One inline parser, lazily initialised, reused for every `(inline)`
/// node we encounter to avoid re-creating it per heading/paragraph.
inline_parser: Option<Parser>,
}
impl ExtractionState {
fn new(file_path: &str, source: &str) -> Self {
let timestamp = SystemTime::now()
.duration_since(UNIX_EPOCH)
.unwrap_or_default()
.as_secs();
Self {
nodes: Vec::new(),
edges: Vec::new(),
file_path: file_path.to_string(),
source: source.as_bytes().to_vec(),
timestamp,
node_stack: Vec::new(),
inline_parser: None,
}
}
fn node_text(&self, node: TsNode<'_>) -> String {
node.utf8_text(&self.source)
.unwrap_or("<invalid utf8>")
.to_string()
}
/// Parse the byte range covered by `inline_node` with the inline grammar
/// and return the resulting tree, or `None` if parsing fails. The tree's
/// byte/row positions are anchored in the original source via
/// `set_included_ranges`.
fn parse_inline(&mut self, inline_node: TsNode<'_>) -> Option<Tree> {
let parser = self.inline_parser.get_or_insert_with(|| {
let mut p = Parser::new();
let _ = p.set_language(&tokensave_large_treesitters::markdown::inline::LANGUAGE.into());
p
});
let range = Range {
start_byte: inline_node.start_byte(),
end_byte: inline_node.end_byte(),
start_point: inline_node.start_position(),
end_point: inline_node.end_position(),
};
if parser.set_included_ranges(&[range]).is_err() {
return None;
}
parser.parse(&self.source, None)
}
}
impl MarkdownExtractor {
pub fn extract_markdown(file_path: &str, source: &str) -> ExtractionResult {
let start = Instant::now();
let mut state = ExtractionState::new(file_path, source);
let file_node = Node {
id: generate_node_id(file_path, &NodeKind::File, file_path, 0),
kind: NodeKind::File,
name: file_path.to_string(),
qualified_name: file_path.to_string(),
file_path: file_path.to_string(),
start_line: 0,
attrs_start_line: 0,
end_line: source.lines().count().saturating_sub(1) as u32,
start_column: 0,
end_column: 0,
signature: None,
docstring: None,
visibility: Visibility::Pub,
is_async: false,
branches: 0,
loops: 0,
returns: 0,
max_nesting: 0,
unsafe_blocks: 0,
unchecked_calls: 0,
assertions: 0,
updated_at: state.timestamp,
};
let file_node_id = file_node.id.clone();
state.nodes.push(file_node);
state
.node_stack
.push((file_path.to_string(), file_node_id, 0));
if let Ok(tree) = Self::parse(source) {
let root = tree.root_node();
Self::visit(&mut state, root);
}
ExtractionResult {
nodes: state.nodes,
edges: state.edges,
unresolved_refs: Vec::new(),
errors: Vec::new(),
duration_ms: start.elapsed().as_millis() as u64,
}
}
fn parse(source: &str) -> Result<Tree, String> {
let mut parser = Parser::new();
parser
.set_language(&tokensave_large_treesitters::markdown::LANGUAGE.into())
.map_err(|e| format!("failed to load markdown grammar: {e}"))?;
parser
.parse(source, None)
.ok_or_else(|| "tree-sitter parse returned None".to_string())
}
/// Walks the block tree. `atx_heading` / `setext_heading` produce `Module`
/// nodes; `(inline)` nodes are re-parsed with the inline grammar to find
/// links. Frontmatter (`(minus_metadata)`, `(plus_metadata)`) is opaque
/// per the grammar — we never descend into it.
fn visit(state: &mut ExtractionState, node: TsNode<'_>) {
match node.kind() {
"atx_heading" | "setext_heading" => Self::visit_heading(state, node),
"minus_metadata" | "plus_metadata" => {
// Opaque YAML/TOML frontmatter — don't descend.
}
"inline" => Self::visit_inline(state, node),
_ => Self::visit_children(state, node),
}
}
fn visit_children(state: &mut ExtractionState, node: TsNode<'_>) {
let mut cursor = node.walk();
if cursor.goto_first_child() {
loop {
Self::visit(state, cursor.node());
if !cursor.goto_next_sibling() {
break;
}
}
}
}
fn visit_heading(state: &mut ExtractionState, node: TsNode<'_>) {
let level = Self::heading_level(node);
// Block grammar exposes the heading text as the `heading_content`
// field, which points to an `(inline)` node containing the text.
let title = node
.child_by_field_name("heading_content")
.map(|n| state.node_text(n).trim().to_string())
.unwrap_or_default();
if title.is_empty() {
// Still recurse into the heading body so any `(inline)` children
// (which would be unusual but not impossible) get their links.
Self::visit_children(state, node);
return;
}
while state.node_stack.len() > 1 {
let last_level = state.node_stack[state.node_stack.len() - 1].2;
if last_level >= level {
state.node_stack.pop();
} else {
break;
}
}
let kind = NodeKind::Module;
let parent_name = &state.node_stack[state.node_stack.len() - 1].0;
let qualified_name = format!("{parent_name}::{title}");
let id = generate_node_id(
&state.file_path,
&kind,
&title,
node.start_position().row as u32,
);
let node_obj = Node {
id: id.clone(),
kind,
name: title.clone(),
qualified_name,
file_path: state.file_path.clone(),
start_line: node.start_position().row as u32,
attrs_start_line: node.start_position().row as u32,
end_line: node.end_position().row as u32,
start_column: node.start_position().column as u32,
end_column: node.end_position().column as u32,
signature: Some(format!("{} {}", "#".repeat(level), title)),
docstring: None,
visibility: Visibility::Pub,
is_async: false,
branches: 0,
loops: 0,
returns: 0,
max_nesting: 0,
unsafe_blocks: 0,
unchecked_calls: 0,
assertions: 0,
updated_at: state.timestamp,
};
if let Some((_, parent_id, _)) = state.node_stack.last() {
state.edges.push(Edge {
source: parent_id.clone(),
target: id.clone(),
kind: EdgeKind::Contains,
line: Some(node.start_position().row as u32),
});
}
state.nodes.push(node_obj);
state.node_stack.push((title, id, level));
// Recurse so links inside the heading text (e.g.
// `## See [main](src/main.rs)`) become `Uses` edges parented to
// this heading.
Self::visit_children(state, node);
}
/// Returns the ATX heading level (1-6) for `atx_heading` / `setext_heading`.
/// ATX uses `atx_h{1..6}_marker` children; setext uses level-1 (`===`) or
/// level-2 (`---`) underlines, identified by the marker child kind.
fn heading_level(node: TsNode<'_>) -> usize {
for child in node.children(&mut node.walk()) {
let k = child.kind();
if let Some(rest) = k.strip_prefix("atx_h") {
if let Some(d) = rest.strip_suffix("_marker") {
if let Ok(n) = d.parse::<usize>() {
return n.clamp(1, 6);
}
}
}
if k == "setext_h1_underline" {
return 1;
}
if k == "setext_h2_underline" {
return 2;
}
}
1
}
/// Re-parse the `(inline)` node's byte range with the inline grammar and
/// collect any `inline_link` nodes as `Uses` edges.
fn visit_inline(state: &mut ExtractionState, inline_node: TsNode<'_>) {
let Some(inline_tree) = state.parse_inline(inline_node) else {
return;
};
let root = inline_tree.root_node();
Self::collect_links(state, root);
}
fn collect_links(state: &mut ExtractionState, node: TsNode<'_>) {
if node.kind() == "inline_link" || node.kind() == "image" {
Self::visit_link(state, node);
// Inline links can nest inside images and vice versa; keep walking.
}
let mut cursor = node.walk();
if cursor.goto_first_child() {
loop {
Self::collect_links(state, cursor.node());
if !cursor.goto_next_sibling() {
break;
}
}
}
}
fn visit_link(state: &mut ExtractionState, node: TsNode<'_>) {
let Some(url_node) = node
.children(&mut node.walk())
.find(|n| n.kind() == "link_destination")
else {
return;
};
let url = state.node_text(url_node);
if url.starts_with("http://") || url.starts_with("https://") {
return;
}
let target_path = url.trim_start_matches("file:");
let target_ext = target_path.rsplit('.').next().unwrap_or("");
if !is_code_extension(target_ext) {
return;
}
let target_id = generate_node_id(target_path, &NodeKind::File, target_path, 0);
if let Some((_, parent_id, _)) = state.node_stack.last() {
state.edges.push(Edge {
source: parent_id.clone(),
target: target_id,
kind: EdgeKind::Uses,
line: Some(node.start_position().row as u32),
});
}
}
}
fn is_code_extension(ext: &str) -> bool {
// Only include actual programming-language source files.
// Config (yaml, toml, json), markup (html, css, markdown), and
// notebook (ipynb) files are excluded to avoid low-signal edges.
matches!(
ext,
"rs" | "py"
| "js"
| "ts"
| "tsx"
| "jsx"
| "go"
| "java"
| "c"
| "cpp"
| "h"
| "hpp"
| "cs"
| "rb"
| "php"
| "swift"
| "kt"
| "scala"
| "R"
| "sh"
| "bash"
| "zsh"
| "fish"
| "ps1"
| "ex"
| "exs"
| "erl"
| "hrl"
| "fs"
| "fsx"
| "ml"
| "mli"
| "hs"
| "lhs"
| "lua"
| "pl"
| "pm"
| "t"
| "nix"
| "sql"
| "proto"
| "v"
| "vhd"
| "vhdl"
| "sage"
| "sagews"
)
}
impl crate::extraction::LanguageExtractor for MarkdownExtractor {
fn extensions(&self) -> &[&str] {
&["md", "markdown"]
}
fn language_name(&self) -> &'static str {
"Markdown"
}
fn extract(&self, file_path: &str, source: &str) -> ExtractionResult {
Self::extract_markdown(file_path, source)
}
}