rto_render/obsidian.rs
1//! The Obsidian-vault renderer: each graph node becomes a markdown note whose
2//! edges are `[[wikilinks]]`, so the provenance-tagged graph is browsable in
3//! Obsidian's graph view. Notes carry frontmatter `tags` (`roteiro/kind/*`,
4//! `roteiro/lang/*`, `roteiro/status/*`) so the graph is colourable/filterable —
5//! edge provenance is shown per-link in the body — surface the node's text as the
6//! knowledge base, show an ADR's status, and (when the repository's web host is
7//! known) a clickable **Source** link to the file.
8//!
9//! That text is the node's captured `meta.content` (a doc comment, PDF or image
10//! text) *except* where the caller supplies a full `body` — which it does for
11//! prose documents, because `meta.content` is an embedding budget and a note
12//! rendered from it is the document capped at 1500 characters and collapsed onto
13//! one line. See [`note_body`].
14//!
15//! A generated `_Home` note is the overview: what was
16//! scanned, counts by kind, provenance breakdown, ADR statuses, intent-debt (with
17//! the files it is densest in), an inventory of secret-**named** config keys and
18//! their redaction state, and the most depended-on symbols by directed call
19//! fan-in.
20//! Built from the same [`Explanation`] the query surface returns, so the vault
21//! and the CLI agree.
22
23use std::fmt::Write as _;
24
25use rto_graph::Explanation;
26
27/// Filename of the generated overview note (sorts first in the file list).
28pub const HOME_NOTE: &str = "_Home.md";
29
30/// A rendered vault note: its filename (with `.md`) and markdown content.
31#[derive(Debug, Clone, PartialEq, Eq)]
32pub struct VaultNote {
33 /// Filename including the `.md` extension.
34 pub filename: String,
35 /// Markdown content.
36 pub content: String,
37}
38
39/// Map a node key to a filesystem- and wikilink-safe note stem that is **unique
40/// per key even after case folding**.
41///
42/// A name is a lowercased, readable *hint* slugged from the key, followed by an
43/// unconditional 64-bit FNV-1a hash of the whole, exact key written as 16 hex
44/// digits — `<hint>-<16 hex digits>`. Characters outside `[a-z0-9._-]` collapse
45/// to a single `-` in the hint; the hash carries everything the hint threw away.
46///
47/// **The hash is the unconditional part, not the hint.** A key of nothing but
48/// separators slugs to an empty hint, and the name is then the bare 16 hex
49/// digits — no hint, and no `-` to join it to. The two forms cannot be confused
50/// for one another, which is what makes the exception safe rather than a second
51/// naming rule: a hinted name is at least 18 characters and contains a `-`,
52/// and a bare one is exactly 16 and contains none. That is argued again at the
53/// branch itself, and asserted by
54/// `every_name_carries_the_hash_however_short_the_key`.
55///
56/// # Why the hash is unconditional (issue #574)
57///
58/// It used to be applied only when the slug overran the filename limit, and the
59/// slug alone was lossy twice over. Measured on this repository — 8,239 nodes
60/// rendering to 8,135 notes, 104 of them silently overwritten:
61///
62/// | mechanism | lost | where |
63/// | --- | --- | --- |
64/// | every character outside the safe set becomes `-` and runs collapse, so `…cytoscape.min.js#$a` and `…cytoscape.min.js#a` are one name | 9 | everywhere |
65/// | macOS and Windows fold filename case, so `…#A` and `…#a` are two *names* but one *file* | 95 | macOS, Windows |
66///
67/// The second mechanism is the trap. A lossless-but-case-sensitive encoding
68/// fixes the 9, verifies clean on Linux CI, and still loses 95 notes on a Mac.
69/// So the requirement is stated after folding:
70///
71/// ```text
72/// lower(note_name(k1)) == lower(note_name(k2)) implies k1 == k2
73/// ```
74///
75/// This matters more than lossiness in a cache would, because the note names are
76/// the vault's **only** stable interface: `reset_vault_dir` deletes and rebuilds
77/// the whole directory on every render, so the one thing that survives a render
78/// is a user's own note *outside* the vault linking in by name (issue #442).
79///
80/// # The trade taken
81///
82/// Two decisions, and what each bought:
83///
84/// **The hint is lowercased rather than case-preserved.** Case-preserving would
85/// also satisfy the requirement — the hash differs for `#A` and `#a`, so the two
86/// names differ in their suffix and stay distinct under folding. It was rejected
87/// because lowercasing makes `note_name(k) == note_name(k).to_lowercase()` an
88/// invariant of the function, and *that* collapses the folded property into the
89/// literal one: there is then no way to write a version of this that is green on
90/// Linux and lossy on macOS, which is the defect shape this repository keeps
91/// finding. The cost is that `parseHTTPHeader` reads as `parsehttpheader`. That
92/// is affordable precisely because the hint is a hint — once a 17-character
93/// suffix is mandatory the name is not something anyone types from memory, so
94/// its job is to be recognisable in a file list, not to be transcribed.
95///
96/// **Readability was spent, deliberately.** Every name grows by 17 characters and
97/// hand-writing a link now needs Obsidian's autocomplete. The alternatives that
98/// keep names short — hashing only the keys observed to collide — make the *set*
99/// of collisions platform-dependent, so one key would get one filename on macOS
100/// and another on Linux and a synced vault would churn. A name that is uglier
101/// everywhere beats a name that is different per platform.
102///
103/// The mapping is not reversible (the hint is lossy and the hash is one-way), but
104/// it does not need to be: every note's frontmatter carries `key:` verbatim, so
105/// name → key is recoverable from the vault itself, which is the direction a
106/// reader actually needs.
107///
108/// # What "unique" rests on
109///
110/// Equal names imply equal hashes, not equal keys — this is a 64-bit hash, not a
111/// proof. Over this repository's 8,239 keys there is no collision, and the
112/// birthday bound at that size is about 2e-12. Should one ever occur it is
113/// *reported*, not silent: `NoteNames` in the render path claims every filename
114/// case-insensitively and warns on a repeat. What is proved outright is the
115/// folding half — the output is lowercase by construction, so case folding is the
116/// identity on it.
117#[must_use]
118pub fn note_name(key: &str) -> String {
119 // Keep the whole stem well under the 255-byte filename limit (leaving room
120 // for ".md"). The hint is ASCII, so byte length equals char count and slicing
121 // is safe.
122 const MAX: usize = 200;
123 // '-' plus the 16 hex digits of the hash.
124 const SUFFIX: usize = 17;
125 const HINT: usize = MAX - SUFFIX;
126
127 let mut hint = String::with_capacity(key.len());
128 let mut prev_dash = false;
129 for c in key.chars() {
130 if c.is_ascii_alphanumeric() || matches!(c, '.' | '_' | '-') {
131 hint.push(c.to_ascii_lowercase());
132 prev_dash = false;
133 } else if !prev_dash {
134 hint.push('-');
135 prev_dash = true;
136 }
137 }
138 let hint = hint.trim_matches('-');
139 // Truncation is only ever cosmetic now: the hash, not the hint, is what keeps
140 // a 300-character grouped `use` distinct from its neighbour.
141 let hint = hint[..hint.len().min(HINT)].trim_end_matches('-');
142 let hash = fnv1a64(key.as_bytes());
143 if hint.is_empty() {
144 // A key of nothing but separators. Bare hex, and it cannot be confused
145 // with a hinted name: those are `<hint>-<16 hex>`, so at least 18
146 // characters, and this is exactly 16 with no `-` in it.
147 format!("{hash:016x}")
148 } else {
149 format!("{hint}-{hash:016x}")
150 }
151}
152
153/// FNV-1a (64-bit) — a dependency-free, deterministic hash carrying everything
154/// [`note_name`]'s hint discards. No cryptographic properties needed: nothing
155/// here defends against a chosen collision, only against an accidental one.
156///
157/// 64 bits rather than fewer because the cost of a collision is exactly the
158/// defect this suffix exists to fix — a note silently overwritten. At 8k keys a
159/// 32-bit hash collides about 0.8% of the time and a 48-bit one about 1e-5;
160/// 64 bits is 2e-12, and stays under 1e-10 for a workspace vault an order of
161/// magnitude larger.
162fn fnv1a64(bytes: &[u8]) -> u64 {
163 let mut hash: u64 = 0xcbf2_9ce4_8422_2325;
164 for &b in bytes {
165 hash ^= u64::from(b);
166 hash = hash.wrapping_mul(0x0000_0100_0000_01b3);
167 }
168 hash
169}
170
171/// Emit `value` as a YAML **double-quoted** scalar, `"`-delimited and escaped so
172/// it parses back to exactly `value`.
173///
174/// The one escaping rule for this module's frontmatter. It exists because the
175/// three hand-rolled variants it replaced disagreed with each other — `key:` and
176/// `project:` turned a `"` into an apostrophe, and `path:` escaped nothing — and
177/// two of the three could emit YAML that does not mean what it says:
178///
179/// | value | was emitted | parsed back as |
180/// | --- | --- | --- |
181/// | `foo\bar` | `"foo\bar"` | `foo<BS>ar` — `\b` is YAML's **backspace** escape |
182/// | `foo\dir` | `"foo\dir"` | *parse error* — `\d` is not a YAML escape |
183/// | `say"hi".rs` | `"say"hi".rs"` | *parse error* — the scalar ends at the `"` |
184///
185/// The first is the dangerous one: seven characters silently become six, and
186/// nothing anywhere reports it. The other two cost the reader every property on
187/// the note, because Obsidian parses this block as the note's properties and a
188/// block that does not parse yields no properties at all rather than an error.
189///
190/// All three inputs are legal path components on Linux and macOS. None occurs in
191/// this repository today, so this is a latent defect rather than an observed one.
192///
193/// Escapes, per YAML 1.2 §7.3.1: the two structural characters `\` and `"`, then
194/// anything a parser is not obliged to accept literally — C0 controls, `DEL`, the
195/// C1 range, and the three separators (`U+2028`, `U+2029`, `U+FEFF`) that some
196/// parsers treat as line breaks. Short escapes where YAML defines one, so the
197/// common cases stay readable, and `\uXXXX` otherwise.
198fn yaml_double_quoted(value: &str) -> String {
199 let mut out = String::with_capacity(value.len() + 2);
200 out.push('"');
201 for ch in value.chars() {
202 match ch {
203 '\\' => out.push_str(r"\\"),
204 '"' => out.push_str("\\\""),
205 '\n' => out.push_str(r"\n"),
206 '\r' => out.push_str(r"\r"),
207 '\t' => out.push_str(r"\t"),
208 '\u{0}' => out.push_str(r"\0"),
209 '\u{7}' => out.push_str(r"\a"),
210 '\u{8}' => out.push_str(r"\b"),
211 '\u{b}' => out.push_str(r"\v"),
212 '\u{c}' => out.push_str(r"\f"),
213 '\u{1b}' => out.push_str(r"\e"),
214 // Everything else a YAML parser may reject or fold: the rest of C0,
215 // DEL, the C1 range, and the separators that can read as line breaks.
216 c if (c < ' ')
217 || c == '\u{7f}'
218 || ('\u{80}'..='\u{9f}').contains(&c)
219 || matches!(c, '\u{2028}' | '\u{2029}' | '\u{feff}') =>
220 {
221 let _ = write!(out, "\\u{:04x}", c as u32);
222 }
223 c => out.push(c),
224 }
225 }
226 out.push('"');
227 out
228}
229
230/// Emit `value` in YAML **plain** (unquoted) style when that round-trips, and as
231/// [`yaml_double_quoted`] when it would not.
232///
233/// For the frontmatter fields that are written bare today — `kind`, `lang`,
234/// `status`. Those are constrained by *today's* producers (an ADR's status is
235/// validated against the house states; kinds and languages come from extraction),
236/// but `roteiro load` installs a caller-supplied graph artifact whose nodes carry
237/// whatever JSON they carry, so "the producer is careful" is not a property this
238/// renderer can rely on. A `status:` of `Accepted: superseded by 0012` emitted
239/// bare is a parse error, and `Accepted # pending` silently truncates to
240/// `Accepted`.
241///
242/// Escalating only when needed is what keeps the bytes of an existing vault
243/// unchanged — every `kind`, `lang` and `status` in this repository is plain-safe
244/// and stays bare. [`is_plain_safe`] is deliberately stricter than YAML's plain
245/// grammar for the same reason it is safe: a value it rejects is merely quoted.
246fn yaml_scalar(value: &str) -> String {
247 if is_plain_safe(value) {
248 value.to_owned()
249 } else {
250 yaml_double_quoted(value)
251 }
252}
253
254/// Whether `value` can be written as a bare YAML scalar and read back unchanged.
255///
256/// A conservative allowlist rather than YAML's actual plain-scalar grammar, which
257/// is subtle enough (indicator characters, `: ` and ` #` only in some positions,
258/// leading and trailing space, implicit typing) that implementing it is how the
259/// bug this replaces gets written a second time. Getting this wrong in the
260/// strict direction costs a pair of quotation marks; getting it wrong in the
261/// permissive direction costs the note's properties.
262///
263/// So: a leading ASCII letter, then letters, digits, `_`, `-`, `.` and `/` — which
264/// covers every kind, language and status this renderer emits — and never a word
265/// YAML resolves to a boolean or null. That last exclusion is not hypothetical:
266/// `no` is the ISO 639-1 code for Norwegian, and YAML 1.1 parsers read a bare `no`
267/// as `false`.
268fn is_plain_safe(value: &str) -> bool {
269 const NOT_STRINGS: [&str; 11] = [
270 "true", "false", "yes", "no", "on", "off", "null", "nil", "none", "y", "n",
271 ];
272 !value.is_empty()
273 && value.starts_with(|c: char| c.is_ascii_alphabetic())
274 && value
275 .chars()
276 .all(|c| c.is_ascii_alphanumeric() || matches!(c, '_' | '-' | '.' | '/'))
277 && !NOT_STRINGS.contains(&value.to_ascii_lowercase().as_str())
278}
279
280/// Which vault a note is being rendered into: a single project's, or one member
281/// of a **workspace** vault spanning several repositories.
282///
283/// This is the whole of the workspace-vault naming rule, in one place. Node keys
284/// are **repository-relative** (`file:README.md` names no repo), so every member
285/// of a workspace produces the same note name for its `README.md` and one would
286/// silently overwrite the rest. Qualifying the key with its project fixes that.
287///
288/// [`VaultScope::PROJECT`] (`project: None`) is not a degenerate case but the
289/// contract: it makes every name in this module reduce to exactly [`note_name`]
290/// of the bare key, with nothing qualified and no `project:` frontmatter.
291///
292/// That reduction is *still* the promise; what it no longer implies is stability
293/// against `main`. #570 could say "a single-project vault's names do not move",
294/// because the only thing moving them would have been workspace qualification.
295/// #574 moves them all, on purpose: the old names were not injective under
296/// filename case folding and the vault lost 104 notes to that. The promise here
297/// was always about **this axis** — turning workspace mode on must not rename a
298/// project's notes — and it holds unchanged. See [`note_name`] for the rename and
299/// what it bought.
300#[derive(Debug, Clone, Copy)]
301pub struct VaultScope<'a> {
302 /// The member project this note belongs to, qualifying its name as
303 /// `<project>::<key>` — the same form ADR-0009's cross-repo links already use.
304 /// `None` ⇒ a single-project vault, and names are unqualified exactly as
305 /// before.
306 pub project: Option<&'a str>,
307 /// The workspace's member project names. An external-ref placeholder whose
308 /// target names one of these is a cross-repo edge the vault can actually
309 /// follow, so it is rendered as a link straight to that member's note. Empty
310 /// for a single-project vault.
311 pub members: &'a std::collections::BTreeSet<String>,
312}
313
314/// The empty member set backing [`VaultScope::PROJECT`] — a single-project vault
315/// has no other members to resolve a cross-repo reference against.
316static NO_MEMBERS: std::collections::BTreeSet<String> = std::collections::BTreeSet::new();
317
318impl VaultScope<'_> {
319 /// A single-project vault: names are unqualified, and no cross-repo reference
320 /// resolves. Every name this produces is byte-identical to [`note_name`] of
321 /// the bare key — see the type's documentation for why that reduction is
322 /// load-bearing, and for what it does *not* promise.
323 pub const PROJECT: Self = Self {
324 project: None,
325 members: &NO_MEMBERS,
326 };
327}
328
329impl Default for VaultScope<'_> {
330 fn default() -> Self {
331 Self::PROJECT
332 }
333}
334
335impl VaultScope<'_> {
336 /// Whether an external-ref placeholder `key` is one this vault resolves for
337 /// itself — its target names a member, so every edge to it points at the real
338 /// note and the placeholder need not be rendered at all.
339 ///
340 /// The single rule behind both halves of that: [`link_target`] redirects
341 /// exactly the keys this accepts, and the caller skips writing exactly the
342 /// notes this accepts. They cannot disagree.
343 #[must_use]
344 pub fn redirects_external_ref(&self, key: &str) -> bool {
345 key.strip_prefix("extref:")
346 .and_then(rto_graph::parse_qualified)
347 .is_some_and(|(project, _)| self.members.contains(project))
348 }
349}
350
351/// The note name for a node `key` owned by `scope`'s project.
352///
353/// In a single-project vault (`scope.project == None`) this *is* [`note_name`].
354/// In a workspace vault it is [`note_name`] of the project-qualified key
355/// `<project>::<key>` — reusing ADR-0009's qualified form rather than inventing a
356/// second one, which is what lets a cross-repo external-ref target (already
357/// stored qualified) map to its note by the very same call.
358#[must_use]
359pub fn scoped_note_name(scope: &VaultScope<'_>, key: &str) -> String {
360 match scope.project {
361 None => note_name(key),
362 Some(project) => note_name(&format!("{project}::{key}")),
363 }
364}
365
366/// The note an edge pointing at `key` should link to.
367///
368/// Almost always [`scoped_note_name`]. The exception is the one cross-repo edge
369/// the graph already models: a spoke's inferred link to a hub is stored as an
370/// edge to a **local external-ref placeholder** (`extref:<project>::<key>`,
371/// [`rto_graph::external_ref_key`]) because store integrity requires both ends of
372/// an edge in one store. A workspace vault holds both repos' notes, so when the
373/// placeholder's target names a member the link is pointed at the **real** note
374/// instead of the stand-in.
375///
376/// This invents no edge. It renders the edge that is there, following the
377/// placeholder exactly as [`rto_graph::Workspace::follow_external_ref`] does at
378/// query time — the cross-repo graph has only ever been *rendered* one repo at a
379/// time.
380fn link_target(scope: &VaultScope<'_>, key: &str) -> String {
381 if scope.redirects_external_ref(key) {
382 // `note_name(qualified)` is by construction the same string
383 // `scoped_note_name` produces for that member's own copy of the node.
384 // `strip_prefix`, not `trim_start_matches`: the latter strips the prefix
385 // repeatedly, which would mangle a target that legitimately starts with it.
386 return note_name(key.strip_prefix("extref:").unwrap_or(key));
387 }
388 scoped_note_name(scope, key)
389}
390
391/// Render a node's [`Explanation`] into an Obsidian note: YAML frontmatter (with
392/// `tags` for the graph view and an ADR's `status`), a clickable **Source** link
393/// (when `source_base` — a web "blob" base like
394/// `https://github.com/org/repo/blob/<sha>` — is known and the node has a path),
395/// the content as the knowledge base, and its edges as provenance-labelled
396/// wikilinks.
397///
398/// `body` is the node's **full source text**, which only the caller can fetch:
399/// this function is a pure function of the `Explanation`, and an `Explanation`
400/// carries no repository, store or blob. When it is `Some`, it replaces
401/// `meta.content` in the note's `## Content` section — see [`note_body`] for why
402/// replacing is the only correct combination of the two.
403#[must_use]
404pub fn render_note(ex: &Explanation, source_base: Option<&str>, body: Option<&str>) -> VaultNote {
405 render_note_scoped(ex, source_base, body, &VaultScope::PROJECT)
406}
407
408/// [`render_note`], for one member of a **workspace** vault: identical except
409/// that the note's own name and every link it emits are resolved through `scope`
410/// (see [`VaultScope`]).
411///
412/// With [`VaultScope::PROJECT`] this is [`render_note`] byte for byte, which is
413/// how the single-project vault's compatibility promise is kept by construction
414/// rather than by a parallel code path that has to be kept in step.
415#[must_use]
416pub fn render_note_scoped(
417 ex: &Explanation,
418 source_base: Option<&str>,
419 body: Option<&str>,
420 scope: &VaultScope<'_>,
421) -> VaultNote {
422 let meta = &ex.meta;
423 let status = meta.get("status").and_then(|v| v.as_str());
424 let content = note_body(meta.get("content").and_then(|v| v.as_str()), body);
425
426 let mut c = String::new();
427 c.push_str("---\n");
428 let _ = writeln!(c, "key: {}", yaml_double_quoted(&ex.node.key));
429 let _ = writeln!(c, "kind: {}", yaml_scalar(ex.node.kind.as_str()));
430 // Which member this note came from. Absent in a single-project vault, where
431 // it would be one constant repeated on every note — and where adding it would
432 // change every note's bytes.
433 if let Some(project) = scope.project {
434 let _ = writeln!(c, "project: {}", yaml_double_quoted(project));
435 }
436 if let Some(path) = &ex.node.path {
437 let _ = writeln!(c, "path: {}", yaml_double_quoted(path));
438 }
439 if let Some(lang) = &ex.node.lang {
440 let _ = writeln!(c, "lang: {}", yaml_scalar(lang));
441 }
442 if let Some(status) = status {
443 let _ = writeln!(c, "status: {}", yaml_scalar(status));
444 }
445 // Nested tags group in Obsidian's tag pane and colour the graph view.
446 c.push_str("tags:\n");
447 let _ = writeln!(c, " - roteiro/kind/{}", tag_slug(&ex.node.kind));
448 // Colours the graph view by member, which is the one thing a workspace vault
449 // is for and a per-project vault has no use for.
450 if let Some(project) = scope.project {
451 let _ = writeln!(c, " - roteiro/project/{}", tag_slug(project));
452 }
453 if let Some(lang) = &ex.node.lang {
454 let _ = writeln!(c, " - roteiro/lang/{}", tag_slug(lang));
455 }
456 if let Some(status) = status {
457 let _ = writeln!(c, " - roteiro/status/{}", tag_slug(status));
458 }
459 c.push_str("---\n\n");
460
461 let _ = writeln!(c, "# {}", ex.node.name);
462 if let Some(status) = status {
463 let _ = writeln!(c, "\n> **Status:** {status}");
464 }
465
466 // A clickable link to the file this node comes from. An absolute URL, so it
467 // works from the downloaded vault too (which has no repo files beside it).
468 if let (Some(base), Some(path)) = (source_base, ex.node.path.as_deref()) {
469 let _ = writeln!(
470 c,
471 "\n**Source:** [`{path}`]({}/{path})",
472 base.trim_end_matches('/')
473 );
474 }
475
476 // The knowledge base: the full source text, or the captured doc comment /
477 // prose / PDF / image text.
478 if let Some(content) = content.map(str::trim).filter(|s| !s.is_empty()) {
479 c.push_str("\n## Content\n\n");
480 c.push_str(content);
481 c.push('\n');
482 }
483
484 if !ex.outgoing.is_empty() {
485 c.push_str("\n## Outgoing\n\n");
486 for e in &ex.outgoing {
487 let _ = writeln!(
488 c,
489 "- {} ({}){} → [[{}]]",
490 e.kind,
491 e.provenance,
492 confidence(e.confidence),
493 link_target(scope, &e.node)
494 );
495 }
496 }
497 if !ex.incoming.is_empty() {
498 c.push_str("\n## Incoming\n\n");
499 for e in &ex.incoming {
500 let _ = writeln!(
501 c,
502 "- [[{}]] {} ({}){} →",
503 link_target(scope, &e.node),
504 e.kind,
505 e.provenance,
506 confidence(e.confidence)
507 );
508 }
509 }
510
511 VaultNote {
512 filename: format!("{}.md", scoped_note_name(scope, &ex.node.key)),
513 content: c,
514 }
515}
516
517/// Choose the text a note shows: the caller's full `body` when it has one, else
518/// the node's stored `content`.
519///
520/// The two are **not** complementary, they are the same text at two fidelities,
521/// so a note shows one of them and never both. `meta.content` is an embedding
522/// budget — extraction caps it (1500 chars) and collapses every whitespace run to
523/// a single space, which is right for a store that ships with the graph and wrong
524/// for a note: a 23 KB document arrives as one 1500-character line with every
525/// heading, table and code fence flattened into it. Where the caller can supply
526/// the source, that is what a reader wants; appending the capped rendering
527/// underneath it would only restate its first 6% badly.
528fn note_body<'a>(content: Option<&'a str>, body: Option<&'a str>) -> Option<&'a str> {
529 body.or(content)
530}
531
532/// `" (0.82)"` for an inferred edge's confidence, else empty.
533fn confidence(c: Option<f64>) -> String {
534 c.map_or_else(String::new, |c| format!(" ({c:.2})"))
535}
536
537/// A tag-safe slug: lowercase, non-alphanumeric runs → `-`. Keeps Obsidian tags
538/// (`roteiro/kind/adr-section`) valid and stable.
539fn tag_slug(s: &str) -> String {
540 let mut out = String::with_capacity(s.len());
541 let mut prev_dash = false;
542 for ch in s.chars() {
543 if ch.is_ascii_alphanumeric() {
544 out.push(ch.to_ascii_lowercase());
545 prev_dash = false;
546 } else if !prev_dash {
547 out.push('-');
548 prev_dash = true;
549 }
550 }
551 out.trim_matches('-').to_owned()
552}
553
554/// One ADR in the overview, with its lifecycle status.
555#[derive(Debug, Clone)]
556pub struct AdrEntry {
557 /// The ADR node key (`adr:<id>`).
558 pub key: String,
559 /// The ADR title.
560 pub name: String,
561 /// Lifecycle status (`Accepted`, …), if recorded.
562 pub status: Option<String>,
563}
564
565/// The `_Home` overview's config-secret inventory figures.
566///
567/// Counts and file paths only — deliberately not the key names, which belong in
568/// `roteiro config-secrets` where the caveat can be stated at length. A vault note
569/// is read casually and out of context, which is exactly the wrong place for a
570/// list that looks like a secret scan's output.
571#[derive(Debug, Clone, Default)]
572pub struct ConfigSecretSummary {
573 /// Config keys whose **name** matched the secret-name heuristic.
574 pub secret_named: usize,
575 /// Of those, how many had their value redacted before persistence.
576 pub redacted: usize,
577 /// Of those, how many are declared in code with no literal value.
578 pub declared: usize,
579 /// Of those, how many carry an unredacted value. Expected to be zero.
580 pub unredacted: usize,
581 /// Distinct files carrying at least one secret-named key, ordered and capped
582 /// by the caller.
583 pub files: Vec<String>,
584}
585
586/// One file in the `_Home` overview's intent-debt density table.
587#[derive(Debug, Clone)]
588pub struct DensityEntry {
589 /// Repository-relative path, used for both the wikilink and the label.
590 pub path: String,
591 /// Retained markers in the file.
592 pub markers: u32,
593 /// The file's length in lines — the denominator.
594 pub lines: u32,
595 /// Markers per 1,000 lines.
596 pub per_kloc: f64,
597}
598
599/// One node in the `_Home` overview's directed-coupling table.
600#[derive(Debug, Clone)]
601pub struct CouplingEntry {
602 /// The node key, for the wikilink.
603 pub key: String,
604 /// The symbol name.
605 pub name: String,
606 /// Distinct callers.
607 pub fan_in: u32,
608 /// Distinct callees.
609 pub fan_out: u32,
610}
611
612/// Aggregate figures for the vault's `_Home` overview note.
613#[derive(Debug, Clone, Default)]
614pub struct VaultSummary {
615 /// Name of the scanned project (repository directory).
616 pub project: String,
617 /// Total node and edge counts.
618 pub total_nodes: usize,
619 /// Total edge count.
620 pub total_edges: usize,
621 /// `(kind, count)` for each node kind, most-frequent first.
622 pub node_counts: Vec<(String, usize)>,
623 /// `(provenance, edge count)` — `derived` / `authored` / `inferred`.
624 pub edge_provenance: Vec<(String, usize)>,
625 /// The ADRs, with status.
626 pub adrs: Vec<AdrEntry>,
627 /// `(category, count)` of intent-debt markers.
628 pub debt: Vec<(String, usize)>,
629 /// The files where that debt is most **concentrated**, already ranked and
630 /// capped by the caller. Empty when the graph has no markers, or when no
631 /// file carrying one has a recorded length.
632 pub densest_files: Vec<DensityEntry>,
633 /// Secret-named config keys and their redaction state. `None` when the graph
634 /// holds no secret-named config key — the section is then absent rather than
635 /// rendering a row of zeroes, which would read as a clean bill of health this
636 /// lens cannot give.
637 pub config_secrets: Option<ConfigSecretSummary>,
638 /// The most depended-on symbols by **directed** call fan-in, already ranked
639 /// and capped by the caller. Empty when the graph has no `calls` edges.
640 pub most_called: Vec<CouplingEntry>,
641 /// Web root of the repository (`https://host/owner/repo`), if derivable from
642 /// the git remote — for a "Repository" link in the overview, and the
643 /// **clone-from** column of a workspace vault's manifest (#442 part 2).
644 ///
645 /// The web root rather than the raw `origin` fetch URL on purpose: a vault is
646 /// made to be handed to someone, and `git@host:owner/repo.git` is only
647 /// actionable for a reader who already has SSH access to that host.
648 ///
649 /// It is **where the code lives, not a guaranteed clone URL**, and the
650 /// manifest says so. `repo_web_root` normalises a remote to `https://host/…`,
651 /// which clones on the common forges and may not on an unusual one, and says
652 /// nothing about whether the reader can read a private repository. A manifest
653 /// that promised "clone from here" would be making a claim it cannot check.
654 pub repo_url: Option<String>,
655 /// Hex commit the graph was rendered from, for a permalink note — and, in a
656 /// workspace vault, the commit this member is pinned at.
657 ///
658 /// With [`Self::repo_url`] it is what makes a workspace vault **replicable**
659 /// rather than merely browsable: *"here is my workspace"* is far less useful
660 /// than *"here is my workspace **at these commits**"*, and it is what lets a
661 /// reader tell a stale vault from a current one instead of guessing.
662 pub commit: Option<String>,
663 /// The **enabled** `[ingest]` toggles this member was extracted under, by
664 /// name (`prose`, `pdf`, …), and its `[debt] ignore` globs.
665 ///
666 /// The manifest's third leg, after clone URL and commit: those two get a
667 /// reader the same *source*, and these two are what decide whether the same
668 /// source produces the same *vault*. `[ingest] prose` off means notes with no
669 /// captured content; a `[debt] ignore` glob means the debt figures on this
670 /// page are already filtered. Without them "reproducible" means "you can
671 /// obtain the code", which is a weaker claim than the section makes.
672 pub settings: RenderedUnder,
673 /// This member's stored analyzer findings (ADR-0012), ordered most severe
674 /// first by the caller. Rendered in a workspace vault's `_Home`; read
675 /// together with [`Self::coverage`], which is what says whether an empty
676 /// list means anything at all.
677 pub findings: Vec<FindingEntry>,
678 /// Whether an analyzer has ever run against this member — the context that
679 /// makes an empty [`Self::findings`] readable rather than reassuring. See
680 /// [`Coverage`].
681 pub coverage: Coverage,
682}
683
684/// Render the vault's overview note: what was scanned, the structure by kind,
685/// the provenance breakdown, the decisions (ADRs) and their status, the
686/// intent-debt summary, and how to navigate. The entry point for the vault.
687#[must_use]
688pub fn render_home(s: &VaultSummary) -> VaultNote {
689 let mut c = String::new();
690 c.push_str("---\ntags:\n - roteiro/home\n---\n\n");
691 let _ = writeln!(c, "# {} — knowledge graph", s.project);
692 c.push_str(
693 "\n*A browsable snapshot of this codebase as one **knowledge graph**, \
694 generated by [Roteiro](https://roteiro.dev). Every symbol, document and \
695 decision is a note, linked to the things it relates to.*\n",
696 );
697 c.push_str(HOW_TO_READ);
698 let _ = writeln!(
699 c,
700 "\n**{} nodes**, **{} edges** across the project.",
701 s.total_nodes, s.total_edges
702 );
703 write_repo_line(&mut c, s);
704 write_summary_sections(&mut c, s, &VaultScope::PROJECT, 2);
705 c.push_str(NAVIGATING);
706
707 VaultNote {
708 filename: HOME_NOTE.to_owned(),
709 content: c,
710 }
711}
712
713/// The "how to read a note" paragraph. Shared verbatim by the single-project and
714/// workspace overviews — the notes themselves are identical in both, so a reader
715/// who learns the format once has learned it for either.
716const HOW_TO_READ: &str = "\n**How to read it.** Open any note to see what a thing is, the intent or \
717 docs behind it (its **Content**), where it lives (its **Source** link), \
718 and how it connects (**Outgoing**/**Incoming** links). Each link is \
719 labelled with how the fact was established — `derived` (extracted from \
720 code), `authored` (human intent: ADRs, blueprints, annotations), or \
721 `inferred` (a scored suggestion). Open Obsidian's **graph view** to see \
722 the whole thing at once.\n";
723
724/// The closing navigation section.
725const NAVIGATING: &str = "\n## Navigating this vault\n\n\
726 - Open the **graph view** to see the whole codebase; notes are coloured/\
727 filterable by their `roteiro/kind/*`, `roteiro/lang/*` and \
728 `roteiro/status/*` tags.\n\
729 - Each note carries its captured **content** (doc comments, prose, PDF/\
730 image text) and its provenance-labelled incoming/outgoing links.\n\
731 - Start from an ADR above, or search the tag pane for a kind.\n";
732
733/// `**Repository:** …` — the web root and the commit the graph was rendered from.
734fn write_repo_line(c: &mut String, s: &VaultSummary) {
735 if let Some(repo) = &s.repo_url {
736 let _ = write!(c, "\n**Repository:** [{repo}]({repo})");
737 if let Some(commit) = &s.commit {
738 let short = &commit[..commit.len().min(12)];
739 let _ = write!(c, " · rendered at commit `{short}`");
740 }
741 c.push('\n');
742 }
743}
744
745/// Every aggregate the overview carries for **one project**: structure by kind,
746/// provenance, ADRs, intent debt (and where it is densest), the config-secret
747/// inventory and directed call coupling.
748///
749/// Factored out of [`render_home`] so a workspace vault's per-member section is
750/// *the same code*, not a reimplementation that can drift: the promise in issue
751/// #442 is that today's per-project view stays a **subset** of the workspace one
752/// rather than a casualty of it. `level` is the markdown heading depth — 2 for a
753/// single-project `_Home`, 3 inside a member's section — and `scope` decides
754/// whether the wikilinks point at bare or project-qualified notes.
755fn write_summary_sections(c: &mut String, s: &VaultSummary, scope: &VaultScope<'_>, level: usize) {
756 let hd = &"#".repeat(level);
757 let sub = &"#".repeat(level + 1);
758 write_structure(c, s, hd);
759 write_decisions(c, s, scope, hd);
760 write_debt(c, s, scope, hd, sub);
761 write_config_secrets(c, s, scope, hd);
762 write_coupling(c, s, scope, hd);
763}
764
765/// `Structure` (nodes by kind) and `Provenance` (edges by how they were established).
766fn write_structure(c: &mut String, s: &VaultSummary, hd: &str) {
767 let _ = write!(c, "\n{hd} Structure\n\n| Kind | Count |\n| --- | --- |\n");
768 for (kind, n) in &s.node_counts {
769 let _ = writeln!(c, "| {kind} | {n} |");
770 }
771
772 if !s.edge_provenance.is_empty() {
773 let _ = write!(
774 c,
775 "\n{hd} Provenance\n\n| Provenance | Edges |\n| --- | --- |\n"
776 );
777 for (prov, n) in &s.edge_provenance {
778 let _ = writeln!(c, "| {prov} | {n} |");
779 }
780 }
781}
782
783/// `Decisions (ADRs)` — the recorded decisions and their lifecycle status.
784fn write_decisions(c: &mut String, s: &VaultSummary, scope: &VaultScope<'_>, hd: &str) {
785 let _ = write!(c, "\n{hd} Decisions (ADRs)\n\n");
786 if s.adrs.is_empty() {
787 c.push_str("*No ADRs found.*\n");
788 } else {
789 for adr in &s.adrs {
790 let status = adr.status.as_deref().unwrap_or("—");
791 let _ = writeln!(
792 c,
793 "- **{status}** — [[{}|{}]]",
794 scoped_note_name(scope, &adr.key),
795 adr.name
796 );
797 }
798 }
799}
800
801/// `Intent debt` — the marker categories, and the files the debt is densest in.
802fn write_debt(c: &mut String, s: &VaultSummary, scope: &VaultScope<'_>, hd: &str, sub: &str) {
803 let _ = write!(c, "\n{hd} Intent debt\n\n");
804 if s.debt.is_empty() {
805 c.push_str("*None recorded.*\n");
806 } else {
807 c.push_str("| Category | Count |\n| --- | --- |\n");
808 for (cat, n) in &s.debt {
809 let _ = writeln!(c, "| {cat} | {n} |");
810 }
811 }
812
813 if !s.densest_files.is_empty() {
814 let _ = write!(
815 c,
816 "\n{sub} Densest files (markers per 1,000 lines)\n\n\
817 *Where the debt above is concentrated, rather than where there is \
818 most of it — a raw count ranks the biggest file first by \
819 construction. The denominator is file length: every line, blanks and \
820 comments included, not source lines of code. Prose matches (`for \
821 now`, `tbd`) count too, so a design document can rank high.*\n\n"
822 );
823 c.push_str("| File | Markers | Lines | Per 1k |\n| --- | --- | --- | --- |\n");
824 for e in &s.densest_files {
825 let _ = writeln!(
826 c,
827 "| [[{}\\|{}]] | {} | {} | {:.2} |",
828 scoped_note_name(scope, &format!("file:{}", e.path)),
829 e.path,
830 e.markers,
831 e.lines,
832 e.per_kloc
833 );
834 }
835 }
836}
837
838/// `Config keys named like secrets` — an inventory and its unconditional caveat.
839fn write_config_secrets(c: &mut String, s: &VaultSummary, scope: &VaultScope<'_>, hd: &str) {
840 if let Some(cs) = &s.config_secrets {
841 let _ = write!(c, "\n{hd} Config keys named like secrets\n\n");
842 let _ = writeln!(
843 c,
844 "**{}** secret-named config key(s): {} redacted before storage, {} \
845 declared in code without a value, {} unredacted.",
846 cs.secret_named, cs.redacted, cs.declared, cs.unredacted
847 );
848 if cs.unredacted > 0 {
849 let _ = writeln!(
850 c,
851 "\n> [!warning] {} key(s) carry an **unredacted** value. Extraction \
852 always redacts, so these came from an import layer — inspect the \
853 importing tool, not this repository.",
854 cs.unredacted
855 );
856 }
857 if !cs.files.is_empty() {
858 c.push_str("\nIn:\n");
859 for path in &cs.files {
860 let _ = writeln!(
861 c,
862 "- [[{}\\|{path}]]",
863 scoped_note_name(scope, &format!("file:{path}"))
864 );
865 }
866 }
867 // The caveat is unconditional and comes last, so it is the final thing read
868 // in this section. A vault note is browsed out of context; this is exactly
869 // where "config keys named like secrets" would otherwise be misread as a
870 // secret scan that came back clean.
871 c.push_str(
872 "\n*An inventory of config keys whose **names** look secret, not a secret \
873 scan. Values are redacted before they are stored, so this reports that \
874 such keys exist and were redacted — never a value. It cannot see a \
875 hardcoded credential in source code, cannot judge whether a value is \
876 valid, and cannot tell a real secret from a placeholder. A credential \
877 under an innocuous key name (`dsn`, `endpoint`) does not appear here at \
878 all, so this section being small says nothing about whether this \
879 repository leaks secrets.*\n",
880 );
881 }
882}
883
884/// `Most depended-on (call fan-in)` — directed call coupling, capped.
885fn write_coupling(c: &mut String, s: &VaultSummary, scope: &VaultScope<'_>, hd: &str) {
886 if !s.most_called.is_empty() {
887 let _ = write!(
888 c,
889 "\n{hd} Most depended-on (call fan-in)\n\n\
890 *Distinct callers and callees over `calls` edges — direction kept, so \
891 \"everything calls this\" and \"this calls everything\" are not the same \
892 row. Call targets are resolved by simple name, so a short, generically-\
893 named function can absorb every call to that name: read a large fan-in on \
894 one as a question, not a finding.*\n\n"
895 );
896 c.push_str("| Symbol | Called by | Calls |\n| --- | --- | --- |\n");
897 for e in &s.most_called {
898 let _ = writeln!(
899 c,
900 "| [[{}\\|{}]] | {} | {} |",
901 scoped_note_name(scope, &e.key),
902 e.name,
903 e.fan_in,
904 e.fan_out
905 );
906 }
907 }
908}
909
910/// One member's **version pin**: the revision of a hub it deploys (ADR-0009 step 8).
911///
912/// # Why it names its hub
913///
914/// A pin is hub-**relative**, and since #623 a workspace is a snowflake rather
915/// than a star: a project can be a spoke of one project and the hub of others, so
916/// "the hub version this member pins" has no single referent. `infra` pins
917/// `chart`, and `chart` pins `app`. Naming the hub is what lets a reader follow
918/// the chain instead of guessing which end of it a bare revision belongs to.
919#[derive(Debug, Clone, PartialEq, Eq)]
920pub struct MemberPin {
921 /// The member doing the pinning — the spoke.
922 pub member: String,
923 /// The workspace member whose version is pinned — this member's parent in the
924 /// dependency graph, not necessarily the workspace's busiest node.
925 pub hub: String,
926 /// The hub git revision pinned: a submodule sha, or an image tag resolved to a
927 /// hub ref. `None` when the member depends on `hub` but declares no version
928 /// this tool can resolve.
929 ///
930 /// `None` is a **finding**, not an absence of one — see the module's rendering
931 /// of it, and issue #505, which established that "asked and found nothing" must
932 /// not render identically to "never asked".
933 pub rev: Option<String>,
934 /// Where the pin was read from (`submodule vendor/app`, `image acme/app:1.4.0`).
935 /// `None` exactly when `rev` is.
936 pub via: Option<String>,
937}
938
939/// One cross-repo edge the workspace vault can actually follow: a spoke's node
940/// linking to a hub's, through the external-ref placeholder ADR-0009 persists.
941///
942/// Collected by the caller, which has every member's store open; the renderer
943/// only lays them out. Nothing here is a new edge — these are the `inferred`
944/// links `roteiro links` already reports, rendered for the first time.
945#[derive(Debug, Clone)]
946pub struct CrossLink {
947 /// The member the edge starts in.
948 pub from_project: String,
949 /// The source node's key, within `from_project`.
950 pub from_key: String,
951 /// The source node's display name.
952 pub from_name: String,
953 /// The edge kind (`links`, …).
954 pub kind: String,
955 /// Confidence, for an `inferred` edge.
956 pub confidence: Option<f64>,
957 /// Whether this link was **declared** (`[[links]]` in the source repo's
958 /// config, ADR-0009) rather than inferred by key matching.
959 ///
960 /// The distinction is the whole of ADR-0009's `authored → gold,
961 /// inferred → slate`: a declaration is a statement of intent by someone who
962 /// knows the topology, a match is a candidate. Until #573 the vault could not
963 /// draw it, because nothing persisted an authored cross-repo edge — so this
964 /// section carried a blanket caveat saying every row was a candidate.
965 pub authored: bool,
966 /// The project-qualified target, `<project>::<key>` (ADR-0009).
967 pub to_qualified: String,
968 /// Whether `to_qualified`'s project is a member of this workspace — and so
969 /// whether the link resolves to a note in this vault, or dangles because the
970 /// target repository is outside it.
971 pub resolves: bool,
972}
973
974/// The settings a member's notes were rendered under — the ones that change what
975/// the vault *contains*, not the whole merged config.
976///
977/// Deliberately a short list rather than the effective configuration in full.
978/// #442 asks the manifest to record "effective settings", and dumping every
979/// resolved key into a shareable artifact would re-open the redaction question
980/// this vault already has to warn about, to record settings that cannot change
981/// what a reader sees. These two can.
982#[derive(Debug, Clone, Default)]
983pub struct RenderedUnder {
984 /// Enabled `[ingest]` toggles by name, in declaration order. Empty means
985 /// every toggle was off — which is a real state and renders as such, not as
986 /// an absent row.
987 pub ingest: Vec<String>,
988 /// `[debt] ignore` globs. Non-empty means the debt figures on this page are
989 /// **already filtered**, and a reader comparing them against a fresh
990 /// `roteiro debt` without the same config will not match.
991 pub debt_ignore: Vec<String>,
992}
993
994/// One analyzer finding, as the vault renders it (ADR-0012).
995///
996/// A render-facing copy rather than `rto_graph::Finding`, matching [`AdrEntry`]
997/// and [`CouplingEntry`]: the renderer takes the fields it prints and stays free
998/// of the findings model. It deliberately does **not** carry `meta` — that is
999/// whatever the analyzer emitted, kept verbatim, and a shareable artifact is the
1000/// worst place to reproduce "whatever the tool said" unread.
1001#[derive(Debug, Clone)]
1002pub struct FindingEntry {
1003 /// The rule, advisory or check id the analyzer fired (`RUSTSEC-2026-0031`).
1004 pub rule: String,
1005 /// The severity the analyzer assigned — **a tool judgement, not a
1006 /// confidence** — rendered as the analyzer's word rather than the vault's.
1007 pub severity: String,
1008 /// One-line summary.
1009 pub title: String,
1010 /// The analyzer's full message.
1011 pub message: String,
1012 /// Repository-relative path the finding is about, if the analyzer located one.
1013 pub path: Option<String>,
1014 /// The analyzer that produced it, so a reader can tell one tool's opinion
1015 /// from another's rather than reading a merged list as a single verdict.
1016 pub analyzer: String,
1017}
1018
1019/// What a member's analyzer coverage actually is — the distinction the vault
1020/// must never blur.
1021///
1022/// An empty findings list means one of two completely different things, and
1023/// printing both as "no findings" is the failure `roteiro security status`
1024/// records as `no-analyzer-on-record`: **nothing has been analyzed** is not
1025/// **nothing is wrong**. A shareable artifact is the worst place to conflate
1026/// them, because its reader is the one person who cannot check.
1027///
1028/// [`Coverage::NotRun`] is the `Default` deliberately. A `VaultSummary` built
1029/// from `Default` has had no analyzer run against it, and defaulting the other
1030/// way would render "no findings" for a member nobody looked at — the exact
1031/// conflation this type exists to prevent, arrived at by omission.
1032#[derive(Debug, Clone, Default)]
1033pub enum Coverage {
1034 /// No analyzer has ever run against this member. **Not** a clean result.
1035 #[default]
1036 NotRun,
1037 /// At least one analyzer ran, as `(analyzer, version)` per run — so an empty
1038 /// findings list is attributable to a tool that actually looked.
1039 Ran(Vec<(String, String)>),
1040}
1041
1042/// Aggregate figures for a **workspace** vault's `_Home` overview: the members,
1043/// each with exactly the aggregates a single-project `_Home` carries, plus the
1044/// cross-repo links between them.
1045#[derive(Debug, Clone, Default)]
1046pub struct WorkspaceSummary {
1047 /// When this vault was rendered, RFC 3339 UTC.
1048 ///
1049 /// A vault is **read-only and point-in-time**. Stamping it is what gives
1050 /// that property teeth: with the per-member commits, a reader can tell
1051 /// whether what they are looking at still describes the workspace, rather
1052 /// than assuming it does.
1053 pub generated_at: String,
1054 /// The workspace name (`--workspace-name`).
1055 pub name: String,
1056 /// One entry per member repository, in stable name order. Each is the very
1057 /// same [`VaultSummary`] a per-project vault would render.
1058 pub members: Vec<VaultSummary>,
1059 /// Version pins: for each member that depends on another, the hub revision it
1060 /// deploys (ADR-0009 step 8). One entry per **dependency**, not per member — a
1061 /// project with two parents pins each of them separately, and a member with no
1062 /// parents contributes nothing.
1063 ///
1064 /// Ordered by the caller, which holds the dependency graph; the renderer only
1065 /// lays them out.
1066 pub pins: Vec<MemberPin>,
1067 /// Cross-repo links between members, already ordered and capped by the caller.
1068 pub cross_links: Vec<CrossLink>,
1069 /// Cross-repo links found in total, which `cross_links` may be a capped view
1070 /// of — so the section can say what it is not showing.
1071 pub cross_links_total: usize,
1072 /// How many of `cross_links_total` were **declared** (`[[links]]`) rather
1073 /// than inferred.
1074 ///
1075 /// Counted before the cap, not from `cross_links`: that vector is truncated
1076 /// to [`WORKSPACE_CROSS_LINK_ROWS`](crate::WORKSPACE_CROSS_LINK_ROWS) rows,
1077 /// so counting it would describe the rows on screen while reading as a
1078 /// statement about the workspace — a caption that quietly changes meaning
1079 /// once a workspace grows past the cap.
1080 pub cross_links_authored: usize,
1081}
1082
1083/// Render a **workspace** vault's overview: the members and their scale, the
1084/// cross-repo links between them, and then each member's own aggregates —
1085/// structure, provenance, ADRs, intent debt, config-secret inventory and call
1086/// coupling — under its own heading.
1087///
1088/// The per-member sections are rendered by the same [`write_summary_sections`]
1089/// the single-project `_Home` uses, so the existing view is a **subset** of this
1090/// one: someone who came for their repository's coupling and debt tables finds
1091/// them, rather than a workspace total that averages them away.
1092#[must_use]
1093pub fn render_workspace_home(ws: &WorkspaceSummary) -> VaultNote {
1094 let members: std::collections::BTreeSet<String> =
1095 ws.members.iter().map(|m| m.project.clone()).collect();
1096
1097 let mut c = String::new();
1098 c.push_str("---\ntags:\n - roteiro/home\n - roteiro/workspace\n---\n\n");
1099 let _ = writeln!(c, "# {} — workspace knowledge graph", ws.name);
1100 c.push_str(
1101 "\n*A browsable snapshot of a whole **workspace** as one **knowledge \
1102 graph**, generated by [Roteiro](https://roteiro.dev). Every symbol, \
1103 document and decision in every member repository is a note, linked to the \
1104 things it relates to — including across repositories.*\n",
1105 );
1106 c.push_str(HOW_TO_READ);
1107 // The example is *rendered* by `note_name` rather than spelled out. A
1108 // hand-written spelling of this sentence survived #574 unchanged, so every
1109 // vault v2.0.0 built stated the pre-#574 naming rule — on the first page a
1110 // reader opens — while `note_name` was writing something else. This is the
1111 // one copy that lives in the crate defining the rule, so it can simply ask:
1112 // a derived example cannot drift, and a spelled one already has.
1113 //
1114 // The *key* it names has to be real too, or the fix trades one false
1115 // sentence in `_Home` for another: `<member>::file:README.md` was fabricated
1116 // from the member list, and workspace membership does not require a README.
1117 // A cross-repo link's **source** end is the strongest key available here —
1118 // `from_project` is a member by definition and `from_key` is a node in that
1119 // member's own store, which the Cross-repo links table below already links
1120 // to by name. The *target* end will not do: `resolves == false` means the
1121 // target repository is outside this vault, so `to_qualified` names no note
1122 // here — the same false claim one remove away.
1123 //
1124 // With no cross-repo links there is no key this function can prove is a
1125 // node, so the sentence says nothing rather than inventing one. The rule it
1126 // states is complete without an example; only the illustration is lost.
1127 let example = ws.cross_links.first().map_or_else(String::new, |l| {
1128 let key = format!("{}::{}", l.from_project, l.from_key);
1129 format!(" Here, `{key}` is the note `{}.md`.", note_name(&key))
1130 });
1131 let _ = writeln!(
1132 c,
1133 "\n**Every note is keyed `<project>::<key>`**, because a node key is \
1134 repository-relative: the same path or symbol can occur in more than one \
1135 member, and without the project the second note would overwrite the \
1136 first. A note's *filename* is derived from that key — a readable \
1137 lowercase hint, then a hash of the whole key — so no filename contains \
1138 `::`.{example} Filter the graph view by a member's `roteiro/project/*` \
1139 tag to see one repository at a time."
1140 );
1141
1142 let total_nodes: usize = ws.members.iter().map(|m| m.total_nodes).sum();
1143 let total_edges: usize = ws.members.iter().map(|m| m.total_edges).sum();
1144 let _ = writeln!(
1145 c,
1146 "\n**{total_nodes} nodes**, **{total_edges} edges** across **{}** member \
1147 repositor{}.",
1148 ws.members.len(),
1149 if ws.members.len() == 1 { "y" } else { "ies" }
1150 );
1151
1152 c.push_str("\n## Members\n\n| Project | Nodes | Edges | Repository | Commit |\n| --- | --- | --- | --- | --- |\n");
1153 for m in &ws.members {
1154 let repo = m
1155 .repo_url
1156 .as_ref()
1157 .map_or_else(|| "—".to_owned(), |u| format!("[{u}]({u})"));
1158 let commit = m.commit.as_ref().map_or_else(
1159 || "—".to_owned(),
1160 |c| format!("`{}`", &c[..c.len().min(12)]),
1161 );
1162 let _ = writeln!(
1163 c,
1164 "| [[#{}\\|{}]] | {} | {} | {repo} | {commit} |",
1165 m.project, m.project, m.total_nodes, m.total_edges
1166 );
1167 }
1168 // `[[#Heading]]`, the form the member rows above already use, rather than a
1169 // `](#slug)` anchor: the slug a renderer computes and the slug a reader's
1170 // tool computes are exactly what issue #524 is about, and a manifest is a
1171 // poor place to find out they disagree.
1172 c.push_str(
1173 "\n*The `Repository` and `Commit` columns say where each member came from \
1174 and what was read. The manifest that makes those reconstructable — and \
1175 states what this vault carries before you share it — is \
1176 [[#Reproducing this vault]], below.*\n",
1177 );
1178
1179 write_cross_links(&mut c, ws);
1180 write_findings(&mut c, ws);
1181 write_manifest(&mut c, ws);
1182
1183 for m in &ws.members {
1184 let _ = writeln!(c, "\n## {}", m.project);
1185 let _ = writeln!(
1186 c,
1187 "\n**{} nodes**, **{} edges** in this member.",
1188 m.total_nodes, m.total_edges
1189 );
1190 write_repo_line(&mut c, m);
1191 let scope = VaultScope {
1192 project: Some(&m.project),
1193 members: &members,
1194 };
1195 write_summary_sections(&mut c, m, &scope, 3);
1196 }
1197
1198 c.push_str(NAVIGATING);
1199
1200 VaultNote {
1201 filename: HOME_NOTE.to_owned(),
1202 content: c,
1203 }
1204}
1205
1206/// The `### Version pins` table — which hub revision each member deploys
1207/// (ADR-0009 step 8, issue #442).
1208///
1209/// # Why this is part of a *manifest* rather than a curiosity
1210///
1211/// The rest of the manifest reconstructs the workspace **as rendered**: clone
1212/// here, check out these commits. Pins answer a different question — what each
1213/// deployment repo is *running*, which is rarely the commit it was rendered at. A
1214/// spoke at `HEAD` deploying `app@1.4.0` describes a live system that its own
1215/// commit id says nothing about.
1216///
1217/// # Each pin names its hub
1218///
1219/// Because since #623 a workspace is a snowflake, not a star: `infra → chart →
1220/// app`, where `chart` is a spoke of one and the hub of the others. A column of
1221/// bare revisions would be ambiguous the moment a workspace has more than one
1222/// level, and silently so.
1223///
1224/// # A member that pins nothing is said to pin nothing
1225///
1226/// A dependency with no resolvable version renders as *(none detected)* rather
1227/// than being dropped. Issue #505 settled the rule this follows: **asked and found
1228/// nothing** must not render identically to **never asked**. Dropping the row
1229/// would make a workspace where pin detection is inert look exactly like one with
1230/// no dependencies at all.
1231fn write_pins(c: &mut String, ws: &WorkspaceSummary) {
1232 if ws.pins.is_empty() {
1233 return;
1234 }
1235 c.push_str("\n### Version pins\n\n");
1236 let found = ws.pins.iter().filter(|p| p.rev.is_some()).count();
1237 // Both halves agree with the count, not just the noun: "1 of 1 dependency
1238 // resolve" is what varying one and fixing the other produces.
1239 let (noun, verb) = if ws.pins.len() == 1 {
1240 ("dependency", "resolves")
1241 } else {
1242 ("dependencies", "resolve")
1243 };
1244 let _ = writeln!(
1245 c,
1246 "*{found} of {} {noun} {verb} to a hub revision. A pin is what the member \
1247 **deploys**, which is not the commit it was rendered at.*\n",
1248 ws.pins.len()
1249 );
1250 c.push_str("| Member | Pins | At revision | Read from |\n| --- | --- | --- | --- |\n");
1251 for p in &ws.pins {
1252 let _ = writeln!(
1253 c,
1254 "| {} | {} | {} | {} |",
1255 // Names are prose, not spans: `inline_untrusted` escapes `|` so they
1256 // are table-safe without becoming monospaced. The revision and the
1257 // source string come from git and from a config file, so they go
1258 // through `table_cell` like every other value this manifest quotes.
1259 inline_untrusted(&p.member),
1260 inline_untrusted(&p.hub),
1261 p.rev
1262 .as_deref()
1263 .map_or_else(|| "*(none detected)*".to_owned(), table_cell),
1264 p.via.as_deref().map_or_else(|| "—".to_owned(), table_cell),
1265 );
1266 }
1267}
1268
1269/// The `### Rendered under` table — the settings half of the manifest.
1270///
1271/// A clone URL and a commit get a reader the same **source**; these get them the
1272/// same **vault**. Rendering the same commits with `[ingest] prose` off, or with
1273/// a different `[debt] ignore`, produces a different document from the same
1274/// code — so a manifest that omits them promises a reproducibility it cannot
1275/// deliver, which is worse than promising less.
1276fn write_rendered_under(c: &mut String, ws: &WorkspaceSummary) {
1277 c.push_str("\n### Rendered under\n\n");
1278 c.push_str(
1279 "*These change what a vault **contains**, so re-rendering the commits \
1280 above under different ones will not reproduce this document.*\n\n",
1281 );
1282 c.push_str("| Member | `[ingest]` | `[debt] ignore` |\n| --- | --- | --- |\n");
1283 for m in &ws.members {
1284 let list = |v: &[String]| {
1285 if v.is_empty() {
1286 // "none" as a word, not an empty cell: a blank reads as unknown,
1287 // and these two are the difference between reproducing this
1288 // vault and something that merely resembles it.
1289 "*none*".to_owned()
1290 } else {
1291 // `[debt] ignore` is user config landing in a table cell: a
1292 // glob carrying a backtick or a `|` reshapes the row. Not
1293 // analyzer-sourced, but untrusted for the same reason — the
1294 // vault did not write it.
1295 v.iter()
1296 .map(|x| table_cell(x))
1297 .collect::<Vec<_>>()
1298 .join(", ")
1299 }
1300 };
1301 let _ = writeln!(
1302 c,
1303 "| {} | {} | {} |",
1304 inline_untrusted(&m.project),
1305 list(&m.settings.ingest),
1306 list(&m.settings.debt_ignore)
1307 );
1308 }
1309}
1310
1311/// The `## Security findings` section: the stored analyzer output (ADR-0012) for
1312/// every member.
1313///
1314/// # Why this is here at all, stated once
1315///
1316/// ADR-0012 keeps findings out of `nodes`/`edges` **specifically** so they cannot
1317/// reach `export_factset`, and records the alternative as rejected because it
1318/// would "silently publish tool output into an artifact". This section is that
1319/// same publication, made **deliberately and visibly** rather than by accident:
1320/// a workspace vault is a hand-over document, and the owner ruled that it should
1321/// answer *"what is wrong with this workspace"* as well as *"what is in it"*.
1322/// The exclusions section says so at the point a reader is about to share it.
1323///
1324/// # The distinction this section exists to preserve
1325///
1326/// A member with no findings is rendered **two different ways** depending on
1327/// [`Coverage`], because "no analyzer has run" and "an analyzer ran and found
1328/// nothing" are opposite facts that look identical when both are printed as an
1329/// empty list. `roteiro security status` names that failure `no-analyzer-on-record`
1330/// and refuses to let it read as clean; a shareable artifact must refuse harder,
1331/// because its reader is the one person who cannot go and check.
1332fn write_findings(c: &mut String, ws: &WorkspaceSummary) {
1333 c.push_str("\n## Security findings\n\n");
1334
1335 let total: usize = ws.members.iter().map(|m| m.findings.len()).sum();
1336 let unanalyzed = ws
1337 .members
1338 .iter()
1339 .filter(|m| matches!(m.coverage, Coverage::NotRun))
1340 .count();
1341
1342 let _ = writeln!(
1343 c,
1344 "*Stored analyzer output (ADR-0012). **{total} finding(s)** across \
1345 {} member(s){}. A severity is the **analyzer's** judgement, not \
1346 Roteiro's, and a finding is a tool's claim rather than a confirmed \
1347 defect — read one as something to check, not something proven.*\n",
1348 ws.members.len(),
1349 if unanalyzed > 0 {
1350 format!(", with {unanalyzed} never analyzed at all")
1351 } else {
1352 String::new()
1353 }
1354 );
1355
1356 for m in &ws.members {
1357 match (&m.coverage, m.findings.is_empty()) {
1358 // Never analyzed. Said loudly, and *not* in the same breath as a
1359 // member that came back clean — the whole point of the split.
1360 (Coverage::NotRun, _) => {
1361 let _ = writeln!(c, "### {} — **not analyzed**\n", m.project);
1362 c.push_str(
1363 "*No analyzer has run against this member, so nothing here is \
1364 a statement about it. An empty list is the absence of a \
1365 question, not a reassuring answer.*\n\n",
1366 );
1367 }
1368 // Analyzed, nothing found — attributable to a tool that looked.
1369 (Coverage::Ran(runs), true) => {
1370 let _ = writeln!(
1371 c,
1372 "### {} — no findings\n\n*{} ran and reported none.*\n",
1373 m.project,
1374 describe_runs(runs)
1375 );
1376 }
1377 (Coverage::Ran(runs), false) => {
1378 let _ = writeln!(
1379 c,
1380 "### {} — {} finding(s)\n\n*Reported by {}.*\n",
1381 m.project,
1382 m.findings.len(),
1383 describe_runs(runs)
1384 );
1385 for f in &m.findings {
1386 let at = f
1387 .path
1388 .as_deref()
1389 .map_or_else(String::new, |p| format!(" · {}", inline_code(p)));
1390 let _ = writeln!(
1391 c,
1392 "**{}** · {}{at} — {}\n",
1393 inline_code(&f.severity),
1394 inline_code(&f.rule),
1395 // Prose, not a span: the title is the sentence a reader
1396 // scans, so it is escaped rather than monospaced.
1397 inline_untrusted(&f.title)
1398 );
1399 write_analyzer_message(c, f);
1400 }
1401 }
1402 }
1403 }
1404}
1405
1406/// An untrusted value as a **complete code span**, delimiters included.
1407///
1408/// A code span renders its content literally — no emphasis, no links, no HTML —
1409/// so the only thing that can escape one is a backtick run long enough to close
1410/// it. Widening the delimiter past the longest run inside the value is therefore
1411/// the whole defence, and it is the *right* defence: backslash-escaping inside a
1412/// span protects nothing and does show, turning `vendor/**` into `vendor\*\*`
1413/// on the page.
1414///
1415/// Whitespace is still collapsed — a newline ends the line the span sits on
1416/// however well the span itself is delimited.
1417///
1418/// **Not sufficient inside a table cell.** GFM splits a row on `|` *before*
1419/// inline parsing, so a pipe inside a code span still ends the cell. Use
1420/// [`table_cell`] there.
1421fn inline_code(s: &str) -> String {
1422 let collapsed = s.split_whitespace().collect::<Vec<_>>().join(" ");
1423 let longest_run = collapsed
1424 .split(|ch| ch != '`')
1425 .map(str::len)
1426 .max()
1427 .unwrap_or(0);
1428 let tick = "`".repeat(longest_run + 1);
1429 // CommonMark strips one leading and trailing space from a span, which is how
1430 // content that itself starts or ends with a backtick is expressed.
1431 let pad = if collapsed.starts_with('`') || collapsed.ends_with('`') {
1432 " "
1433 } else {
1434 ""
1435 };
1436 format!("{tick}{pad}{collapsed}{pad}{tick}")
1437}
1438
1439/// One line of analyzer-sourced text, made safe to interpolate into the note.
1440///
1441/// The message gets a fence; **these fields did not**, and they come from the
1442/// same place. A `title` carrying a newline ends the line it was placed on and
1443/// starts a new block — one that can open a heading, a list or a table row — so
1444/// the "quoted, not absorbed" property held for one field out of four. A `rule`
1445/// or `path` carrying a backtick escapes the code span it is wrapped in.
1446///
1447/// Four steps, in order:
1448///
1449/// 1. **Collapse every whitespace run to one space.** This is what removes the
1450/// structural attack: Markdown block constructs need a line start, and after
1451/// this there are no line starts left inside the value.
1452/// 2. **HTML-escape `&`, `<`, `>`.** Some of these values land inside raw HTML
1453/// (`<sub>`, `<details><summary>`), where a backslash escapes nothing — an
1454/// analyzer named `</sub><script>` would close the tag and keep going. This
1455/// step is why there is **one** helper rather than a Markdown one and an HTML
1456/// one: two helpers means two contexts to keep straight, and the reason this
1457/// function exists at all is that the first version secured one context and
1458/// missed its neighbour.
1459/// 3. **Backslash-escape the remaining Markdown punctuation.** A link in a
1460/// document handed to someone can point anywhere, and a stray backtick or
1461/// pipe silently reshapes the line it lands in. `<` and `>` are absent from
1462/// that set because step 2 already turned them into entities.
1463/// 4. **Defuse bare URLs.** Escaping `[` stops explicit `[text](url)` syntax and
1464/// nothing else: Obsidian and GFM *linkify* a plain `https://…`, so analyzer
1465/// text could still hand the reader a clickable link to anywhere. Replacing
1466/// the scheme's colon with `:` renders identically and matches no
1467/// linkifier — the reader still sees the URL and can copy it deliberately.
1468///
1469/// Safe in both contexts, so a caller never has to know which one it is in.
1470fn inline_untrusted(s: &str) -> String {
1471 let collapsed = s.split_whitespace().collect::<Vec<_>>().join(" ");
1472 let mut out = String::with_capacity(collapsed.len());
1473 for ch in collapsed.chars() {
1474 match ch {
1475 '&' => out.push_str("&"),
1476 '<' => out.push_str("<"),
1477 '>' => out.push_str(">"),
1478 // `&` is deliberately absent below: the entities written above must
1479 // survive intact, and backslash-escaping their `&` would render them
1480 // literally.
1481 '\\' | '`' | '*' | '_' | '[' | ']' | '|' | '#' => {
1482 out.push('\\');
1483 out.push(ch);
1484 }
1485 _ => out.push(ch),
1486 }
1487 }
1488 defuse_autolinks(&out)
1489}
1490
1491/// Stop a bare URL in untrusted text from being turned into a clickable link.
1492///
1493/// Escaping `[` covers explicit link syntax; it does nothing about linkification,
1494/// which is on by default in Obsidian and GFM. `https://evil.example` renders
1495/// as `https://evil.example` and matches no autolinker, so the URL stays
1496/// readable and copyable while ceasing to be a thing the reader can click by
1497/// accident in a document someone handed them.
1498fn defuse_autolinks(s: &str) -> String {
1499 s.replace("https://", "https://")
1500 .replace("http://", "http://")
1501}
1502
1503/// [`inline_code`] for a value that lands in a **table cell**.
1504///
1505/// A code span is not enough there: GFM splits a row on `|` before it parses
1506/// inline content, so a pipe inside the span still ends the cell and shifts
1507/// every column after it. `\|` is the documented escape, and it is applied to
1508/// the finished span so the delimiter widening still holds.
1509///
1510/// The distinction is not cosmetic — it is a third rendering context, and this
1511/// file has now been wrong about a context twice.
1512fn table_cell(s: &str) -> String {
1513 inline_code(s).replace('|', "\\|")
1514}
1515
1516/// One finding's full analyzer message, collapsed and **verbatim**.
1517///
1518/// Collapsed because the messages are long — a single GHSA advisory runs to
1519/// paragraphs, and seventeen of them inline turn `_Home` into a wall of text
1520/// nobody reads, which loses the findings as surely as omitting them. `<details>`
1521/// keeps every byte in the file (the ruling was to include them in full) while
1522/// letting the reader see the list first.
1523///
1524/// **Verbatim, in a fence, rather than as Markdown.** An analyzer message is
1525/// tool output travelling into a document that gets handed to people: rendered
1526/// as Markdown it can open headings that restructure the note, or links that
1527/// point anywhere. A fence is what makes it text the vault *quotes* rather than
1528/// text the vault *becomes* — and it preserves the advisory's own line structure,
1529/// which flattening to one line destroys.
1530///
1531/// The fence is sized to beat the longest backtick run in the message, for the
1532/// reason [`crate::docs`] handles multi-backtick spans: a three-backtick fence
1533/// around a message that itself contains one ends the block early and spills the
1534/// remainder into the note as prose.
1535fn write_analyzer_message(c: &mut String, f: &FindingEntry) {
1536 let message = f.message.trim();
1537 if message.is_empty() {
1538 let _ = writeln!(
1539 c,
1540 "<sub>reported by `{}`</sub>\n",
1541 inline_untrusted(&f.analyzer)
1542 );
1543 return;
1544 }
1545 let longest_run = message
1546 .split(|ch| ch != '`')
1547 .map(str::len)
1548 .max()
1549 .unwrap_or(0);
1550 let fence = "`".repeat(longest_run.max(2) + 1);
1551 let _ = writeln!(
1552 c,
1553 "<details><summary><sub>what `{}` said</sub></summary>\n\n\
1554 {fence}text\n{message}\n{fence}\n\n</details>\n",
1555 inline_untrusted(&f.analyzer)
1556 );
1557}
1558
1559/// `analyzer version` for each run that produced a member's findings, joined —
1560/// so "no findings" names the tool that looked rather than asserting a state of
1561/// the world.
1562fn describe_runs(runs: &[(String, String)]) -> String {
1563 if runs.is_empty() {
1564 // `Coverage::Ran` with no runs should not occur; say something true
1565 // rather than rendering an empty clause that reads as a name.
1566 return "an analyzer".to_owned();
1567 }
1568 runs.iter()
1569 .map(|(a, v)| {
1570 // `inline_untrusted`, not `inline_code`: this same value is also
1571 // interpolated into `<summary>`/`<sub>` by `write_analyzer_message`,
1572 // and keeping one treatment for it means there is no second context
1573 // to get wrong later. That is the mistake this pair of helpers was
1574 // introduced to fix, so it is not worth re-creating for monospace.
1575 let (a, v) = (inline_untrusted(a), inline_untrusted(v));
1576 // An ingested report often carries no version the producer stated,
1577 // and the store keeps that as `unknown`. Printing "`osv-scanner`
1578 // unknown" reads as a version string; saying nothing reads as the
1579 // absence it is.
1580 if v.is_empty() || v == "unknown" {
1581 format!("`{a}`")
1582 } else {
1583 format!("`{a}` {v}")
1584 }
1585 })
1586 .collect::<Vec<_>>()
1587 .join(", ")
1588}
1589
1590/// The `## Reproducing this vault` section — the **manifest** half of #442.
1591///
1592/// A vault that says *"here is my workspace"* is far less useful than one that
1593/// says *"here is my workspace **at these commits**"*. With an origin and a
1594/// commit per member, a reader can clone, check out, and hold exactly what this
1595/// vault describes; without them they have a picture and no way back to the
1596/// thing pictured.
1597///
1598/// It is also where the reader is told what a shared vault **does not** contain,
1599/// at the moment they are most likely to share it. Each exclusion is a decision
1600/// recorded on #442 rather than an oversight, so each is named with its reason.
1601fn write_manifest(c: &mut String, ws: &WorkspaceSummary) {
1602 c.push_str("\n## Reproducing this vault\n\n");
1603 let _ = writeln!(
1604 c,
1605 "*Rendered **{}**. A vault is read-only and point-in-time: it describes \
1606 these repositories at these commits, and does not change when they do.*\n",
1607 ws.generated_at
1608 );
1609 // "Repository", not "Clone from". The value is the **web root derived from
1610 // the `origin` remote**, which is the right thing to show a reader — an
1611 // `git@host:owner/repo.git` is only actionable for someone who already has
1612 // SSH to that host — but it is not guaranteed to be a working clone URL for
1613 // every forge or for a private repository. Naming the column "Clone from"
1614 // promised something this cannot deliver, and the manifest's whole value is
1615 // that its promises hold.
1616
1617 let any_origin = ws.members.iter().any(|m| m.repo_url.is_some());
1618 if any_origin {
1619 // Stated here rather than above the `if`: it says "each repository
1620 // below", and above the branch it would be followed immediately by
1621 // "no member has an `origin` remote".
1622 c.push_str(
1623 "*Each repository below is the web root derived from that member's \
1624 `origin` remote — where the code lives, not a guaranteed clone URL. \
1625 A private repository, or a forge with a different clone path, still \
1626 needs whatever access you would normally use.*\n\n",
1627 );
1628 c.push_str("| Member | Repository | At commit |\n| --- | --- | --- |\n");
1629 for m in &ws.members {
1630 let _ = writeln!(
1631 c,
1632 "| {} | {} | {} |",
1633 // The name is prose, not a span — `inline_untrusted` already
1634 // backslash-escapes `|`, so it is table-safe without becoming
1635 // monospaced. The two values below *are* spans, and go through
1636 // `table_cell`: a remote URL and a sha come from git, not from
1637 // us, and this table sits thirty lines above the one that
1638 // already knew that.
1639 inline_untrusted(&m.project),
1640 m.repo_url
1641 .as_deref()
1642 .map_or_else(|| "*(no `origin` remote)*".to_owned(), table_cell),
1643 m.commit
1644 .as_deref()
1645 .map_or_else(|| "*(unknown)*".to_owned(), table_cell),
1646 );
1647 }
1648 } else {
1649 // Every member is local-only. Say so rather than rendering a table of
1650 // "no origin" rows, which reads as a fault rather than as a workspace
1651 // that was never pushed anywhere.
1652 c.push_str(
1653 "*No member has an `origin` remote, so this vault cannot be \
1654 reconstructed from it — it describes repositories that exist only \
1655 where it was rendered.*\n",
1656 );
1657 }
1658
1659 write_pins(c, ws);
1660 write_rendered_under(c, ws);
1661
1662 c.push_str(
1663 "\n### Before you share this\n\n\
1664 Stated here because this is the point at which a vault is handed to \
1665 someone.\n\n\
1666 - **It lists this workspace's known security findings** — see *Security \
1667 findings* above. That is deliberate: a hand-over document should say \
1668 what is wrong as well as what is there. But it means this file is a \
1669 list of **unpatched weaknesses and where they are**, and unlike a \
1670 local store it cannot be un-shared once sent. Treat it as you would \
1671 the analyzer reports themselves.\n\
1672 - **Config keys are redacted by *name*, and that is narrower than it \
1673 looks.** The check matches ten well-known key names and inspects no \
1674 values, so a secret in a value whose key is not named like one — \
1675 `DATABASE_URL=postgres://user:pw@host` — is **not** redacted. \
1676 Tolerable in a local store; materially different in an artifact whose \
1677 purpose is to be handed on.\n\
1678 - **Agent memory** (ADR-0013) is the one thing deliberately left out. It \
1679 is per-developer and uncommitted, it records prose that can carry \
1680 pasted tokens, stack traces and customer names, and it has **no \
1681 redaction chokepoint** at all — so unlike the two above, there is no \
1682 version of it that is safe to include.\n",
1683 );
1684}
1685
1686/// The `## Cross-repo links` section: the edges that only a workspace vault can
1687/// show, and the honest statement of what is missing from them.
1688fn write_cross_links(c: &mut String, ws: &WorkspaceSummary) {
1689 c.push_str("\n## Cross-repo links\n\n");
1690 if ws.cross_links.is_empty() {
1691 c.push_str(
1692 "*None. These are the `inferred` cross-repo links `roteiro links \
1693 --infer --write` persists (ADR-0009); a workspace whose members have \
1694 never been inferred over has none recorded yet.*\n",
1695 );
1696 return;
1697 }
1698 // The caveat is per-row now, because the two provenances are no longer the
1699 // same claim: an **authored** row was declared by someone who knows the
1700 // topology, an **inferred** row is a scored guess. Saying "these are all
1701 // candidates" over a table containing declarations would understate the
1702 // declarations exactly as saying nothing would overstate the matches.
1703 let authored = ws.cross_links_authored;
1704 // `saturating_sub`: `WorkspaceSummary` is public, so a caller can hand us a
1705 // count larger than the total. In release that subtraction wraps and the
1706 // caption reports billions of inferred links — a rendering function should
1707 // not be the place an inconsistent input becomes nonsense. The debug
1708 // assertion says which caller was wrong, in the build that can afford to.
1709 debug_assert!(
1710 authored <= ws.cross_links_total,
1711 "cross_links_authored ({authored}) exceeds cross_links_total ({})",
1712 ws.cross_links_total
1713 );
1714 let inferred = ws.cross_links_total.saturating_sub(authored);
1715 let _ = writeln!(
1716 c,
1717 "*A spoke's config key and the hub key it corresponds to, across \
1718 repositories — the one thing a per-project vault structurally cannot \
1719 show. **{authored} declared** (`[[links]]`, ADR-0009 — a statement of \
1720 intent) and **{inferred} inferred** (`roteiro links --infer --write` — \
1721 read those as candidate correspondences).*\n"
1722 );
1723 c.push_str("| From | | To | Kind |\n| --- | --- | --- | --- |\n");
1724 for l in &ws.cross_links {
1725 let from_scope = VaultScope {
1726 project: Some(&l.from_project),
1727 members: &NO_MEMBERS,
1728 };
1729 let to = if l.resolves {
1730 format!("[[{}\\|{}]]", note_name(&l.to_qualified), l.to_qualified)
1731 } else {
1732 // Outside this workspace: there is no note to link to, and a wikilink
1733 // to a note that does not exist reads in Obsidian as one that is
1734 // merely unwritten.
1735 format!("`{}` *(outside this workspace)*", l.to_qualified)
1736 };
1737 // `declared` rather than a confidence score: an authored link carries no
1738 // score by construction, so an empty cell there would read as "confidence
1739 // unknown" instead of "not that kind of claim".
1740 let how = if l.authored {
1741 " *(declared)*".to_owned()
1742 } else {
1743 confidence(l.confidence)
1744 };
1745 let _ = writeln!(
1746 c,
1747 "| [[{}\\|{}]] | {} | {to} | {}{how} |",
1748 scoped_note_name(&from_scope, &l.from_key),
1749 l.from_name,
1750 l.from_project,
1751 l.kind,
1752 );
1753 }
1754 if ws.cross_links_total > ws.cross_links.len() {
1755 let _ = writeln!(
1756 c,
1757 "\n*Showing {} of {} — the full report is `roteiro links --matrix`.*",
1758 ws.cross_links.len(),
1759 ws.cross_links_total
1760 );
1761 }
1762 c.push_str(
1763 "\n*Shown in one direction only. The edge lives in the spoke's store, \
1764 pointing at a local placeholder for the hub's node, so the hub's own note \
1765 carries no matching **Incoming** entry — Obsidian's **Backlinks** pane \
1766 still shows it, because the link is in the vault.*\n",
1767 );
1768}
1769
1770#[cfg(test)]
1771mod tests {
1772 use super::{
1773 AdrEntry, ConfigSecretSummary, CouplingEntry, Coverage, CrossLink, DensityEntry,
1774 FindingEntry, HOME_NOTE, MemberPin, RenderedUnder, VaultScope, VaultSummary,
1775 WorkspaceSummary, note_name, render_home, render_note, render_note_scoped,
1776 render_workspace_home, scoped_note_name,
1777 };
1778 use rto_graph::{EdgeRef, Explanation, NodeSummary};
1779
1780 /// The shape of a name, pinned once so a change to it is a deliberate edit
1781 /// here rather than a diff spread over twenty other assertions.
1782 ///
1783 /// Everything else in this module composes `note_name` instead of repeating
1784 /// its output, because those tests are about *which key a link points at* and
1785 /// were never about the spelling.
1786 #[test]
1787 fn note_name_is_a_lowercase_hint_and_a_hash_of_the_whole_key() {
1788 assert_eq!(
1789 note_name("sym:rust:src/a.rs#Store"),
1790 "sym-rust-src-a.rs-store-b4cbf6633003361f"
1791 );
1792 assert_eq!(note_name("adr:0001"), "adr-0001-559a2e837953b2ff");
1793 assert_eq!(
1794 note_name("file:src/main.rs"),
1795 "file-src-main.rs-4a72627453f6780e"
1796 );
1797 // Deterministic: the suffix is a pure function of the key, so a vault
1798 // renders the same names on every machine and every run.
1799 assert_eq!(note_name("adr:0001"), note_name("adr:0001"));
1800 }
1801
1802 /// **The property `note_name` exists to have** (issue #574): distinct keys
1803 /// give distinct notes *on a case-folding filesystem*, which is where the
1804 /// vault was losing them.
1805 ///
1806 /// Asserted over lowercased names, not names. On macOS and Windows two names
1807 /// differing only in case are one file, so a name set that is distinct as
1808 /// strings can still be a vault with notes missing — and Linux CI cannot see
1809 /// it. Folding here makes the assertion say what the filesystem says, on
1810 /// every platform.
1811 ///
1812 /// The keys are the two mechanisms that were actually losing notes, taken
1813 /// from this repository's own render rather than invented: the vendored
1814 /// `cytoscape.min.js` bundle whose minified single-letter symbols differ only
1815 /// by a sigil or by case, and a pair of grouped Rust `use` keys differing
1816 /// only by a trailing comma. `render_cli` runs the same assertion end to end
1817 /// over a rendered vault; this is the unit-level statement of it.
1818 #[test]
1819 fn distinct_keys_give_distinct_notes_even_after_case_folding() {
1820 const JS: &str = "sym:javascript:crates/roteiro/src/assets/cytoscape.min.js";
1821 let keys: Vec<String> = [
1822 // Slug lossiness: the sigil and the letter both slugged to the same
1823 // thing (9 notes lost this way, on every platform).
1824 format!("{JS}#$a"),
1825 format!("{JS}#a"),
1826 format!("{JS}#$o"),
1827 format!("{JS}#o"),
1828 // Case folding: distinct names, one file (95 notes lost this way, and
1829 // only on macOS and Windows).
1830 format!("{JS}#A"),
1831 format!("{JS}#O"),
1832 format!("{JS}#S"),
1833 format!("{JS}#s"),
1834 // Real source symbols, same shape.
1835 "sym:rust:crates/rto-exec/src/sandbox_store.rs#Store".into(),
1836 "sym:rust:crates/rto-exec/src/sandbox_store.rs#store".into(),
1837 // A trailing comma is the whole difference between these two.
1838 "import:rust:crate::engine::{ChatRequest,Engine,ModelInfo,}".into(),
1839 "import:rust:crate::engine::{ChatRequest,Engine,ModelInfo}".into(),
1840 // Nothing but separators: no hint at all, so the name is bare hash.
1841 "::".into(),
1842 "##".into(),
1843 // Over the length bound, differing only past the truncation point —
1844 // the case truncation alone used to merge.
1845 format!("import:rust:{}A", "a::b::c,".repeat(60)),
1846 format!("import:rust:{}a", "a::b::c,".repeat(60)),
1847 ]
1848 .into();
1849
1850 let folded: std::collections::BTreeSet<String> =
1851 keys.iter().map(|k| note_name(k).to_lowercase()).collect();
1852 assert_eq!(
1853 folded.len(),
1854 keys.len(),
1855 "two keys share a note after case folding; the vault would hold one \
1856 file for both and report two"
1857 );
1858 }
1859
1860 /// Case folding is the identity on a note name, so the assertion above is not
1861 /// weaker than the filesystem it stands in for.
1862 ///
1863 /// This is the reason the hint is lowercased rather than case-preserved: it
1864 /// makes "distinct names" and "distinct files on macOS" the same statement,
1865 /// so there is no version of this module that passes on Linux and loses notes
1866 /// on a Mac. Without it, the two assertions could drift apart and only the
1867 /// weaker one would ever run in CI.
1868 #[test]
1869 fn a_note_name_is_already_lowercase() {
1870 for key in [
1871 "sym:rust:src/a.rs#Store",
1872 "file:README.md",
1873 "app::file:CHANGELOG.md",
1874 "sym:javascript:a.js#ABC",
1875 ] {
1876 let name = note_name(key);
1877 assert_eq!(name, name.to_lowercase(), "`{key}` kept case in its name");
1878 }
1879 }
1880
1881 /// `_Home` is a name in the same namespace as every note, and it is not
1882 /// derived from a key — so nothing must be able to collide with it. The
1883 /// mandatory suffix gives that for free: every generated name either ends in
1884 /// `-<16 hex>` or *is* 16 hex digits, and `_home` is neither.
1885 #[test]
1886 fn no_key_can_claim_the_home_note() {
1887 for key in ["_Home", "file:_Home", "_home", "::_Home::"] {
1888 assert_ne!(
1889 format!("{}.md", note_name(key)).to_lowercase(),
1890 HOME_NOTE.to_lowercase(),
1891 "`{key}` would overwrite the overview note"
1892 );
1893 }
1894 }
1895
1896 #[test]
1897 fn render_note_emits_frontmatter_and_wikilinks() {
1898 let ex = Explanation {
1899 schema: rto_graph::SCHEMA,
1900 node: NodeSummary {
1901 key: "sym:rust:a.rs#main".into(),
1902 kind: "fn".into(),
1903 name: "main".into(),
1904 path: Some("a.rs".into()),
1905 lang: Some("rust".into()),
1906 },
1907 meta: serde_json::Value::Null,
1908 outgoing: vec![EdgeRef {
1909 kind: "calls".into(),
1910 provenance: "derived",
1911 confidence: None,
1912 node: "sym:rust:a.rs#helper".into(),
1913 }],
1914 incoming: vec![EdgeRef {
1915 kind: "references".into(),
1916 provenance: "authored",
1917 confidence: None,
1918 node: "adr:0001".into(),
1919 }],
1920 };
1921 let note = render_note(&ex, None, None);
1922 assert_eq!(
1923 note.filename,
1924 format!("{}.md", note_name("sym:rust:a.rs#main"))
1925 );
1926 assert!(note.content.contains("kind: fn"));
1927 // No source base → no Source link.
1928 assert!(!note.content.contains("**Source:**"));
1929 assert!(note.content.contains("# main"));
1930 assert!(note.content.contains(&format!(
1931 "- calls (derived) → [[{}]]",
1932 note_name("sym:rust:a.rs#helper")
1933 )));
1934 assert!(note.content.contains(&format!(
1935 "- [[{}]] references (authored) →",
1936 note_name("adr:0001")
1937 )));
1938 // Tags for the graph view.
1939 assert!(note.content.contains("- roteiro/kind/fn"));
1940 assert!(note.content.contains("- roteiro/lang/rust"));
1941 }
1942
1943 #[test]
1944 fn note_name_bounds_long_keys_deterministically() {
1945 let long = format!("import:rust:{}", "a::b::c,".repeat(60));
1946 let a = note_name(&long);
1947 let b = note_name(&long);
1948 assert_eq!(a, b, "deterministic");
1949 assert!(
1950 a.len() <= 205,
1951 "bounded under the filename limit: {}",
1952 a.len()
1953 );
1954 assert_ne!(
1955 note_name(&format!("{long}x")),
1956 a,
1957 "different keys stay distinct after truncation"
1958 );
1959 // Truncation must not leave a doubled separator before the suffix — the
1960 // hint is trimmed after cutting, not before.
1961 assert!(!a.contains("--"), "{a}");
1962 }
1963
1964 /// A short key is bounded too, and every name carries the suffix — the hash
1965 /// is no longer reached for only when the hint overruns.
1966 ///
1967 /// That gating was the defect (#574): two keys short enough to skip the hash
1968 /// had nothing left to tell them apart once the slug had flattened them.
1969 #[test]
1970 fn every_name_carries_the_hash_however_short_the_key() {
1971 for key in ["a", "adr:0001", "file:README.md"] {
1972 let name = note_name(key);
1973 let (hint, hash) = name.rsplit_once('-').expect("a suffixed name");
1974 assert!(!hint.is_empty(), "{name}");
1975 assert_eq!(hash.len(), 16, "{name}");
1976 assert!(
1977 hash.chars().all(|c| c.is_ascii_hexdigit()),
1978 "the suffix is the key's hash, not part of the hint: {name}"
1979 );
1980 }
1981 // A key with no hint at all is the bare hash, which cannot be mistaken
1982 // for a hinted name (those are at least 18 characters).
1983 let bare = note_name("::");
1984 assert_eq!(bare.len(), 16, "{bare}");
1985 assert!(!bare.contains('-'), "{bare}");
1986 }
1987
1988 #[test]
1989 fn render_note_surfaces_content_and_status() {
1990 let ex = Explanation {
1991 schema: rto_graph::SCHEMA,
1992 node: NodeSummary {
1993 key: "adr:0001".into(),
1994 kind: "adr".into(),
1995 name: "Build Roteiro".into(),
1996 path: Some("docs/adr/0001.md".into()),
1997 lang: None,
1998 },
1999 meta: serde_json::json!({ "status": "Accepted", "content": "The decision text." }),
2000 outgoing: vec![],
2001 incoming: vec![],
2002 };
2003 let note = render_note(&ex, Some("https://github.com/org/repo/blob/abc123"), None);
2004 assert!(note.content.contains("status: Accepted"));
2005 assert!(note.content.contains("- roteiro/status/accepted"));
2006 assert!(note.content.contains("> **Status:** Accepted"));
2007 assert!(note.content.contains("## Content\n\nThe decision text."));
2008 // A clickable link to the actual ADR file on the repository host.
2009 assert!(
2010 note.content.contains(
2011 "**Source:** [`docs/adr/0001.md`](https://github.com/org/repo/blob/abc123/docs/adr/0001.md)"
2012 ),
2013 "{}",
2014 note.content
2015 );
2016 }
2017
2018 /// The structured document a prose note is supposed to reproduce: headings, a
2019 /// table and a fenced code block, none of which survive whitespace collapse.
2020 const DOC: &str = "# Working offline\n\nRoteiro is **offline-capable**.\n\n| Host | What |\n| --- | --- |\n| `example.com` | models |\n\n```sh\nroteiro model pull\n```\n";
2021
2022 fn prose_note(content: Option<&str>) -> Explanation {
2023 Explanation {
2024 schema: rto_graph::SCHEMA,
2025 node: NodeSummary {
2026 key: "file:docs/OFFLINE_SETUP.md".into(),
2027 kind: "file".into(),
2028 name: "OFFLINE_SETUP.md".into(),
2029 path: Some("docs/OFFLINE_SETUP.md".into()),
2030 lang: None,
2031 },
2032 meta: content.map_or(
2033 serde_json::Value::Null,
2034 |c| serde_json::json!({ "content": c }),
2035 ),
2036 outgoing: vec![],
2037 incoming: vec![],
2038 }
2039 }
2040
2041 /// The whole readability defect, in one assertion pair: a note built from
2042 /// `meta.content` alone is the document whitespace-collapsed onto one line,
2043 /// and a note built from the source is the document.
2044 ///
2045 /// The newline count is the claim. A character count alone would pass on a
2046 /// note that had merely grown longer while staying flat, which is exactly the
2047 /// failure being fixed — `meta.content` is capped *and* collapsed, and only
2048 /// the collapse is what makes it unreadable.
2049 #[test]
2050 fn a_supplied_body_supersedes_the_collapsed_stored_content() {
2051 // What extraction stores: the same text, whitespace-collapsed.
2052 let collapsed = DOC.split_whitespace().collect::<Vec<_>>().join(" ");
2053 let ex = prose_note(Some(&collapsed));
2054
2055 let note = render_note(&ex, None, Some(DOC));
2056 assert!(
2057 note.content.contains(DOC.trim()),
2058 "the source document is reproduced verbatim: {}",
2059 note.content
2060 );
2061 assert!(
2062 !note.content.contains(&collapsed),
2063 "the collapsed rendering is replaced, not appended: {}",
2064 note.content
2065 );
2066 assert!(
2067 note.content.contains("\n| Host | What |\n"),
2068 "a table needs its own lines to be a table: {}",
2069 note.content
2070 );
2071 assert!(
2072 note.content.contains("\n```sh\n"),
2073 "a fenced block needs its own lines to be a fence: {}",
2074 note.content
2075 );
2076
2077 // The flat control: the same node with no body is the one-line note.
2078 let flat = render_note(&ex, None, None);
2079 assert!(
2080 flat.content.contains(&collapsed),
2081 "without a body the stored content is still shown: {}",
2082 flat.content
2083 );
2084 assert!(
2085 content_lines(¬e.content) > content_lines(&flat.content),
2086 "structure restored: {} line(s) with a body vs {} without",
2087 content_lines(¬e.content),
2088 content_lines(&flat.content)
2089 );
2090 assert_eq!(
2091 content_lines(&flat.content),
2092 1,
2093 "the defect: the stored content is a single line"
2094 );
2095 }
2096
2097 /// A doc comment is a summary of a definition, not a document, and its note is
2098 /// correct as it stands. The caller supplies no body for these, so this pins
2099 /// the unchanged path — the fix must not depend on every node gaining one.
2100 #[test]
2101 fn a_note_with_no_body_is_unchanged() {
2102 let ex = Explanation {
2103 schema: rto_graph::SCHEMA,
2104 node: NodeSummary {
2105 key: "sym:rust:a.rs#main".into(),
2106 kind: "fn".into(),
2107 name: "main".into(),
2108 path: Some("a.rs".into()),
2109 lang: Some("rust".into()),
2110 },
2111 meta: serde_json::json!({ "content": "Entry point." }),
2112 outgoing: vec![],
2113 incoming: vec![],
2114 };
2115 assert!(
2116 render_note(&ex, None, None)
2117 .content
2118 .contains("## Content\n\nEntry point.")
2119 );
2120 }
2121
2122 /// Lines in the note's `## Content` section.
2123 fn content_lines(note: &str) -> usize {
2124 let body = note
2125 .split_once("## Content\n\n")
2126 .map_or("", |(_, rest)| rest);
2127 let body = body.split_once("\n## ").map_or(body, |(head, _)| head);
2128 body.trim_end().lines().count()
2129 }
2130
2131 #[test]
2132 fn render_note_shows_inferred_confidence() {
2133 let ex = Explanation {
2134 schema: rto_graph::SCHEMA,
2135 node: NodeSummary {
2136 key: "file:a.md".into(),
2137 kind: "file".into(),
2138 name: "a.md".into(),
2139 path: Some("a.md".into()),
2140 lang: None,
2141 },
2142 meta: serde_json::Value::Null,
2143 outgoing: vec![EdgeRef {
2144 kind: "related".into(),
2145 provenance: "inferred",
2146 confidence: Some(0.82),
2147 node: "file:b.md".into(),
2148 }],
2149 incoming: vec![],
2150 };
2151 let note = render_note(&ex, None, None);
2152 assert!(
2153 note.content.contains(&format!(
2154 "related (inferred) (0.82) → [[{}]]",
2155 note_name("file:b.md")
2156 )),
2157 "{}",
2158 note.content
2159 );
2160 }
2161
2162 /// The single-project `_Home` fixture, extracted so the test that reads its
2163 /// rendering stays about the rendering. Every field is populated: the
2164 /// sections it drives are individually omitted when empty, so a partial
2165 /// fixture would silently stop exercising them.
2166 fn home_summary() -> VaultSummary {
2167 VaultSummary {
2168 project: "demo".into(),
2169 total_nodes: 3,
2170 total_edges: 2,
2171 node_counts: vec![("fn".into(), 2), ("adr".into(), 1)],
2172 edge_provenance: vec![("derived".into(), 1), ("authored".into(), 1)],
2173 adrs: vec![AdrEntry {
2174 key: "adr:0001".into(),
2175 name: "First".into(),
2176 status: Some("Accepted".into()),
2177 }],
2178 debt: vec![("todo".into(), 4)], // roteiro:ignore
2179 densest_files: vec![DensityEntry {
2180 path: "src/small.rs".into(),
2181 markers: 3,
2182 lines: 120,
2183 per_kloc: 25.0,
2184 }],
2185 config_secrets: Some(ConfigSecretSummary {
2186 secret_named: 4,
2187 redacted: 3,
2188 declared: 1,
2189 unredacted: 0,
2190 files: vec![".env".into()],
2191 }),
2192 most_called: vec![CouplingEntry {
2193 key: "sym:rust:a.rs#helper".into(),
2194 name: "helper".into(),
2195 fan_in: 7,
2196 fan_out: 1,
2197 }],
2198 repo_url: Some("https://github.com/org/repo".into()),
2199 commit: Some("abcdef0123456789".into()),
2200 findings: vec![],
2201 coverage: Coverage::NotRun,
2202 settings: RenderedUnder::default(),
2203 }
2204 }
2205
2206 #[test]
2207 fn render_home_summarises_the_graph() {
2208 let summary = home_summary();
2209 let note = render_home(&summary);
2210 assert_eq!(note.filename, HOME_NOTE);
2211 assert!(note.content.contains("# demo — knowledge graph"));
2212 assert!(note.content.contains("**3 nodes**, **2 edges**"));
2213 assert!(note.content.contains("| fn | 2 |"));
2214 assert!(note.content.contains("| derived | 1 |"));
2215 assert!(note.content.contains(&format!(
2216 "**Accepted** — [[{}|First]]",
2217 note_name("adr:0001")
2218 )));
2219 assert!(note.content.contains("| todo | 4 |")); // roteiro:ignore
2220 // Directed coupling: the two fans are separate columns, and the wikilink's
2221 // own `|` is escaped so it cannot break the table it sits in.
2222 assert!(
2223 note.content.contains(&format!(
2224 "| [[{}\\|helper]] | 7 | 1 |",
2225 note_name("sym:rust:a.rs#helper")
2226 )),
2227 "{}",
2228 note.content
2229 );
2230 assert!(
2231 note.content.contains("resolved by simple name"),
2232 "the precision caveat travels with the figures"
2233 );
2234 // Density: the count and the denominator are both shown, so the ratio can
2235 // be checked rather than taken on trust, and the wikilink's own `|` is
2236 // escaped so it cannot break the table it sits in.
2237 assert!(
2238 note.content.contains(&format!(
2239 "| [[{}\\|src/small.rs]] | 3 | 120 | 25.00 |",
2240 note_name("file:src/small.rs")
2241 )),
2242 "{}",
2243 note.content
2244 );
2245 assert!(
2246 note.content.contains("not source lines of code"),
2247 "the denominator caveat travels with the figures"
2248 );
2249 // Config secrets: counts and files, and no key names — a vault note is
2250 // browsed out of context, which is the wrong place for a list that would
2251 // read as a secret scan's output.
2252 assert!(
2253 note.content.contains(
2254 "**4** secret-named config key(s): 3 redacted before storage, 1 \
2255 declared in code without a value, 0 unredacted."
2256 ),
2257 "{}",
2258 note.content
2259 );
2260 assert!(
2261 note.content
2262 .contains(&format!("- [[{}\\|.env]]", note_name("file:.env"))),
2263 "{}",
2264 note.content
2265 );
2266 assert!(
2267 note.content.contains("not a secret scan")
2268 && note.content.contains("cannot see a hardcoded credential"),
2269 "the limitation travels with the figures: {}",
2270 note.content
2271 );
2272 assert!(
2273 !note.content.contains("[!warning]"),
2274 "no warning when nothing is unredacted: {}",
2275 note.content
2276 );
2277 // A repository link + short-commit permalink note.
2278 assert!(
2279 note.content
2280 .contains("**Repository:** [https://github.com/org/repo](https://github.com/org/repo) · rendered at commit `abcdef012345`"),
2281 "{}",
2282 note.content
2283 );
2284 }
2285
2286 #[test]
2287 fn render_home_omits_density_for_a_graph_with_no_markers() {
2288 // A clean repository has no markers, so there is no density to rank. An
2289 // empty table under a heading reads as "measured, and there is nothing";
2290 // the section is absent instead. Same rule as the coupling table below.
2291 let note = render_home(&VaultSummary {
2292 project: "clean".into(),
2293 total_nodes: 1,
2294 ..VaultSummary::default()
2295 });
2296 assert!(
2297 !note.content.contains("Densest files"),
2298 "no heading without rows: {}",
2299 note.content
2300 );
2301 // The intent-debt section itself still renders — density is an addition
2302 // to it, not a replacement.
2303 assert!(note.content.contains("## Intent debt"));
2304 assert!(note.content.contains("*None recorded.*"));
2305 }
2306
2307 #[test]
2308 fn render_home_omits_config_secrets_rather_than_rendering_zeroes() {
2309 // A row of zeroes under this heading would read as "scanned, and clean" —
2310 // a conclusion the lens cannot support, since a credential under an
2311 // innocuous key name never appears in it. The section is absent instead.
2312 let note = render_home(&VaultSummary {
2313 project: "clean".into(),
2314 total_nodes: 1,
2315 ..VaultSummary::default()
2316 });
2317 assert!(
2318 !note.content.contains("named like secrets"),
2319 "no heading without figures: {}",
2320 note.content
2321 );
2322 }
2323
2324 #[test]
2325 fn render_home_warns_loudly_about_an_unredacted_value() {
2326 // Extraction cannot produce this state, so if it appears something else
2327 // put an unredacted value in the store — and the note must say where to
2328 // look rather than implicating the repository.
2329 let note = render_home(&VaultSummary {
2330 project: "imported".into(),
2331 total_nodes: 1,
2332 config_secrets: Some(ConfigSecretSummary {
2333 secret_named: 1,
2334 redacted: 0,
2335 declared: 0,
2336 unredacted: 1,
2337 files: vec!["imported.env".into()],
2338 }),
2339 ..VaultSummary::default()
2340 });
2341 assert!(
2342 note.content.contains("[!warning]") && note.content.contains("**unredacted**"),
2343 "{}",
2344 note.content
2345 );
2346 assert!(
2347 note.content.contains("came from an import layer"),
2348 "and it points at the importing tool, not the repository: {}",
2349 note.content
2350 );
2351 }
2352
2353 #[test]
2354 fn render_home_omits_coupling_for_a_graph_with_no_calls() {
2355 // A prose-only vault has no `calls` edges. An empty table under a heading
2356 // reads as "measured, and there is nothing" — the section is absent instead.
2357 let note = render_home(&VaultSummary {
2358 project: "docs".into(),
2359 total_nodes: 1,
2360 ..VaultSummary::default()
2361 });
2362 assert!(
2363 !note.content.contains("Most depended-on"),
2364 "no heading without rows: {}",
2365 note.content
2366 );
2367 // The rest of the overview is unaffected.
2368 assert!(note.content.contains("# docs — knowledge graph"));
2369 }
2370
2371 // ---- Workspace vaults (issue #442 part 1) --------------------------------
2372
2373 /// A `Explanation` for `key`, with one outgoing edge to `to`.
2374 fn node_linking_to(key: &str, name: &str, to: &str) -> Explanation {
2375 Explanation {
2376 schema: rto_graph::SCHEMA,
2377 node: NodeSummary {
2378 key: key.into(),
2379 kind: "config_key".into(),
2380 name: name.into(),
2381 path: Some("config.toml".into()),
2382 lang: None,
2383 },
2384 meta: serde_json::Value::Null,
2385 outgoing: vec![EdgeRef {
2386 kind: "links".into(),
2387 provenance: "inferred",
2388 confidence: Some(0.91),
2389 node: to.into(),
2390 }],
2391 incoming: vec![],
2392 }
2393 }
2394
2395 fn members(names: &[&str]) -> std::collections::BTreeSet<String> {
2396 names.iter().map(|s| (*s).to_owned()).collect()
2397 }
2398
2399 /// **Rewritten deliberately under #574.** #570 landed this as "a project
2400 /// scope leaves every note name exactly as it was", and read that two ways at
2401 /// once: `PROJECT` reduces to `note_name`, *and* `note_name` itself does not
2402 /// move. #574 breaks the second half on purpose — the old names were not
2403 /// injective under filename case folding and this repository's vault lost 104
2404 /// notes to it — so the two halves are separated here rather than having
2405 /// expected values quietly updated underneath the old title.
2406 ///
2407 /// What survives is the half #570 was actually about, and it is unweakened:
2408 /// **turning workspace mode on must not rename a project's notes.** Names may
2409 /// move when `note_name` changes, for a reason argued at `note_name`; they may
2410 /// never move because a repository happens to sit inside a configured
2411 /// workspace, because that would happen by inference rather than by a release.
2412 ///
2413 /// The other half of #570's promise — that a project render is byte-identical
2414 /// apart from names — is now [`render_note_is_the_project_scoped_render_byte_for_byte`]
2415 /// and `render_cli`'s end-to-end pair.
2416 #[test]
2417 fn a_project_scope_never_qualifies_a_name() {
2418 // A user's own notes live outside the vault and link into it *by name*
2419 // (#442), so a rename breaks them silently, with no error and nothing to
2420 // grep for. Whatever workspace mode does, `VaultScope::PROJECT` must
2421 // reduce to `note_name` of the bare key.
2422 for key in [
2423 "file:README.md",
2424 "adr:0001",
2425 "sym:rust:src/a.rs#Store",
2426 "extref:other::file:README.md",
2427 "cfgkey:config.toml#serve.addr",
2428 ] {
2429 assert_eq!(
2430 scoped_note_name(&VaultScope::PROJECT, key),
2431 note_name(key),
2432 "single-project name moved for `{key}`"
2433 );
2434 // And the qualified form really is a different name, so the assertion
2435 // above is not vacuously true of every scope.
2436 let ms = members(&["app"]);
2437 assert_ne!(
2438 scoped_note_name(
2439 &VaultScope {
2440 project: Some("app"),
2441 members: &ms,
2442 },
2443 key
2444 ),
2445 note_name(key),
2446 "qualification must move the name for `{key}`, or nothing above holds"
2447 );
2448 }
2449 }
2450
2451 #[test]
2452 fn render_note_is_the_project_scoped_render_byte_for_byte() {
2453 let ex = node_linking_to("cfgkey:config.toml#addr", "addr", "sym:rust:a.rs#A");
2454 assert_eq!(
2455 render_note(&ex, Some("https://h/b"), Some("body")),
2456 render_note_scoped(&ex, Some("https://h/b"), Some("body"), &VaultScope::PROJECT),
2457 "the unscoped entry point must stay the scoped one at PROJECT, so the \
2458 two cannot drift apart"
2459 );
2460 }
2461
2462 #[test]
2463 fn each_member_gets_its_own_note_for_the_same_key() {
2464 // The collision the whole feature exists for: node keys are
2465 // repository-relative, so every member's `README.md` is `file:README.md`.
2466 let ms = members(&["api", "sdk"]);
2467 let names: Vec<String> = ["api", "sdk"]
2468 .iter()
2469 .map(|p| {
2470 scoped_note_name(
2471 &VaultScope {
2472 project: Some(p),
2473 members: &ms,
2474 },
2475 "file:README.md",
2476 )
2477 })
2478 .collect();
2479 assert_eq!(
2480 names,
2481 [
2482 note_name("api::file:README.md"),
2483 note_name("sdk::file:README.md")
2484 ]
2485 );
2486 assert_ne!(names[0], names[1], "two members must not share one note");
2487 }
2488
2489 /// The two names this feature has, pinned together in one place.
2490 ///
2491 /// They are easy to conflate and were, in this PR, described inconsistently
2492 /// in two doc comments — the **key** is `<project>::<key>` (ADR-0009's
2493 /// cross-repo form, which is why cross-repo links resolve), and the **note
2494 /// name** is [`note_name`] of that key, in which `::` has become `-`. A
2495 /// reader told the wrong one goes looking for a file with `::` in it.
2496 ///
2497 /// Asserting both here means the next description that drifts has something
2498 /// to disagree with, rather than waiting for a reviewer to read two comments
2499 /// side by side.
2500 #[test]
2501 fn the_qualified_key_and_the_note_name_are_different_strings() {
2502 let ms = members(&["app"]);
2503 let scope = VaultScope {
2504 project: Some("app"),
2505 members: &ms,
2506 };
2507 // The key: project-qualified, `::` intact — this is what the graph and
2508 // ADR-0009's external refs use.
2509 let qualified = "app::file:README.md";
2510 // The note name: `note_name` of exactly that key, `::` slugged to `-`,
2511 // the whole hint lowercased, and the key's own hash appended.
2512 assert_eq!(
2513 scoped_note_name(&scope, "file:README.md"),
2514 "app-file-readme.md-a114bde6dcaba1c1"
2515 );
2516 assert_eq!(note_name(qualified), "app-file-readme.md-a114bde6dcaba1c1");
2517 assert!(
2518 !scoped_note_name(&scope, "file:README.md").contains("::"),
2519 "no note name ever contains `::`"
2520 );
2521 // And on disk the stem gains the extension, which is the string a reader
2522 // actually looks for.
2523 let note = render_note_scoped(
2524 &node_with("file:README.md", Some("README.md"), None),
2525 None,
2526 None,
2527 &scope,
2528 );
2529 assert_eq!(note.filename, "app-file-readme.md-a114bde6dcaba1c1.md");
2530 }
2531
2532 /// `_Home` must *show* a name, not spell the form out.
2533 ///
2534 /// The test above pins the distinction in the code. It did not stop the
2535 /// distinction being described wrongly in the same file, because it guards
2536 /// the function and not the sentences: `render_workspace_home` went on
2537 /// writing the pre-#574 form into the `_Home` of every workspace vault
2538 /// v2.0.0 built, and nothing here disagreed with it.
2539 ///
2540 /// So this asserts the property that made that possible is gone — the
2541 /// paragraph now contains a string `note_name` actually produced for a key
2542 /// the workspace really holds, which a hand-written spelling cannot
2543 /// satisfy. It is not a tautology despite both sides calling `note_name`:
2544 /// what it rejects is the *shape* of the old copy, a form written out by
2545 /// hand next to the function that could have rendered it.
2546 ///
2547 /// That the *key* is real is the other half, and the reason the example is
2548 /// drawn from `cross_links` rather than invented from the member list —
2549 /// a name rendered for a node the vault does not hold is a true sentence
2550 /// about a note nobody can open. The empty case is
2551 /// `the_workspace_home_claims_no_example_note_when_it_has_no_real_key`.
2552 fn finding(rule: &str, severity: &str, message: &str) -> FindingEntry {
2553 FindingEntry {
2554 rule: rule.to_owned(),
2555 severity: severity.to_owned(),
2556 title: "a title".to_owned(),
2557 message: message.to_owned(),
2558 path: Some("Cargo.lock".to_owned()),
2559 analyzer: "osv-scanner".to_owned(),
2560 }
2561 }
2562
2563 fn ws_with(members: Vec<VaultSummary>) -> WorkspaceSummary {
2564 WorkspaceSummary {
2565 generated_at: "2026-08-22T10:00:00Z".to_owned(),
2566 name: "platform".into(),
2567 members,
2568 pins: Vec::new(),
2569 cross_links: vec![],
2570 cross_links_total: 0,
2571 cross_links_authored: 0,
2572 }
2573 }
2574
2575 /// An analyzer message is **tool output** being embedded in a document that
2576 /// gets handed to people. It must be quoted, never absorbed: a message
2577 /// carrying its own fence would otherwise close the block early and spill the
2578 /// rest into the note as vault prose — headings and links included.
2579 #[test]
2580 fn an_analyzer_message_cannot_break_out_of_its_fence() {
2581 let mut api = member_summary("api", 1);
2582 api.coverage = Coverage::Ran(vec![("osv-scanner".into(), "1.9.0".into())]);
2583 api.findings = vec![finding(
2584 "GHSA-x",
2585 "high",
2586 "before
2587```
2588## not a vault heading
2589```
2590after",
2591 )];
2592 let c = render_workspace_home(&ws_with(vec![api])).content;
2593
2594 // The fence opened must be longer than any run inside the message, so the
2595 // message's own ``` is content rather than a terminator.
2596 assert!(
2597 c.contains("````text"),
2598 "fence widened past the message's own: {c}"
2599 );
2600 assert!(
2601 c.contains("## not a vault heading"),
2602 "and the text is still all there: {c}"
2603 );
2604 // Nothing between the finding and the next section may be loose prose.
2605 let after = c.split("````text").nth(1).expect("a fenced block");
2606 let body = after.split("````").next().expect("a closing fence");
2607 assert!(
2608 body.contains("## not a vault heading"),
2609 "the heading is INSIDE the fence, not outside it: {body}"
2610 );
2611 }
2612
2613 /// The fence secured the **message**. Every other analyzer-sourced field on a
2614 /// finding is interpolated into a line of prose, and they come from the same
2615 /// place — so a title carrying a newline used to end its line and start a new
2616 /// block, which can open a heading the vault never wrote.
2617 #[test]
2618 fn every_analyzer_sourced_field_is_quoted_not_absorbed() {
2619 let mut api = member_summary("api", 1);
2620 api.coverage = Coverage::Ran(vec![("osv-scanner".into(), "1.9.0".into())]);
2621 api.findings = vec"
2628 .to_owned(),
2629 message: "fine".to_owned(),
2630 path: Some("src/a`b.rs".to_owned()),
2631 analyzer: "osv-scanner".to_owned(),
2632 }];
2633 let c = render_workspace_home(&ws_with(vec![api])).content;
2634
2635 assert!(
2636 !c.contains("\n## an injected heading"),
2637 "a newline in a title must not start a block: {c}"
2638 );
2639 assert!(
2640 !c.contains("[a link](http://evil)"),
2641 "and a link must not survive as a link: {c}"
2642 );
2643 // The text is still readable — escaped, not deleted. Losing it would be a
2644 // different failure: a finding whose title vanished is a finding nobody
2645 // acts on.
2646 assert!(c.contains("an injected heading"), "the words remain: {c}");
2647 // A widened delimiter, not a backslash: inside a span the backslash
2648 // would show, and `vendor/**` would reach the reader as `vendor\*\*`.
2649 assert!(
2650 c.contains("``R-1`x``"),
2651 "the span widens past the backtick inside it: {c}"
2652 );
2653
2654 // The structural property, asserted directly rather than inferred from
2655 // the absence of a heading: the whole finding stays on **one line**. An
2656 // escape set can neutralise `#` and `[`, and a newline would still end
2657 // the line and start a block — and nothing escapes a `-` list marker.
2658 let line = c
2659 .lines()
2660 .find(|l| l.contains("R-1"))
2661 .unwrap_or_else(|| panic!("the finding renders: {c}"));
2662 assert!(
2663 line.contains("a link") && line.contains("a list item"),
2664 "every part of the title stays on the finding's own line: {line}"
2665 );
2666 }
2667
2668 /// The analyzer name is interpolated into **raw HTML** (`<sub>`,
2669 /// `<details><summary>`), where a backslash escapes nothing. Fixing the
2670 /// Markdown surface and leaving this one is how the first version of the
2671 /// escaping shipped.
2672 #[test]
2673 fn analyzer_text_cannot_break_out_of_the_html_it_sits_in() {
2674 let mut api = member_summary("api", 1);
2675 let hostile = "osv</sub><script>alert(1)</script>".to_owned();
2676 api.coverage = Coverage::Ran(vec![(hostile.clone(), "1.0".into())]);
2677 api.findings = vec![FindingEntry {
2678 rule: "R-1".to_owned(),
2679 severity: "high".to_owned(),
2680 title: "t".to_owned(),
2681 message: "m".to_owned(),
2682 path: None,
2683 analyzer: hostile,
2684 }];
2685 let c = render_workspace_home(&ws_with(vec![api])).content;
2686
2687 assert!(
2688 !c.contains("<script>"),
2689 "a tag in analyzer text must not survive as a tag: {c}"
2690 );
2691 assert!(
2692 !c.contains("osv</sub>"),
2693 "and must not close the element it was placed inside: {c}"
2694 );
2695 // Escaped, not dropped: the reader still sees which analyzer said it.
2696 assert!(c.contains("<script>"), "the text remains, inert: {c}");
2697 }
2698
2699 /// `[debt] ignore` is user config, not analyzer output — but the vault did
2700 /// not write it either, and it lands in a table cell where a `|` reshapes
2701 /// the row.
2702 #[test]
2703 fn a_config_glob_cannot_reshape_the_settings_table() {
2704 let mut api = member_summary("api", 1);
2705 api.settings = RenderedUnder {
2706 ingest: vec!["prose".into()],
2707 debt_ignore: vec!["a`b|c".into()],
2708 };
2709 let c = render_workspace_home(&ws_with(vec![api])).content;
2710 // Scoped to the settings table. The manifest table above it also has a
2711 // row starting `| api |`, and a bare `.find` matched *that* one — so the
2712 // first version of this test asserted against a row containing no glob
2713 // at all, and stayed green with the escaping removed.
2714 let section = c
2715 .split("### Rendered under")
2716 .nth(1)
2717 .unwrap_or_else(|| panic!("the settings section: {c}"));
2718 let row = section
2719 .lines()
2720 .find(|l| l.starts_with("| api |"))
2721 .unwrap_or_else(|| panic!("a settings row: {section}"));
2722 // Count **cell delimiters**, not `|` characters: an escaped `\|` is
2723 // content, and GFM does not split on it. Counting raw pipes would fail
2724 // on the correct output and pass on some wrong ones.
2725 let delimiters = row
2726 .char_indices()
2727 .filter(|&(i, ch)| ch == '|' && !row[..i].ends_with('\\'))
2728 .count();
2729 assert_eq!(
2730 delimiters, 4,
2731 "the row keeps exactly its own four cell delimiters: {row}"
2732 );
2733 }
2734
2735 /// The manifest is a table too. `table_cell` was written for the settings
2736 /// table and not applied to this one, thirty lines above it — a remote URL
2737 /// and a project name both come from git rather than from us.
2738 #[test]
2739 fn a_remote_url_cannot_reshape_the_manifest_row() {
2740 let mut api = member_summary("api", 1);
2741 api.repo_url = Some("https://host/a|b/c".to_owned());
2742 api.commit = Some("dead`beef".to_owned());
2743 let c = render_workspace_home(&ws_with(vec![api])).content;
2744
2745 let section = c
2746 .split("## Reproducing this vault")
2747 .nth(1)
2748 .unwrap_or_else(|| panic!("the manifest: {c}"));
2749 let row = section
2750 .lines()
2751 .find(|l| l.starts_with("| api |"))
2752 .unwrap_or_else(|| panic!("a manifest row: {section}"));
2753 let delimiters = row
2754 .char_indices()
2755 .filter(|&(i, ch)| ch == '|' && !row[..i].ends_with('\\'))
2756 .count();
2757 assert_eq!(
2758 delimiters, 4,
2759 "the row keeps exactly its own four cell delimiters: {row}"
2760 );
2761 assert!(
2762 row.contains("``dead`beef``"),
2763 "and a backtick in a sha widens its span: {row}"
2764 );
2765 }
2766
2767 /// Escaping `[` stops explicit link syntax and nothing else. Obsidian and
2768 /// GFM linkify a bare URL, so analyzer text could still hand the reader
2769 /// something clickable in a document they were given.
2770 #[test]
2771 fn a_bare_url_in_analyzer_text_is_not_left_clickable() {
2772 let mut api = member_summary("api", 1);
2773 api.coverage = Coverage::Ran(vec![("osv-scanner".into(), "1.9.0".into())]);
2774 api.findings = vec![FindingEntry {
2775 rule: "R-1".to_owned(),
2776 severity: "high".to_owned(),
2777 title: "see https://evil.example/path for details".to_owned(),
2778 message: "m".to_owned(),
2779 path: None,
2780 analyzer: "osv-scanner".to_owned(),
2781 }];
2782 let c = render_workspace_home(&ws_with(vec![api])).content;
2783
2784 assert!(
2785 !c.contains("https://evil.example"),
2786 "a bare URL must not survive in linkifiable form: {c}"
2787 );
2788 // Rendered identically for a reader, and still copyable — defusing it
2789 // must not amount to hiding it.
2790 assert!(c.contains("https://evil.example/path"), "{c}");
2791 }
2792
2793 /// The two empty states are opposite facts and must never render alike.
2794 #[test]
2795 fn an_unanalyzed_member_never_reads_as_one_that_came_back_clean() {
2796 let mut looked = member_summary("api", 1);
2797 looked.coverage = Coverage::Ran(vec![("osv-scanner".into(), "1.9.0".into())]);
2798 let never = member_summary("sdk", 1); // Coverage::NotRun by default
2799 let c = render_workspace_home(&ws_with(vec![looked, never])).content;
2800
2801 assert!(c.contains("### api — no findings"), "{c}");
2802 assert!(
2803 c.contains("osv-scanner` 1.9.0 ran and reported none"),
2804 "{c}"
2805 );
2806 assert!(c.contains("### sdk — **not analyzed**"), "{c}");
2807 assert!(
2808 !c.contains("### sdk — no findings"),
2809 "an unanalyzed member must never be rendered as clean: {c}"
2810 );
2811 assert!(c.contains("1 never analyzed at all"), "counted too: {c}");
2812 }
2813
2814 /// The severity a finding carries is the **analyzer's** word. An unrecognised
2815 /// level is rendered as given rather than mapped onto a known rung, because
2816 /// mapping it would be inventing a judgement the tool did not make.
2817 #[test]
2818 fn an_unrecognised_severity_keeps_the_analyzers_own_label() {
2819 let mut api = member_summary("api", 1);
2820 api.coverage = Coverage::Ran(vec![("semgrep".into(), "1.2.3".into())]);
2821 api.findings = vec![finding("R-1", "WARNING", "msg")];
2822 let c = render_workspace_home(&ws_with(vec![api])).content;
2823 assert!(c.contains("**`WARNING`**"), "{c}");
2824 }
2825
2826 /// #442: the manifest records what each member **deploys**, not only the
2827 /// commit it was rendered at — and each pin names the hub it is relative to.
2828 ///
2829 /// A snowflake fixture on purpose (`infra → chart → app`), because a star
2830 /// would let a bare-revision column pass: it is only when a workspace has two
2831 /// levels that "the hub version" stops having one answer. And deliberately
2832 /// mixed — one dependency resolves, one does not — so the row that reports
2833 /// *finding nothing* is exercised rather than assumed.
2834 #[test]
2835 fn the_manifest_records_each_members_pin_and_which_hub_it_is_relative_to() {
2836 let ws = WorkspaceSummary {
2837 generated_at: "2026-08-26T10:00:00Z".to_owned(),
2838 name: "platform".into(),
2839 members: vec![
2840 member_summary("infra", 3),
2841 member_summary("chart", 4),
2842 member_summary("app", 9),
2843 ],
2844 pins: vec![
2845 MemberPin {
2846 member: "chart".into(),
2847 hub: "app".into(),
2848 rev: Some("1.4.0".into()),
2849 via: Some("image acme/app:1.4.0".into()),
2850 },
2851 MemberPin {
2852 member: "infra".into(),
2853 hub: "chart".into(),
2854 rev: None,
2855 via: None,
2856 },
2857 ],
2858 cross_links: vec![],
2859 cross_links_total: 0,
2860 cross_links_authored: 0,
2861 };
2862 let c = render_workspace_home(&ws).content;
2863
2864 assert!(c.contains("### Version pins"), "{c}");
2865 // The hub is named per row. Without this the table is ambiguous the moment
2866 // a workspace has more than one level — which is exactly this fixture.
2867 assert!(
2868 c.contains("| chart | app | `1.4.0` |") || c.contains("| chart | app |"),
2869 "the pin names its hub: {c}"
2870 );
2871 assert!(
2872 c.contains("image acme/app:1.4.0"),
2873 "says where it read it: {c}"
2874 );
2875
2876 // The unresolved dependency is REPORTED, not dropped. #505's rule: asked
2877 // and found nothing must not render as never asked.
2878 assert!(c.contains("| infra | chart |"), "the row survives: {c}");
2879 assert!(c.contains("(none detected)"), "and says so: {c}");
2880
2881 // The count separates what was asked from what was found.
2882 assert!(
2883 c.contains("1 of 2 dependencies resolve"),
2884 "counts found against asked: {c}"
2885 );
2886 }
2887
2888 /// A one-dependency workspace reads as English.
2889 ///
2890 /// Varying the noun and fixing the verb produced "1 of 1 dependency resolve",
2891 /// which is the failure mode of pluralising by hand: the two halves have to
2892 /// agree with the same count, not one of them with the count and the other
2893 /// with a guess. Two-member workspaces are common enough that this renders.
2894 #[test]
2895 fn the_pins_caption_agrees_with_a_single_dependency() {
2896 let ws = WorkspaceSummary {
2897 generated_at: "2026-08-26T10:00:00Z".to_owned(),
2898 name: "pair".into(),
2899 members: vec![member_summary("deploy", 2), member_summary("app", 5)],
2900 pins: vec![MemberPin {
2901 member: "deploy".into(),
2902 hub: "app".into(),
2903 rev: Some("1.4.0".into()),
2904 via: Some("image acme/app:1.4.0".into()),
2905 }],
2906 cross_links: vec![],
2907 cross_links_total: 0,
2908 cross_links_authored: 0,
2909 };
2910 let c = render_workspace_home(&ws).content;
2911 assert!(
2912 c.contains("1 of 1 dependency resolves"),
2913 "singular noun and singular verb: {c}"
2914 );
2915 }
2916
2917 /// A workspace whose members depend on nothing has no pins section at all —
2918 /// rather than an empty table, which reads as a fault.
2919 #[test]
2920 fn a_workspace_with_no_dependencies_has_no_pins_table() {
2921 let ws = WorkspaceSummary {
2922 generated_at: "2026-08-26T10:00:00Z".to_owned(),
2923 name: "solo".into(),
2924 members: vec![member_summary("only", 1)],
2925 pins: Vec::new(),
2926 cross_links: vec![],
2927 cross_links_total: 0,
2928 cross_links_authored: 0,
2929 };
2930 let c = render_workspace_home(&ws).content;
2931 assert!(!c.contains("Version pins"), "{c}");
2932 }
2933
2934 /// #442 part 2: the vault says how to reconstruct the workspace it describes,
2935 /// and what it deliberately leaves out.
2936 ///
2937 /// A vault that says *"here is my workspace"* is far less useful than one
2938 /// that says *"here is my workspace at these commits"* — and the exclusions
2939 /// are stated at the point a reader is most likely to share it, because that
2940 /// is when they matter.
2941 #[test]
2942 fn the_workspace_home_says_how_to_reproduce_itself_and_what_it_omits() {
2943 let mut api = member_summary("api", 7);
2944 api.repo_url = Some("https://github.com/acme/api".to_owned());
2945 api.commit = Some("4e0d5a6afd0b1c2d".to_owned());
2946 // Deliberately **mixed**: one member reproducible, one not. A fixture where
2947 // every member has a remote never exercises the row that says so, and the
2948 // gap is the thing a reader most needs to see.
2949 let mut sdk = member_summary("sdk", 4);
2950 sdk.repo_url = None;
2951 sdk.commit = None;
2952 let ws = WorkspaceSummary {
2953 generated_at: "2026-08-22T10:00:00Z".to_owned(),
2954 name: "platform".into(),
2955 members: vec![api, sdk],
2956 pins: Vec::new(),
2957 cross_links: vec![],
2958 cross_links_total: 0,
2959 cross_links_authored: 0,
2960 };
2961 let c = render_workspace_home(&ws).content;
2962
2963 assert!(c.contains("2026-08-22T10:00:00Z"), "stamped: {c}");
2964 assert!(c.contains("read-only and point-in-time"), "{c}");
2965 assert!(c.contains("https://github.com/acme/api"), "clone-from: {c}");
2966 assert!(c.contains("4e0d5a6afd0b1c2d"), "pinned commit: {c}");
2967 // A member with no remote is named as such rather than omitted — a gap in
2968 // the manifest is the reader's problem to know about, not ours to hide.
2969 assert!(c.contains("no `origin` remote"), "{c}");
2970
2971 // The share-time warning, each item with its reason. Two of the three are
2972 // about what the vault *carries*, not what it omits — the owner ruled
2973 // findings in, so this section warns rather than reassures.
2974 assert!(
2975 c.contains("cannot be un-shared"),
2976 "including findings has a consequence, stated where it is acted on: {c}"
2977 );
2978 assert!(
2979 c.contains("Agent memory"),
2980 "the one genuine exclusion is still named: {c}"
2981 );
2982 assert!(
2983 c.contains("no redaction chokepoint"),
2984 "with its reason: {c}"
2985 );
2986 // The limit of what *is* included, which is narrower than it looks.
2987 assert!(
2988 c.contains("DATABASE_URL"),
2989 "the redaction gap is shown, not described: {c}"
2990 );
2991 }
2992
2993 /// The manifest records the settings a re-render must match, because a clone
2994 /// URL and a commit get a reader the same **source** and not the same
2995 /// **vault**: `[ingest] prose` off produces notes with no captured content,
2996 /// and a `[debt] ignore` glob means the debt figures on the page are already
2997 /// filtered.
2998 #[test]
2999 fn the_manifest_records_the_settings_a_re_render_would_have_to_match() {
3000 let mut api = member_summary("api", 1);
3001 api.settings = RenderedUnder {
3002 ingest: vec!["prose".into(), "pdf".into()],
3003 debt_ignore: vec!["vendor/**".into()],
3004 };
3005 // A member with everything off is a real state, and must not render as a
3006 // blank cell that reads as "unknown".
3007 let sdk = member_summary("sdk", 1);
3008 let c = render_workspace_home(&ws_with(vec![api, sdk])).content;
3009
3010 assert!(c.contains("### Rendered under"), "{c}");
3011 assert!(c.contains("`prose`, `pdf`"), "the enabled toggles: {c}");
3012 assert!(c.contains("`vendor/**`"), "the debt filter: {c}");
3013 assert!(
3014 c.contains("| sdk | *none* | *none* |"),
3015 "all-off is stated, not left blank: {c}"
3016 );
3017 assert!(
3018 c.contains("will not reproduce this document"),
3019 "and the section says why it is there: {c}"
3020 );
3021 }
3022
3023 /// A workspace of purely local repositories cannot be reconstructed, and the
3024 /// vault says so instead of rendering a table of blanks — which would read as
3025 /// a fault rather than as repositories that were never pushed anywhere.
3026 #[test]
3027 fn a_vault_with_no_remotes_says_it_cannot_be_reconstructed() {
3028 // `member_summary` gives every member a remote, so it must be cleared —
3029 // the fixture has to actually be the case it names.
3030 let mut api = member_summary("api", 1);
3031 api.repo_url = None;
3032 api.commit = None;
3033 let ws = WorkspaceSummary {
3034 generated_at: "2026-08-22T10:00:00Z".to_owned(),
3035 name: "local".into(),
3036 members: vec![api],
3037 pins: Vec::new(),
3038 cross_links: vec![],
3039 cross_links_total: 0,
3040 cross_links_authored: 0,
3041 };
3042 let c = render_workspace_home(&ws).content;
3043 assert!(c.contains("cannot be reconstructed"), "{c}");
3044 assert!(
3045 !c.contains("| Member | Repository |"),
3046 "no empty table: {c}"
3047 );
3048 }
3049
3050 /// #573: the cross-repo section distinguishes a **declaration** from a
3051 /// **match**, and says how many of each.
3052 ///
3053 /// Before authored links could be persisted, every row was a candidate and
3054 /// the section said so in one blanket caveat. That caveat is now false for
3055 /// declared rows, and an edge that exists but renders as a guess leaves
3056 /// ADR-0009's `authored → gold` path just as unreachable as no edge at all —
3057 /// so the rendering is part of the contract, not decoration.
3058 #[test]
3059 fn the_cross_repo_section_separates_declared_links_from_inferred_ones() {
3060 let link = |authored: bool, key: &str| CrossLink {
3061 from_project: "sdk".into(),
3062 from_key: format!("cfgkey:config.toml#{key}"),
3063 from_name: key.into(),
3064 kind: "references".into(),
3065 // A declaration carries no score by construction; a match does.
3066 confidence: if authored { None } else { Some(0.91) },
3067 to_qualified: format!("api::cfgkey:config.toml#{key}"),
3068 resolves: true,
3069 authored,
3070 };
3071 let ws = WorkspaceSummary {
3072 generated_at: "2026-08-22T10:00:00Z".to_owned(),
3073 name: "platform".into(),
3074 members: vec![member_summary("api", 7), member_summary("sdk", 4)],
3075 pins: Vec::new(),
3076 cross_links: vec![link(true, "addr"), link(false, "port")],
3077 // Totals deliberately **larger** than the two rows shown: the caption
3078 // is a statement about the workspace, and `cross_links` is a capped
3079 // view of it. Counting the rows would give 1 and 1 here and read as
3080 // correct — which is exactly the bug, invisible until a workspace
3081 // outgrows the cap.
3082 cross_links_total: 9,
3083 cross_links_authored: 4,
3084 };
3085 let c = render_workspace_home(&ws).content;
3086
3087 assert!(
3088 c.contains("**4 declared**"),
3089 "the caption counts the workspace, not the rows on screen: {c}"
3090 );
3091 assert!(c.contains("**5 inferred**"), "{c}");
3092 assert!(
3093 !c.contains("not authored facts"),
3094 "the blanket caveat is false once a declared row can appear: {c}"
3095 );
3096 // Per row: a declaration is marked as one, and a match keeps its score.
3097 assert!(c.contains("references *(declared)*"), "{c}");
3098 assert!(c.contains("references (0.91)"), "{c}");
3099 }
3100
3101 #[test]
3102 fn the_workspace_home_names_an_example_note_name_actually_produces() {
3103 let ws = WorkspaceSummary {
3104 generated_at: "2026-08-22T10:00:00Z".to_owned(),
3105 name: "platform".into(),
3106 members: vec![member_summary("api", 7), member_summary("sdk", 4)],
3107 pins: Vec::new(),
3108 cross_links: vec![CrossLink {
3109 from_project: "sdk".into(),
3110 from_key: "cfgkey:config.toml#addr".into(),
3111 from_name: "addr".into(),
3112 kind: "links".into(),
3113 confidence: Some(0.91),
3114 to_qualified: "api::cfgkey:config.toml#addr".into(),
3115 resolves: true,
3116 authored: false,
3117 }],
3118 cross_links_total: 1,
3119 cross_links_authored: 0,
3120 };
3121 let note = render_workspace_home(&ws);
3122
3123 // The source end of the first cross-repo link, rendered through the real
3124 // function. `from_project` is a member and `from_key` is one of its own
3125 // nodes, so this is a note the render writes rather than one the
3126 // sentence assumes.
3127 let expected = format!("{}.md", note_name("sdk::cfgkey:config.toml#addr"));
3128 assert!(
3129 note.content.contains(&expected),
3130 "the naming paragraph must show a real name ({expected}), not a \
3131 hand-written form:\n{}",
3132 note.content
3133 );
3134 // And the key form it is derived *from* is still stated, because that is
3135 // the half a reader needs to look a note up by its frontmatter.
3136 assert!(
3137 note.content.contains("`<project>::<key>`"),
3138 "{}",
3139 note.content
3140 );
3141 // No filename anywhere in the vault carries `::`.
3142 assert!(!expected.contains("::"), "{expected}");
3143 }
3144
3145 /// With no cross-repo links there is no key the renderer can prove is a
3146 /// node, so it must say nothing rather than fabricate one.
3147 ///
3148 /// The example this replaced was `<first member>::file:README.md`, invented
3149 /// from the member list — and membership does not require a README, so
3150 /// `_Home` could assert a note that was never written. That is the very
3151 /// defect this PR exists to fix, one remove away, so the empty case gets an
3152 /// assertion of its own rather than an assumption.
3153 #[test]
3154 fn the_workspace_home_claims_no_example_note_when_it_has_no_real_key() {
3155 let ws = WorkspaceSummary {
3156 generated_at: "2026-08-22T10:00:00Z".to_owned(),
3157 name: "platform".into(),
3158 members: vec![member_summary("api", 7), member_summary("sdk", 4)],
3159 pins: Vec::new(),
3160 cross_links: vec![],
3161 cross_links_total: 0,
3162 cross_links_authored: 0,
3163 };
3164 let note = render_workspace_home(&ws);
3165
3166 assert!(
3167 !note.content.contains("is the note "),
3168 "no cross-repo link means no provable key, so no `Here, X is the \
3169 note Y` claim:\n{}",
3170 note.content
3171 );
3172 // The fabricated form specifically: never emitted, with or without links.
3173 assert!(
3174 !note.content.contains("::file:README.md"),
3175 "{}",
3176 note.content
3177 );
3178 // The rule itself is still stated — only the illustration is absent.
3179 assert!(
3180 note.content.contains("`<project>::<key>`")
3181 && note.content.contains("no filename contains `::`"),
3182 "{}",
3183 note.content
3184 );
3185 }
3186
3187 #[test]
3188 fn a_member_note_declares_which_member_it_came_from() {
3189 let ms = members(&["api"]);
3190 let ex = node_linking_to("cfgkey:config.toml#addr", "addr", "sym:rust:a.rs#A");
3191 let note = render_note_scoped(
3192 &ex,
3193 None,
3194 None,
3195 &VaultScope {
3196 project: Some("api"),
3197 members: &ms,
3198 },
3199 );
3200 assert_eq!(
3201 note.filename,
3202 format!("{}.md", note_name("api::cfgkey:config.toml#addr"))
3203 );
3204 assert!(
3205 note.content.contains("project: \"api\""),
3206 "{}",
3207 note.content
3208 );
3209 assert!(
3210 note.content.contains("- roteiro/project/api"),
3211 "the tag is what filters the graph view to one repository: {}",
3212 note.content
3213 );
3214 // A within-member edge is qualified to the same member, not left bare.
3215 assert!(
3216 note.content
3217 .contains(&format!("→ [[{}]]", note_name("api::sym:rust:a.rs#A"))),
3218 "{}",
3219 note.content
3220 );
3221 }
3222
3223 #[test]
3224 fn a_project_note_declares_no_project() {
3225 let ex = node_linking_to("cfgkey:config.toml#addr", "addr", "sym:rust:a.rs#A");
3226 let note = render_note(&ex, None, None);
3227 assert!(!note.content.contains("project:"), "{}", note.content);
3228 assert!(
3229 !note.content.contains("roteiro/project/"),
3230 "a per-project vault would carry one constant on every note — and \
3231 adding it would change every note's bytes: {}",
3232 note.content
3233 );
3234 }
3235
3236 #[test]
3237 fn a_cross_repo_edge_links_straight_to_the_other_members_note() {
3238 // ADR-0009: the spoke's edge points at a *local placeholder* for the hub's
3239 // node, because store integrity needs both ends in one store. A workspace
3240 // vault holds both, so the link goes to the real note. No new edge — the
3241 // resolver already follows this placeholder at query time.
3242 let ms = members(&["spoke", "hub"]);
3243 let scope = VaultScope {
3244 project: Some("spoke"),
3245 members: &ms,
3246 };
3247 let ex = node_linking_to(
3248 "cfgkey:config.toml#addr",
3249 "addr",
3250 &rto_graph::external_ref_key("hub::cfgkey:config.toml#addr"),
3251 );
3252 let note = render_note_scoped(&ex, None, None, &scope);
3253 assert!(
3254 note.content.contains(&format!(
3255 "→ [[{}]]",
3256 note_name("hub::cfgkey:config.toml#addr")
3257 )),
3258 "the edge must land on the hub's own note: {}",
3259 note.content
3260 );
3261 assert!(
3262 !note.content.contains("extref"),
3263 "and never on the placeholder: {}",
3264 note.content
3265 );
3266 // The same rule decides that the placeholder is not written as a note, so
3267 // the two halves cannot disagree.
3268 assert!(
3269 scope.redirects_external_ref(&rto_graph::external_ref_key(
3270 "hub::cfgkey:config.toml#addr"
3271 ))
3272 );
3273 }
3274
3275 #[test]
3276 fn a_cross_repo_edge_out_of_the_workspace_keeps_its_placeholder() {
3277 // The target repo is not in this vault, so there is no note to point at.
3278 // Redirecting anyway would produce a link that resolves to nothing —
3279 // Obsidian shows that as merely unwritten, which is a worse lie than a
3280 // placeholder that honestly says "elsewhere".
3281 let ms = members(&["spoke"]);
3282 let scope = VaultScope {
3283 project: Some("spoke"),
3284 members: &ms,
3285 };
3286 let key = rto_graph::external_ref_key("elsewhere::cfgkey:config.toml#addr");
3287 assert!(!scope.redirects_external_ref(&key));
3288 let ex = node_linking_to("cfgkey:config.toml#addr", "addr", &key);
3289 let note = render_note_scoped(&ex, None, None, &scope);
3290 assert!(
3291 note.content.contains(&format!(
3292 "→ [[{}]]",
3293 note_name("spoke::extref:elsewhere::cfgkey:config.toml#addr")
3294 )),
3295 "{}",
3296 note.content
3297 );
3298 }
3299
3300 #[test]
3301 fn a_single_project_vault_never_redirects_an_external_ref() {
3302 // No members ⇒ nothing to resolve against, so today's vault keeps rendering
3303 // the placeholder exactly as it does now.
3304 let key = rto_graph::external_ref_key("hub::cfgkey:config.toml#addr");
3305 assert!(!VaultScope::PROJECT.redirects_external_ref(&key));
3306 assert_eq!(
3307 scoped_note_name(&VaultScope::PROJECT, &key),
3308 note_name(&key)
3309 );
3310 }
3311
3312 fn member_summary(project: &str, fan_in: u32) -> VaultSummary {
3313 VaultSummary {
3314 project: project.to_owned(),
3315 total_nodes: 3,
3316 total_edges: 2,
3317 node_counts: vec![("fn".into(), 2)],
3318 edge_provenance: vec![("derived".into(), 2)],
3319 adrs: vec![AdrEntry {
3320 key: "adr:0001".into(),
3321 name: "First".into(),
3322 status: Some("Accepted".into()),
3323 }],
3324 debt: vec![("todo".into(), 4)], // roteiro:ignore
3325 densest_files: vec![DensityEntry {
3326 path: "src/small.rs".into(),
3327 markers: 3,
3328 lines: 120,
3329 per_kloc: 25.0,
3330 }],
3331 config_secrets: None,
3332 most_called: vec![CouplingEntry {
3333 key: "sym:rust:a.rs#helper".into(),
3334 name: "helper".into(),
3335 fan_in,
3336 fan_out: 1,
3337 }],
3338 repo_url: Some(format!("https://github.com/org/{project}")),
3339 commit: Some("abcdef0123456789".into()),
3340 findings: vec![],
3341 coverage: Coverage::NotRun,
3342 settings: RenderedUnder::default(),
3343 }
3344 }
3345
3346 #[test]
3347 fn the_workspace_home_keeps_every_members_own_aggregates() {
3348 // The promise in issue #442: the existing per-project `_Home` view is a
3349 // *subset* of the workspace one, not a casualty of it. Someone who came for
3350 // their repository's coupling and debt tables must still find them —
3351 // not a workspace total that averages them away.
3352 let ws = WorkspaceSummary {
3353 generated_at: "2026-08-22T10:00:00Z".to_owned(),
3354 name: "platform".into(),
3355 members: vec![member_summary("api", 7), member_summary("sdk", 4)],
3356 pins: Vec::new(),
3357 cross_links: vec![],
3358 cross_links_total: 0,
3359 cross_links_authored: 0,
3360 };
3361 let note = render_workspace_home(&ws);
3362 assert_eq!(note.filename, HOME_NOTE);
3363 assert!(
3364 note.content
3365 .contains("# platform — workspace knowledge graph")
3366 );
3367 // Summed, and the members listed.
3368 assert!(
3369 note.content
3370 .contains("**6 nodes**, **4 edges** across **2** member")
3371 );
3372 assert!(note.content.contains("| [[#api\\|api]] | 3 | 2 |"));
3373
3374 for project in ["api", "sdk"] {
3375 assert!(
3376 note.content.contains(&format!("\n## {project}\n")),
3377 "each member gets its own section"
3378 );
3379 }
3380 // Today's sections, one level deeper, once per member.
3381 for section in [
3382 "### Structure",
3383 "### Provenance",
3384 "### Decisions (ADRs)",
3385 "### Intent debt",
3386 "#### Densest files",
3387 "### Most depended-on",
3388 ] {
3389 assert_eq!(
3390 note.content.matches(section).count(),
3391 2,
3392 "`{section}` must appear once per member: {}",
3393 note.content
3394 );
3395 }
3396 // And every link inside a member's section resolves within that member.
3397 assert!(note.content.contains(&format!(
3398 "**Accepted** — [[{}|First]]",
3399 note_name("api::adr:0001")
3400 )));
3401 assert!(note.content.contains(&format!(
3402 "**Accepted** — [[{}|First]]",
3403 note_name("sdk::adr:0001")
3404 )));
3405 assert!(note.content.contains(&format!(
3406 "[[{}\\|helper]] | 7 |",
3407 note_name("api::sym:rust:a.rs#helper")
3408 )));
3409 assert!(note.content.contains(&format!(
3410 "[[{}\\|src/small.rs]]",
3411 note_name("sdk::file:src/small.rs")
3412 )));
3413 }
3414
3415 #[test]
3416 fn the_workspace_home_renders_cross_repo_links_and_marks_the_ones_it_cannot_follow() {
3417 let ws = WorkspaceSummary {
3418 generated_at: "2026-08-22T10:00:00Z".to_owned(),
3419 name: "platform".into(),
3420 members: vec![member_summary("api", 7), member_summary("sdk", 4)],
3421 pins: Vec::new(),
3422 cross_links: vec![
3423 CrossLink {
3424 from_project: "sdk".into(),
3425 from_key: "cfgkey:config.toml#addr".into(),
3426 from_name: "addr".into(),
3427 kind: "links".into(),
3428 confidence: Some(0.91),
3429 to_qualified: "api::cfgkey:config.toml#addr".into(),
3430 resolves: true,
3431 authored: false,
3432 },
3433 CrossLink {
3434 from_project: "sdk".into(),
3435 from_key: "cfgkey:config.toml#other".into(),
3436 from_name: "other".into(),
3437 kind: "links".into(),
3438 confidence: None,
3439 to_qualified: "absent::cfgkey:config.toml#other".into(),
3440 resolves: false,
3441 authored: false,
3442 },
3443 ],
3444 cross_links_total: 2,
3445 cross_links_authored: 0,
3446 };
3447 let note = render_workspace_home(&ws);
3448 // Resolvable: a link to the other member's note, with its confidence.
3449 assert!(
3450 note.content.contains(&format!(
3451 "| [[{}\\|addr]] | sdk | [[{}\\|api::cfgkey:config.toml#addr]] | links (0.91) |",
3452 note_name("sdk::cfgkey:config.toml#addr"),
3453 note_name("api::cfgkey:config.toml#addr"),
3454 )),
3455 "{}",
3456 note.content
3457 );
3458 // Outside the workspace: stated as such, never as a wikilink — Obsidian
3459 // renders a link to a missing note as one that is merely unwritten.
3460 assert!(
3461 note.content
3462 .contains("`absent::cfgkey:config.toml#other` *(outside this workspace)*"),
3463 "{}",
3464 note.content
3465 );
3466 assert!(
3467 !note.content.contains("[[absent-"),
3468 "a dangling wikilink would read as a note someone forgot to write: {}",
3469 note.content
3470 );
3471 }
3472
3473 #[test]
3474 fn the_workspace_home_says_when_it_has_truncated_the_cross_links() {
3475 // A capped table that does not say it is capped reads as the whole set.
3476 let ws = WorkspaceSummary {
3477 generated_at: "2026-08-22T10:00:00Z".to_owned(),
3478 name: "platform".into(),
3479 members: vec![member_summary("api", 7)],
3480 pins: Vec::new(),
3481 cross_links: vec![CrossLink {
3482 from_project: "api".into(),
3483 from_key: "cfgkey:config.toml#addr".into(),
3484 from_name: "addr".into(),
3485 kind: "links".into(),
3486 confidence: None,
3487 to_qualified: "api::cfgkey:config.toml#addr".into(),
3488 resolves: true,
3489 authored: false,
3490 }],
3491 cross_links_total: 40,
3492 cross_links_authored: 0,
3493 };
3494 let note = render_workspace_home(&ws);
3495 assert!(note.content.contains("Showing 1 of 40"), "{}", note.content);
3496 assert!(note.content.contains("roteiro links --matrix"));
3497 }
3498
3499 #[test]
3500 fn a_workspace_with_no_cross_repo_links_says_why_rather_than_showing_nothing() {
3501 let ws = WorkspaceSummary {
3502 generated_at: "2026-08-22T10:00:00Z".to_owned(),
3503 name: "platform".into(),
3504 members: vec![member_summary("api", 7)],
3505 pins: Vec::new(),
3506 cross_links: vec![],
3507 cross_links_total: 0,
3508 cross_links_authored: 0,
3509 };
3510 let note = render_workspace_home(&ws);
3511 assert!(note.content.contains("## Cross-repo links"));
3512 assert!(
3513 note.content.contains("links --infer --write"),
3514 "an empty section must name what would fill it, or it reads as \
3515 \"these repos are unrelated\": {}",
3516 note.content
3517 );
3518 // Singular, because getting this wrong on a one-member workspace is the
3519 // kind of thing nobody notices until it ships.
3520 assert!(note.content.contains("**1** member repository."));
3521 }
3522
3523 // ---- YAML frontmatter escaping -------------------------------------------
3524
3525 /// Parse a note's frontmatter block with a **real** YAML parser and return
3526 /// `field`'s value, or the parse error.
3527 ///
3528 /// Every assertion below goes through this rather than checking the emitted
3529 /// bytes. An escaper that is wrong in a self-consistent way passes a
3530 /// byte-comparison — that is precisely how `"foo\bar"` survived: it looks
3531 /// exactly like what was asked for, and means something else.
3532 fn frontmatter_field(note: &str, field: &str) -> Result<Option<String>, String> {
3533 let block = note
3534 .strip_prefix("---\n")
3535 .and_then(|rest| rest.split_once("\n---\n"))
3536 .map(|(block, _)| block)
3537 .expect("note must open with a frontmatter block");
3538 let docs = yaml_rust2::YamlLoader::load_from_str(block).map_err(|e| e.to_string())?;
3539 Ok(docs[0][field].as_str().map(ToOwned::to_owned))
3540 }
3541
3542 /// A node whose key, path and language are whatever the test needs.
3543 fn node_with(key: &str, path: Option<&str>, lang: Option<&str>) -> Explanation {
3544 Explanation {
3545 schema: rto_graph::SCHEMA,
3546 node: NodeSummary {
3547 key: key.into(),
3548 kind: "fn".into(),
3549 name: "n".into(),
3550 path: path.map(ToOwned::to_owned),
3551 lang: lang.map(ToOwned::to_owned),
3552 },
3553 meta: serde_json::Value::Null,
3554 outgoing: vec![],
3555 incoming: vec![],
3556 }
3557 }
3558
3559 /// The three measured failure modes of the escaping this replaced, each
3560 /// asserted on the **parsed** value.
3561 ///
3562 /// Before the fix: `foo\bar` parsed back as `foo<BS>ar` (silently six
3563 /// characters, not seven), and the other two made the whole block
3564 /// unparseable — which in Obsidian costs the note *every* property, with no
3565 /// error shown.
3566 #[test]
3567 fn a_backslash_or_quote_in_a_path_still_parses_back_to_itself() {
3568 for path in [
3569 r"foo\bar", // `\b` was YAML's backspace escape: silent corruption
3570 r"foo\dir", // `\d` is not a YAML escape at all: parse error
3571 "say\"hi\".rs", // an unescaped `"` ended the scalar early: parse error
3572 r"a\\b",
3573 "trailing-backslash\\",
3574 ] {
3575 let note = render_note(&node_with("file:x", Some(path), None), None, None);
3576 assert_eq!(
3577 frontmatter_field(¬e.content, "path"),
3578 Ok(Some(path.to_owned())),
3579 "path {path:?} must round-trip"
3580 );
3581 }
3582 }
3583
3584 /// `key:` is not hypothetical for this: node keys already carry `:` and `#`,
3585 /// and a symbol name can contain a quotation mark.
3586 #[test]
3587 fn a_node_key_round_trips_whatever_punctuation_it_carries() {
3588 for key in [
3589 "sym:rust:src/a.rs#Store",
3590 r"sym:rust:src\weird.rs#Thing",
3591 "sym:rust:a.rs#say\"hi\"",
3592 "cfgkey:config.toml#serve.addr",
3593 ] {
3594 let note = render_note(&node_with(key, None, None), None, None);
3595 assert_eq!(
3596 frontmatter_field(¬e.content, "key"),
3597 Ok(Some(key.to_owned())),
3598 "key {key:?} must round-trip"
3599 );
3600 }
3601 // The old rule turned a `"` into an apostrophe, so the note reported a key
3602 // that was not the node's key — parseable, and wrong.
3603 let note = render_note(
3604 &node_with("sym:rust:a.rs#say\"hi\"", None, None),
3605 None,
3606 None,
3607 );
3608 assert!(
3609 !note.content.contains("say'hi'"),
3610 "a quotation mark must be escaped, not rewritten: {}",
3611 note.content
3612 );
3613 }
3614
3615 /// A member directory name is a path component, so it reaches the same rule.
3616 #[test]
3617 fn a_member_project_name_round_trips() {
3618 let ms: std::collections::BTreeSet<String> =
3619 std::iter::once(r"odd\name".to_owned()).collect();
3620 let note = render_note_scoped(
3621 &node_with("file:x", None, None),
3622 None,
3623 None,
3624 &VaultScope {
3625 project: Some(r"odd\name"),
3626 members: &ms,
3627 },
3628 );
3629 assert_eq!(
3630 frontmatter_field(¬e.content, "project"),
3631 Ok(Some(r"odd\name".to_owned()))
3632 );
3633 }
3634
3635 /// The **bare** fields are the other half of the same class, and were missed
3636 /// by the review that found the quoted ones: `status` is written unquoted, and
3637 /// `roteiro load` installs a caller-supplied artifact whose nodes carry
3638 /// whatever they carry.
3639 #[test]
3640 fn a_bare_field_is_quoted_only_when_being_bare_would_change_it() {
3641 let with_status = |status: &str| {
3642 let mut ex = node_with("adr:0001", None, None);
3643 ex.meta = serde_json::json!({ "status": status });
3644 render_note(&ex, None, None)
3645 };
3646
3647 // Would be a parse error bare; would silently truncate bare.
3648 for status in [
3649 "Accepted: superseded by 0012",
3650 "Accepted # pending",
3651 "{draft}",
3652 "",
3653 ] {
3654 let note = with_status(status);
3655 assert_eq!(
3656 frontmatter_field(¬e.content, "status"),
3657 Ok(Some(status.to_owned())),
3658 "status {status:?} must round-trip"
3659 );
3660 }
3661
3662 // …and a safe one stays bare, which is what keeps an existing vault's
3663 // bytes unchanged.
3664 let note = with_status("Accepted");
3665 assert!(
3666 note.content.contains("\nstatus: Accepted\n"),
3667 "a plain-safe status must not gain quotes: {}",
3668 note.content
3669 );
3670 }
3671
3672 /// `no` is Norwegian, and a bare `no` reads as `false` to a YAML **1.1**
3673 /// parser.
3674 ///
3675 /// The only assertion here that pins emitted bytes, and deliberately so:
3676 /// `yaml-rust2` implements YAML 1.2, whose core schema resolves a bare `no`
3677 /// to the *string* `no`, so a round-trip through this test's own oracle
3678 /// cannot see the problem — it passes either way. The exposure is to the
3679 /// parser on the other side, and Obsidian's is not this one. Quoting costs
3680 /// two characters on a value that never occurs here; guessing which YAML
3681 /// version every downstream reader implements does not seem like the better
3682 /// bet.
3683 #[test]
3684 fn a_language_that_spells_a_yaml_boolean_is_quoted() {
3685 let note = render_note(&node_with("file:x", None, Some("no")), None, None);
3686 assert!(
3687 note.content.contains("\nlang: \"no\"\n"),
3688 "a bare `no` is `false` to a 1.1 parser and must be quoted: {}",
3689 note.content
3690 );
3691 assert_eq!(
3692 frontmatter_field(¬e.content, "lang"),
3693 Ok(Some("no".to_owned())),
3694 "and it must still read back as the string: {}",
3695 note.content
3696 );
3697 // And an ordinary language is untouched.
3698 let rust = render_note(&node_with("file:x", None, Some("rust")), None, None);
3699 assert!(rust.content.contains("\nlang: rust\n"), "{}", rust.content);
3700 }
3701
3702 /// Control characters and the separators some parsers fold as line breaks.
3703 #[test]
3704 fn control_characters_cannot_break_out_of_the_block() {
3705 for path in [
3706 "a\nb",
3707 "a\tb",
3708 "a\u{0}b",
3709 "a\u{2028}b",
3710 "a\u{7f}b",
3711 "a\u{85}b",
3712 ] {
3713 let note = render_note(&node_with("file:x", Some(path), None), None, None);
3714 assert_eq!(
3715 frontmatter_field(¬e.content, "path"),
3716 Ok(Some(path.to_owned())),
3717 "path {path:?} must round-trip"
3718 );
3719 // A raw newline would end the scalar and inject a sibling key.
3720 assert_eq!(
3721 note.content.matches("\npath: ").count(),
3722 1,
3723 "the value must stay on one line: {}",
3724 note.content
3725 );
3726 }
3727 }
3728
3729 /// The escaping is *only* an escaping: for a value with nothing to escape it
3730 /// must emit the same bytes it always did, or #442's promise that a
3731 /// single-project vault is byte-identical does not hold.
3732 #[test]
3733 fn an_ordinary_value_is_emitted_exactly_as_before() {
3734 let note = render_note(
3735 &node_with("sym:rust:src/a.rs#Store", Some("src/a.rs"), Some("rust")),
3736 None,
3737 None,
3738 );
3739 assert!(
3740 note.content
3741 .contains("\nkey: \"sym:rust:src/a.rs#Store\"\n")
3742 );
3743 assert!(note.content.contains("\nkind: fn\n"));
3744 assert!(note.content.contains("\npath: \"src/a.rs\"\n"));
3745 assert!(note.content.contains("\nlang: rust\n"));
3746 }
3747
3748 /// The plain-style decision is checked against a real parser rather than
3749 /// against itself: whatever `is_plain_safe` accepts must actually round-trip
3750 /// bare, and whatever it rejects must round-trip quoted.
3751 #[test]
3752 fn the_plain_style_decision_agrees_with_a_real_yaml_parser() {
3753 for value in [
3754 "fn",
3755 "config_key",
3756 "rust",
3757 "Accepted",
3758 "a.b",
3759 "a/b",
3760 "a-b_c",
3761 "no",
3762 "yes",
3763 "true",
3764 "null",
3765 "y",
3766 "N",
3767 "",
3768 " lead",
3769 "trail ",
3770 "a: b",
3771 "a #c",
3772 "{x}",
3773 "[x]",
3774 "*x",
3775 "&x",
3776 "!x",
3777 "#x",
3778 ">x",
3779 "|x",
3780 "%x",
3781 "@x",
3782 "`x",
3783 "\"x",
3784 "'x",
3785 ",x",
3786 "123",
3787 "1.5",
3788 "-x",
3789 ".x",
3790 "a\\b",
3791 ] {
3792 let emitted = super::yaml_scalar(value);
3793 let doc = format!("v: {emitted}");
3794 let parsed = yaml_rust2::YamlLoader::load_from_str(&doc)
3795 .unwrap_or_else(|e| panic!("{value:?} emitted {emitted:?}: {e}"));
3796 assert_eq!(
3797 parsed[0]["v"].as_str(),
3798 Some(value),
3799 "{value:?} emitted as {emitted:?} did not round-trip"
3800 );
3801 }
3802 }
3803}