Improve local evidence search and AI capabilities

This commit is contained in:
lix-2026
2026-06-05 23:00:53 +08:00
parent a1dcfc9f76
commit 4dbd9a978b
52 changed files with 6415 additions and 527 deletions
+276 -22
View File
@@ -7,10 +7,10 @@ use axum::extract::{Extension, State};
use axum::http::{HeaderMap, HeaderValue, StatusCode};
use axum::Json;
use core_protocol::{
EvidenceBBox, EvidenceLocator, EvidenceOpenAction, EvidenceOpenRequest, EvidenceReadRequest,
EvidenceResourceKind, EvidenceSearchMode, EvidenceSearchRequest, EvidenceSearchResponse,
EvidenceSearchResult, ResourceSourceMap, SourceMapBlock, SourceMapTextItem,
EVIDENCE_LOCATOR_SCHEMA,
EvidenceBBox, EvidenceLocator, EvidenceOpenAction, EvidenceOpenRequest, EvidenceRange,
EvidenceReadRequest, EvidenceResourceKind, EvidenceSearchMode, EvidenceSearchRequest,
EvidenceSearchResponse, EvidenceSearchResult, ResourceSourceMap, SourceMapBlock,
SourceMapTextItem, EVIDENCE_LOCATOR_SCHEMA,
};
use serde_json::Map;
use serde_json::{json, Value};
@@ -47,18 +47,31 @@ pub(crate) async fn search_payload(
)
.map_err(|error| error.with_context(context))?;
let page_id = body.scope.target_document_id.as_deref();
let evidence_owner_filter = if body.scope.include_resources || body.scope.include_ocr {
None
} else {
page_id
};
let query = body.query.trim().to_string();
if matches!(body.mode, EvidenceSearchMode::Graph) {
if let Some(results) = local_search_index::query_evidence_graph_results(
&root_path, &query, page_id, body.top_k,
if let Some(mut results) = local_search_index::query_evidence_graph_results(
&root_path,
&query,
evidence_owner_filter,
body.top_k,
)? {
enrich_citation_links(&mut results);
return Ok(json!(EvidenceSearchResponse { ok: true, results }));
}
}
if let Some(mut results) =
local_search_index::query_evidence_sqlite_results(&root_path, &query, page_id, body.top_k)?
{
if let Some(mut results) = local_search_index::query_evidence_sqlite_results(
&root_path,
&query,
evidence_owner_filter,
body.top_k,
)? {
enrich_sqlite_evidence_results(&mut results, &root_path, &query);
enrich_citation_links(&mut results);
if !results.is_empty() || !query.is_empty() {
let response = EvidenceSearchResponse { ok: true, results };
return Ok(json!(response));
@@ -77,9 +90,13 @@ pub(crate) async fn search_payload(
)?;
let response = EvidenceSearchResponse {
ok: true,
results: evidence_results_from_local_search(
&search, &root_path, &root_uri, body.mode, &query,
),
results: {
let mut results = evidence_results_from_local_search(
&search, &root_path, &root_uri, body.mode, &query,
);
enrich_citation_links(&mut results);
results
},
};
Ok(json!(response))
}
@@ -127,6 +144,149 @@ fn enrich_sqlite_evidence_results(
}
}
fn enrich_citation_links(results: &mut [EvidenceSearchResult]) {
for result in results {
let citation_url = citation_url_for_locator(&result.source);
result.source.open_action.url = citation_url.clone();
result.citation_label = Some(citation_label_for_locator(&result.source));
result.citation_markdown = Some(citation_markdown_for_locator(&result.source));
result.citation_url = Some(citation_url);
}
}
fn citation_markdown_for_locator(locator: &EvidenceLocator) -> String {
let label = citation_label_for_locator(locator);
let url = citation_url_for_locator(locator);
format!(
"[{}]({})",
markdown_link_label_escape(&label),
url.replace(')', "%29")
)
}
fn citation_label_for_locator(locator: &EvidenceLocator) -> String {
let source_path = locator
.resource_path
.as_deref()
.filter(|value| !value.trim().is_empty())
.unwrap_or(locator.owner_document_path.as_str());
let name = source_path
.rsplit('/')
.find(|part| !part.trim().is_empty())
.unwrap_or(source_path)
.trim();
let mut parts = vec![if name.is_empty() {
"证据".to_string()
} else {
name.to_string()
}];
if let Some(page) = locator.page {
parts.push(format!("p.{page}"));
}
if let Some(section) = locator
.section_path
.last()
.filter(|value| !value.trim().is_empty())
{
parts.push(section.to_string());
}
parts.join(" · ")
}
fn citation_url_for_locator(locator: &EvidenceLocator) -> String {
let owner_document_id = locator.owner_document_id.trim();
let mut url = if owner_document_id.is_empty() {
let existing = locator.open_action.url.trim();
if existing.is_empty() {
"/".to_string()
} else {
existing.to_string()
}
} else {
format!("/documents/{owner_document_id}")
};
append_query_param(&mut url, "sourceKind", "local_folder");
append_query_param(&mut url, "rootUri", locator.root_uri.trim());
if let Some(resource_path) = locator
.resource_path
.as_deref()
.filter(|value| !value.trim().is_empty())
{
append_query_param(
&mut url,
"resourceTab",
&format!(
"resource:file:{}:{}",
locator.root_uri.trim(),
resource_path
),
);
append_query_param(&mut url, "resourcePath", resource_path);
}
if let Some(page) = locator.page {
append_query_param(&mut url, "page", &page.to_string());
}
if let Some(bbox) = &locator.bbox {
append_query_param(
&mut url,
"bbox",
&format!("{},{},{},{}", bbox.x0, bbox.y0, bbox.x1, bbox.y1),
);
}
if let Some(block_id) = locator.block_id.as_deref() {
append_query_param(&mut url, "blockId", block_id);
}
if let Some(source_map_path) = locator.source_map_path.as_deref() {
append_query_param(&mut url, "sourceMapPath", source_map_path);
}
if let Some(line_range) = &locator.line_range {
append_query_param(
&mut url,
"lineRange",
&format!("{}-{}", line_range.start, line_range.end),
);
}
if let Some(char_range) = &locator.char_range {
append_query_param(
&mut url,
"charRange",
&format!("{}-{}", char_range.start, char_range.end),
);
}
url
}
fn append_query_param(url: &mut String, key: &str, value: &str) {
let value = value.trim();
if value.is_empty() {
return;
}
let separator = if url.contains('?') { '&' } else { '?' };
url.push(separator);
url.push_str(&encode_query_component(key));
url.push('=');
url.push_str(&encode_query_component(value));
}
fn encode_query_component(value: &str) -> String {
let mut encoded = String::with_capacity(value.len());
for byte in value.as_bytes() {
if byte.is_ascii_alphanumeric() || matches!(byte, b'-' | b'_' | b'.' | b'~') {
encoded.push(*byte as char);
} else {
encoded.push_str(&format!("%{byte:02X}"));
}
}
encoded
}
fn markdown_link_label_escape(value: &str) -> String {
value
.replace('\\', "\\\\")
.replace('[', "\\[")
.replace(']', "\\]")
}
pub async fn read(
State(state): State<AppState>,
Extension(context): Extension<RequestContext>,
@@ -154,12 +314,13 @@ pub(crate) async fn read_payload(
state, context, &root_uri,
)
.map_err(|error| error.with_context(context))?;
if let Some(results) = read_source_map_context(
if let Some(mut results) = read_source_map_context(
&root_path,
&body.locator,
body.context.before_blocks,
body.context.after_blocks,
) {
enrich_citation_links(&mut results);
let quote = results
.iter()
.find(|item| locator_matches(&item.source, &body.locator))
@@ -176,17 +337,21 @@ pub(crate) async fn read_payload(
"locator": body.locator,
"quote": quote,
"sectionPath": section_path,
"citationUrl": citation_url_for_locator(&body.locator),
"citationLabel": citation_label_for_locator(&body.locator),
"citationMarkdown": citation_markdown_for_locator(&body.locator),
"contextBlocks": results,
});
return Ok(response);
}
if let Some(results) = local_search_index::read_evidence_sqlite_context(
if let Some(mut results) = local_search_index::read_evidence_sqlite_context(
&root_path,
&body.locator,
body.context.before_blocks,
body.context.after_blocks,
)? {
if !results.is_empty() {
enrich_citation_links(&mut results);
let quote = results
.iter()
.find(|item| locator_matches(&item.source, &body.locator))
@@ -197,6 +362,9 @@ pub(crate) async fn read_payload(
"ok": true,
"locator": body.locator,
"quote": quote,
"citationUrl": citation_url_for_locator(&body.locator),
"citationLabel": citation_label_for_locator(&body.locator),
"citationMarkdown": citation_markdown_for_locator(&body.locator),
"contextBlocks": results,
});
return Ok(response);
@@ -222,13 +390,14 @@ pub(crate) async fn read_payload(
false,
true,
)?;
let results = evidence_results_from_local_search(
let mut results = evidence_results_from_local_search(
&search,
&root_path,
&root_uri,
EvidenceSearchMode::Tree,
&query,
);
enrich_citation_links(&mut results);
let quote = results
.first()
.map(|item| item.quote.clone())
@@ -237,6 +406,9 @@ pub(crate) async fn read_payload(
"ok": true,
"locator": body.locator,
"quote": quote,
"citationUrl": citation_url_for_locator(&body.locator),
"citationLabel": citation_label_for_locator(&body.locator),
"citationMarkdown": citation_markdown_for_locator(&body.locator),
"contextBlocks": results,
});
Ok(response)
@@ -365,6 +537,9 @@ fn source_map_block_result(
quote: block.text.clone(),
score,
source,
citation_url: None,
citation_label: None,
citation_markdown: None,
}
}
@@ -394,7 +569,9 @@ pub(crate) async fn open_payload(
context: &RequestContext,
body: EvidenceOpenRequest,
) -> Result<Value, WebError> {
let locator = body.locator;
let mut locator = body.locator;
let citation_url = citation_url_for_locator(&locator);
locator.open_action.url = citation_url.clone();
let root_uri = locator.root_uri.trim().to_string();
if root_uri.is_empty() {
return Err(
@@ -410,6 +587,9 @@ pub(crate) async fn open_payload(
"ok": true,
"locator": locator.clone(),
"openAction": locator.open_action,
"citationUrl": citation_url,
"citationLabel": citation_label_for_locator(&locator),
"citationMarkdown": citation_markdown_for_locator(&locator),
});
Ok(response)
}
@@ -499,19 +679,29 @@ pub(crate) fn evidence_results_from_local_search(
})
.or_else(|| source_map_hit.as_ref().and_then(|hit| hit.bbox.clone()));
let block_id = result
.get("ocrEvidence")
.and_then(|_| source_map_hit.as_ref().and_then(|hit| hit.block_id.clone()))
.or_else(|| result.get("id").and_then(Value::as_str).map(str::to_string))
.get("blockId")
.and_then(Value::as_str)
.map(str::to_string)
.or_else(|| {
result
.get("ocrEvidence")
.and_then(|_| source_map_hit.as_ref().and_then(|hit| hit.block_id.clone()))
})
.or_else(|| source_map_hit.as_ref().and_then(|hit| hit.block_id.clone()));
let line_range = result.get("lineRange").and_then(evidence_range_from_value);
let char_range = source_map_hit
.as_ref()
.and_then(|hit| hit.char_range.clone());
.and_then(|hit| hit.char_range.clone())
.or_else(|| result.get("charRange").and_then(evidence_range_from_value));
let open_action_params = evidence_open_action_params(
&result,
query,
&mode,
page,
bbox.clone(),
block_id.clone(),
line_range.clone(),
char_range.clone(),
source_map_path.clone(),
);
let locator = EvidenceLocator {
@@ -534,7 +724,7 @@ pub(crate) fn evidence_results_from_local_search(
.collect::<Vec<_>>()
})
.unwrap_or_default(),
line_range: None,
line_range,
char_range,
block_id,
source_map_path: source_map_path.clone(),
@@ -557,6 +747,9 @@ pub(crate) fn evidence_results_from_local_search(
.to_string(),
score: result.get("score").and_then(Value::as_f64).unwrap_or(0.0),
source: locator,
citation_url: None,
citation_label: None,
citation_markdown: None,
}
})
.collect()
@@ -568,6 +761,9 @@ fn evidence_open_action_params(
mode: &EvidenceSearchMode,
page: Option<u32>,
bbox: Option<EvidenceBBox>,
block_id: Option<String>,
line_range: Option<EvidenceRange>,
char_range: Option<EvidenceRange>,
source_map_path: Option<String>,
) -> Map<String, Value> {
let mut params = Map::new();
@@ -586,12 +782,35 @@ fn evidence_open_action_params(
if let Some(bbox) = bbox {
params.insert("bbox".into(), json!(bbox));
}
if let Some(block_id) = block_id {
params.insert("blockId".into(), json!(block_id));
}
if let Some(line_range) = line_range {
params.insert("lineRange".into(), json!(line_range));
}
if let Some(char_range) = char_range {
params.insert("charRange".into(), json!(char_range));
}
if let Some(source_map_path) = source_map_path {
params.insert("sourceMapPath".into(), json!(source_map_path));
}
params
}
fn evidence_range_from_value(value: &Value) -> Option<EvidenceRange> {
if let Some(map) = value.as_object() {
let start = map.get("start")?.as_u64()?;
let end = map.get("end")?.as_u64()?;
return Some(EvidenceRange { start, end });
}
let text = value.as_str()?.trim();
let (start, end) = text.split_once('-')?;
Some(EvidenceRange {
start: start.trim().parse().ok()?,
end: end.trim().parse().ok()?,
})
}
fn infer_resource_kind(resource_type: &str, resource_path: Option<&str>) -> EvidenceResourceKind {
let extension = resource_path
.and_then(|path| Path::new(path).extension())
@@ -788,6 +1007,15 @@ mod tests {
)
.expect("readme");
let root_uri = format!("file://{}", root.display());
local_search_index::write_local_index_settings(
&root,
&[String::from(".")],
None,
None,
None,
None,
)
.expect("settings");
local_search_index::refresh_local_search_index(&root, &root_uri, "local-ws-evidence-route")
.expect("refresh");
fs::write(
@@ -838,6 +1066,7 @@ mod tests {
"scope": {
"workspaceId": "local-ws-evidence-route",
"rootUri": root_uri,
"targetDocumentId": "local-md:missing.md",
"includeResources": true,
"includeOcr": true
},
@@ -925,13 +1154,14 @@ mod tests {
}]
});
let results = evidence_results_from_local_search(
let mut results = evidence_results_from_local_search(
&payload,
&root,
"file:///workspace",
EvidenceSearchMode::Hybrid,
"OCR-only-token",
);
enrich_citation_links(&mut results);
fs::remove_dir_all(&root).ok();
@@ -954,6 +1184,16 @@ mod tests {
locator.open_action.params["sourceMapPath"].as_str(),
Some("docs/Page.ocr/photo.png.source-map.json")
);
let citation_url = results[0].citation_url.as_deref().unwrap_or_default();
assert!(citation_url.starts_with("/documents/local-md:docs~2FPage.md?"));
assert!(citation_url.contains("resourceTab=resource%3Afile%3Afile%3A%2F%2F%2Fworkspace%3Adocs%2FPage.assets%2Fphoto.png"));
assert!(citation_url.contains("page=1"));
assert!(citation_url.contains("blockId=p1_b1"));
assert!(results[0]
.citation_markdown
.as_deref()
.unwrap_or_default()
.contains("photo.png"));
}
#[test]
@@ -1011,9 +1251,13 @@ mod tests {
quote: "NeedleToken 原文定位".into(),
score: 1.0,
source: locator,
citation_url: None,
citation_label: None,
citation_markdown: None,
}];
enrich_sqlite_evidence_results(&mut results, &root, "NeedleToken");
enrich_citation_links(&mut results);
fs::remove_dir_all(&root).ok();
let locator = &results[0].source;
@@ -1029,6 +1273,16 @@ mod tests {
locator.open_action.params["blockId"].as_str(),
Some("p2_b7")
);
assert!(results[0]
.citation_url
.as_deref()
.unwrap_or_default()
.contains("sourceMapPath=docs%2FPage.ocr%2Fspec.pdf.source-map.json"));
assert!(results[0]
.citation_markdown
.as_deref()
.unwrap_or_default()
.contains("spec.pdf"));
}
#[test]
@@ -1318,6 +1318,123 @@ pub async fn toggle_skill(
))
}
pub async fn list_capabilities(
State(state): State<AppState>,
Extension(context): Extension<RequestContext>,
Query(query): Query<HashMap<String, String>>,
) -> Result<(StatusCode, HeaderMap, Json<Value>), WebError> {
ensure_authenticated(&context)?;
let runtime = query.get("runtime").map(String::as_str).unwrap_or("mnote");
if runtime != "mnote" {
return Ok((
StatusCode::OK,
stamp_client_headers(),
Json(json!({
"ok": true,
"runtime": runtime,
"categories": [],
"archived": []
})),
));
}
let fallback_profile = active_profile_name().unwrap_or_else(|| "default".into());
let profile = query
.get("profile")
.map(String::as_str)
.unwrap_or(fallback_profile.as_str());
let payload = mnote_capabilities_payload(
&state,
&context,
query.get("agentId").map(String::as_str),
profile,
)?;
Ok((StatusCode::OK, stamp_client_headers(), Json(payload)))
}
pub async fn toggle_capability(
State(state): State<AppState>,
Extension(context): Extension<RequestContext>,
Json(payload): Json<Value>,
) -> Result<(StatusCode, HeaderMap, Json<Value>), WebError> {
ensure_authenticated(&context)?;
let capability_id = payload
.get("id")
.or_else(|| payload.get("name"))
.and_then(Value::as_str)
.map(str::trim)
.filter(|value| !value.is_empty())
.ok_or_else(|| {
WebError::bad_request_code("hermes_client_bad_request", "缺少 capability id")
.with_context(&context)
.with_header(HEADER_MNOTE_WEB_OWNER, "mnote-web")
.with_header(HEADER_HERMES_CLIENT_OWNER, "mnote-web-hermes-client")
})?;
let enabled = payload
.get("enabled")
.and_then(Value::as_bool)
.ok_or_else(|| {
WebError::bad_request_code("hermes_client_bad_request", "缺少 enabled")
.with_context(&context)
.with_header(HEADER_MNOTE_WEB_OWNER, "mnote-web")
.with_header(HEADER_HERMES_CLIENT_OWNER, "mnote-web-hermes-client")
})?;
let runtime = payload
.get("runtime")
.and_then(Value::as_str)
.map(str::trim)
.unwrap_or("mnote");
if runtime != "mnote" {
return Err(WebError::bad_request_code(
"hermes_client_capability_runtime_unsupported",
"当前只支持 MNote 内置能力开关",
)
.with_context(&context));
}
let fallback_profile = active_profile_name().unwrap_or_else(|| "default".into());
let profile = payload
.get("profile")
.and_then(Value::as_str)
.map(str::trim)
.filter(|value| !value.is_empty())
.unwrap_or(fallback_profile.as_str());
let skill = crate::hermes_tools::skill::find_skill(capability_id, None).ok_or_else(|| {
WebError::new(
StatusCode::NOT_FOUND,
"mnote_capability_not_found",
"未知 MNote AI 能力",
)
.with_context(&context)
})?;
let actor_id = page_ai_actor_id(&state, &context)?;
ensure_page_ai_actor_user(&state, &actor_id, page_ai_actor_is_admin(&context))?;
set_mnote_builtin_capability_enabled(&state, &actor_id, capability_id, enabled, &context)?;
for tool_name in skill.tool_names {
if mnote_capability_tool_toggleable(tool_name) {
set_mnote_tool_enabled(profile, tool_name, enabled).map_err(|error| {
WebError::bad_gateway_code(
"hermes_client_capability_tool_toggle_failed",
format!("更新 MNote 能力工具设置失败: {error}"),
)
.with_context(&context)
.with_header(HEADER_MNOTE_WEB_OWNER, "mnote-web")
.with_header(HEADER_HERMES_CLIENT_OWNER, "mnote-web-hermes-client")
})?;
}
}
Ok((
StatusCode::OK,
stamp_client_headers(),
Json(json!({
"ok": true,
"runtime": "mnote",
"id": capability_id,
"enabled": enabled,
"profile": profile,
"configScope": "user_sqlite+profile_tool_policy"
})),
))
}
pub async fn toggle_tool(
Extension(context): Extension<RequestContext>,
Json(payload): Json<Value>,
@@ -4068,6 +4185,217 @@ fn mnote_tool_entry(tool: &Value, disabled: &[String]) -> Option<Value> {
}))
}
fn mnote_capabilities_payload(
state: &AppState,
context: &RequestContext,
agent_id: Option<&str>,
profile: &str,
) -> Result<Value, WebError> {
let mut skills_payload = mnote_builtin_skills_payload(agent_id);
stamp_mnote_builtin_skill_payload_policy(state, context, &mut skills_payload)?;
let tools_by_name = mnote_tools_payload(profile)
.into_iter()
.filter_map(|tool| {
let name = tool.get("name").and_then(Value::as_str)?.to_string();
Some((name, tool))
})
.collect::<BTreeMap<_, _>>();
let mut capability_categories: BTreeMap<String, Vec<Value>> = BTreeMap::new();
for category in skills_payload
.get("categories")
.and_then(Value::as_array)
.into_iter()
.flatten()
{
let mut capabilities = Vec::new();
for skill in category
.get("skills")
.and_then(Value::as_array)
.into_iter()
.flatten()
{
let Some(id) = skill.get("id").and_then(Value::as_str) else {
continue;
};
if id == "mnote-chat-only" {
continue;
}
let tool_names = skill
.get("toolNames")
.and_then(Value::as_array)
.map(|items| {
items
.iter()
.filter_map(Value::as_str)
.filter(|name| !name.trim().is_empty())
.map(ToOwned::to_owned)
.collect::<Vec<_>>()
})
.unwrap_or_default();
let tools = tool_names
.iter()
.filter_map(|name| tools_by_name.get(name).cloned())
.collect::<Vec<_>>();
let disabled_tool_count = tools
.iter()
.filter(|tool| tool.get("enabled").and_then(Value::as_bool) == Some(false))
.count();
let enabled = skill
.get("enabled")
.and_then(Value::as_bool)
.unwrap_or(true);
let status = if !enabled {
"disabled"
} else if disabled_tool_count > 0 {
"partial"
} else {
"available"
};
let capability_category = skill
.get("category")
.and_then(Value::as_str)
.map(str::trim)
.filter(|value| !value.is_empty())
.unwrap_or("mnote");
capabilities.push(json!({
"id": id,
"name": id,
"title": skill.get("title").cloned().unwrap_or_else(|| json!(id)),
"description": skill.get("description").cloned().unwrap_or(Value::Null),
"enabled": enabled,
"toggleable": skill.get("toggleable").cloned().unwrap_or_else(|| json!(true)),
"builtin": true,
"configurable": true,
"configScope": "user_sqlite+profile_tool_policy",
"skillKind": "mnote_capability",
"source": "mnote",
"origin": "builtin",
"category": capability_category,
"categoryTitle": mnote_capability_category_title(capability_category),
"capabilityId": id,
"capabilityKind": "mnote_builtin",
"uiKind": mnote_capability_ui_kind(capability_category),
"skillId": id,
"readOnly": skill.get("readOnly").cloned().unwrap_or(Value::Bool(false)),
"agentIds": skill.get("agentIds").cloned().unwrap_or(Value::Null),
"requiresContextRefs": skill.get("requiresContextRefs").cloned().unwrap_or(Value::Null),
"toolNames": tool_names,
"tools": tools,
"toolCount": tools.len(),
"disabledToolCount": disabled_tool_count,
"status": status,
}));
}
for capability in capabilities {
let category_name = capability
.get("category")
.and_then(Value::as_str)
.unwrap_or("mnote")
.to_string();
capability_categories
.entry(category_name)
.or_default()
.push(capability);
}
}
let categories = ordered_mnote_capability_categories(capability_categories)
.into_iter()
.map(|(name, capabilities)| {
json!({
"name": name,
"title": mnote_capability_category_title(&name),
"description": mnote_capability_category_description(&name),
"capabilities": capabilities.clone(),
"skills": capabilities
})
})
.collect::<Vec<_>>();
Ok(json!({
"ok": true,
"runtime": "mnote",
"profile": profile,
"categories": categories,
"archived": []
}))
}
fn ordered_mnote_capability_categories(
mut categories: BTreeMap<String, Vec<Value>>,
) -> Vec<(String, Vec<Value>)> {
let mut ordered = Vec::new();
for name in ["mnote", "knowledge", "file", "resource", "office", "chat"] {
if let Some(capabilities) = categories.remove(name) {
ordered.push((name.to_string(), capabilities));
}
}
ordered.extend(categories);
ordered
}
fn mnote_capability_category_title(category: &str) -> &'static str {
match category {
"knowledge" => "知识库与索引",
"file" => "本地文件",
"resource" => "资源编辑",
"office" => "Office / ONLYOFFICE",
"chat" => "聊天",
_ => "MNote",
}
}
fn mnote_capability_category_description(category: &str) -> &'static str {
match category {
"knowledge" => "本地索引、证据检索和资料范围管理。",
"file" => "授权目录内的本地 Markdown 文件读写。",
"resource" => "MNote 资源型编辑器能力,例如思维导图。",
"office" => "Office 摘要、建议和 ONLYOFFICE 实时编辑桥。",
"chat" => "不读取文档上下文的普通对话能力。",
_ => "MNote 页面上下文与基础能力。",
}
}
fn mnote_capability_ui_kind(category: &str) -> &'static str {
if category == "chat" {
"chat"
} else {
"ai_capability"
}
}
fn set_mnote_builtin_capability_enabled(
state: &AppState,
actor_id: &str,
capability_id: &str,
enabled: bool,
context: &RequestContext,
) -> Result<(), WebError> {
let key = format!("ai.agent.mnote_builtin.skill.{capability_id}.enabled");
state
.control_plane()
.upsert_user_ui_preference(control_plane::UpsertUserUiPreferenceInput {
id: None,
user_id: actor_id.to_string(),
workspace_id: None,
source_kind: None,
scope_kind: "page_ai_capability".to_string(),
scope_id: "mnote_builtin".to_string(),
key,
value_json: Value::Bool(enabled).to_string(),
})
.map(|_| ())
.map_err(|error| {
WebError::internal(format!("SQLite MNote AI 能力偏好写入失败: {error}"))
.with_context(context)
})
}
fn mnote_capability_tool_toggleable(tool_name: &str) -> bool {
!matches!(
tool_name,
"mnote.skill.read" | "mnote.context.snapshot" | "mnote.context.resolve_target"
)
}
fn extract_skill_description(markdown: &str) -> String {
markdown
.lines()
@@ -4451,6 +4779,7 @@ fn mnote_builtin_skills_payload(agent_id: Option<&str>) -> Value {
"skillKind": "mnote_builtin",
"source": "mnote",
"origin": "builtin",
"category": skill.get("category").cloned().unwrap_or_else(|| json!("mnote")),
"agentIds": skill.get("agentIds").cloned().unwrap_or(Value::Null),
"toolNames": skill.get("toolNames").cloned().unwrap_or(Value::Null),
"requiresContextRefs": skill.get("requiresContextRefs").cloned().unwrap_or(Value::Null)
@@ -10144,6 +10473,223 @@ mod tests {
let _ = fs::remove_dir_all(&hermes_home);
}
#[tokio::test]
async fn page_ai_capabilities_expose_local_index_and_toggle_tools() {
let _env_guard = env_lock().lock().expect("env lock");
let hermes_home = std::env::temp_dir().join(format!(
"mnote-web-ai-capability-policy-{}",
std::process::id()
));
let _ = fs::remove_dir_all(&hermes_home);
fs::create_dir_all(&hermes_home).expect("hermes home");
std::env::set_var("HERMES_HOME", &hermes_home);
let app = build_app(test_state());
let response = app
.clone()
.oneshot(
Request::builder()
.method("GET")
.uri("/api/hermes/client/capabilities?runtime=mnote&agentId=reasonix&profile=chemist")
.header("x-mnote-actor-id", "user_1")
.header("x-mnote-actor-type", "user")
.body(Body::empty())
.expect("request"),
)
.await
.expect("capabilities");
assert_eq!(response.status(), StatusCode::OK);
let body = to_bytes(response.into_body(), usize::MAX)
.await
.expect("body");
let payload: Value = serde_json::from_slice(&body).expect("capabilities json");
let all_capabilities = payload["categories"]
.as_array()
.expect("categories")
.iter()
.flat_map(|category| category["capabilities"].as_array().into_iter().flatten())
.collect::<Vec<_>>();
assert!(
all_capabilities
.iter()
.all(|capability| capability["id"] != "mnote-chat-only"),
"纯聊天是 agent 模式,不应作为 MNote 公共能力展示"
);
let local_index = payload["categories"]
.as_array()
.expect("categories")
.iter()
.flat_map(|category| category["capabilities"].as_array().into_iter().flatten())
.find(|capability| capability["id"] == "mnote-local-index")
.expect("local index capability");
assert_eq!(local_index["enabled"], true);
assert_eq!(local_index["uiKind"], "ai_capability");
assert!(local_index["tools"]
.as_array()
.expect("local index tools")
.iter()
.any(|tool| tool["name"] == "mnote.index.status"));
assert!(local_index["tools"]
.as_array()
.expect("local index tools")
.iter()
.any(|tool| tool["name"] == "mnote.index.update_settings"));
let toggle_response = app
.clone()
.oneshot(
Request::builder()
.method("PUT")
.uri("/api/hermes/client/capabilities/toggle")
.header("content-type", "application/json")
.header("x-mnote-actor-id", "user_1")
.header("x-mnote-actor-type", "user")
.body(Body::from(
json!({
"runtime": "mnote",
"profile": "chemist",
"id": "mnote-local-index",
"enabled": false
})
.to_string(),
))
.expect("request"),
)
.await
.expect("toggle capability");
assert_eq!(toggle_response.status(), StatusCode::OK);
let response = app
.clone()
.oneshot(
Request::builder()
.method("GET")
.uri("/api/hermes/client/capabilities?runtime=mnote&agentId=reasonix&profile=chemist")
.header("x-mnote-actor-id", "user_1")
.header("x-mnote-actor-type", "user")
.body(Body::empty())
.expect("request"),
)
.await
.expect("capabilities after toggle");
assert_eq!(response.status(), StatusCode::OK);
let body = to_bytes(response.into_body(), usize::MAX)
.await
.expect("body");
let payload: Value = serde_json::from_slice(&body).expect("capabilities json");
let local_index = payload["categories"]
.as_array()
.expect("categories")
.iter()
.flat_map(|category| category["capabilities"].as_array().into_iter().flatten())
.find(|capability| capability["id"] == "mnote-local-index")
.expect("local index capability");
assert_eq!(local_index["enabled"], false);
assert_eq!(local_index["status"], "disabled");
assert!(local_index["tools"]
.as_array()
.expect("local index tools")
.iter()
.any(|tool| tool["name"] == "mnote.index.status" && tool["enabled"] == false));
let tools_response = app
.oneshot(
Request::builder()
.method("GET")
.uri("/api/hermes/client/tools?scope=mnote&profile=chemist")
.header("x-mnote-actor-id", "user_1")
.body(Body::empty())
.expect("request"),
)
.await
.expect("tools after toggle");
assert_eq!(tools_response.status(), StatusCode::OK);
let body = to_bytes(tools_response.into_body(), usize::MAX)
.await
.expect("body");
let payload: Value = serde_json::from_slice(&body).expect("tools json");
let tools = payload["tools"]
.as_array()
.expect("tools")
.iter()
.map(|tool| {
(
tool["name"].as_str().unwrap_or_default().to_string(),
tool.clone(),
)
})
.collect::<HashMap<_, _>>();
assert_eq!(tools["mnote.index.status"]["enabled"], false);
assert_eq!(tools["mnote.index.update_settings"]["enabled"], false);
std::env::remove_var("HERMES_HOME");
let _ = fs::remove_dir_all(&hermes_home);
}
#[tokio::test]
async fn page_ai_capabilities_group_onlyoffice_live_bridge() {
let _env_guard = env_lock().lock().expect("env lock");
let hermes_home = std::env::temp_dir().join(format!(
"mnote-web-ai-onlyoffice-capability-{}",
std::process::id()
));
let _ = fs::remove_dir_all(&hermes_home);
fs::create_dir_all(&hermes_home).expect("hermes home");
std::env::set_var("HERMES_HOME", &hermes_home);
let response = build_app(test_state())
.oneshot(
Request::builder()
.method("GET")
.uri("/api/hermes/client/capabilities?runtime=mnote&agentId=reasonix&profile=chemist")
.header("x-mnote-actor-id", "user_1")
.header("x-mnote-actor-type", "user")
.body(Body::empty())
.expect("request"),
)
.await
.expect("capabilities");
assert_eq!(response.status(), StatusCode::OK);
let body = to_bytes(response.into_body(), usize::MAX)
.await
.expect("body");
let payload: Value = serde_json::from_slice(&body).expect("capabilities json");
let office_category = payload["categories"]
.as_array()
.expect("categories")
.iter()
.find(|category| category["name"] == "office")
.expect("office category");
assert_eq!(office_category["title"], "Office / ONLYOFFICE");
let onlyoffice = office_category["capabilities"]
.as_array()
.expect("office capabilities")
.iter()
.find(|capability| capability["id"] == "mnote-onlyoffice-live")
.expect("onlyoffice capability");
assert_eq!(onlyoffice["title"], "ONLYOFFICE 实时编辑");
assert_eq!(onlyoffice["categoryTitle"], "Office / ONLYOFFICE");
assert_eq!(onlyoffice["enabled"], true);
assert!(onlyoffice["tools"]
.as_array()
.expect("onlyoffice tools")
.iter()
.any(|tool| tool["name"] == "mnote.onlyoffice.session.current"));
assert!(onlyoffice["tools"]
.as_array()
.expect("onlyoffice tools")
.iter()
.any(|tool| tool["name"] == "mnote.onlyoffice.sheet.batch_set_values"));
assert!(onlyoffice["tools"]
.as_array()
.expect("onlyoffice tools")
.iter()
.any(|tool| tool["name"] == "mnote.onlyoffice.presentation.add_shape"));
std::env::remove_var("HERMES_HOME");
let _ = fs::remove_dir_all(&hermes_home);
}
#[test]
fn reasonix_memory_policy_defaults_off_and_reads_user_preference() {
let state = test_state();
@@ -3,8 +3,8 @@ use crate::app::AppState;
use crate::context::RequestContext;
use crate::error::WebError;
use crate::hermes_tools::{
artifact, block, context_tools, doc, evidence, manifest, onlyoffice_live, page, resource,
skill, ToolCallInput,
artifact, block, context_tools, doc, evidence, index, manifest, onlyoffice_live, page,
resource, skill, ToolCallInput,
};
use axum::extract::{Extension, Query, State};
use axum::http::{HeaderMap, HeaderName, HeaderValue, StatusCode};
@@ -365,6 +365,11 @@ pub(crate) async fn execute_mnote_tool_call(
"mnote.evidence.search" => evidence::evidence_search(&state, &context, &input).await,
"mnote.evidence.read" => evidence::evidence_read(&state, &context, &input).await,
"mnote.evidence.open" => evidence::evidence_open(&state, &context, &input).await,
"mnote.index.status" => index::index_status(&state, &context, &input).await,
"mnote.index.refresh" => index::index_refresh(&state, &context, &input).await,
"mnote.index.update_settings" => {
index::index_update_settings(&state, &context, &input).await
}
"mnote.doc.plan_update" => doc::plan_update(&state, &context, &input).await,
"mnote.block.fetch" => block::block_fetch(&state, &context, &input).await,
"mnote.block.replace" => block::block_replace(&state, &context, &input).await,
@@ -704,6 +709,8 @@ fn is_read_tool(tool_name: &str) -> bool {
| "mnote.evidence.search"
| "mnote.evidence.read"
| "mnote.evidence.open"
| "mnote.index.status"
| "mnote.index.refresh"
| "mnote.block.fetch"
| "mnote.mindmap.fetch"
| "mnote.office.fetch_summary"
@@ -1578,6 +1585,70 @@ mod tests {
);
}
#[tokio::test]
async fn hermes_tools_manifest_exposes_mnote_capability_packs() {
let response = app()
.oneshot(
Request::builder()
.uri("/api/hermes/tools/mnote/manifest")
.header("x-mnote-actor-id", "user_1")
.body(Body::empty())
.expect("request"),
)
.await
.expect("response");
assert_eq!(response.status(), StatusCode::OK);
let body = to_bytes(response.into_body(), usize::MAX)
.await
.expect("body");
let payload: Value = serde_json::from_slice(&body).expect("json");
let manifest = &payload["manifest"];
let capabilities = manifest["capabilities"].as_array().expect("capabilities");
assert!(capabilities
.iter()
.any(|capability| capability["id"] == "mnote-local-index"));
assert!(capabilities
.iter()
.any(|capability| capability["id"] == "mnote-onlyoffice-live"));
let tools = manifest["tools"].as_array().expect("tools");
let index_status = tools
.iter()
.find(|tool| tool["name"] == "mnote.index.status")
.expect("index status tool");
assert!(index_status["capabilityIds"]
.as_array()
.expect("index capability ids")
.iter()
.any(|id| id == "mnote-local-index"));
let onlyoffice_batch_set = tools
.iter()
.find(|tool| tool["name"] == "mnote.onlyoffice.sheet.batch_set_values")
.expect("onlyoffice batch set tool");
assert_eq!(
onlyoffice_batch_set["capabilityId"],
"mnote-onlyoffice-live"
);
assert!(onlyoffice_batch_set["capabilityIds"]
.as_array()
.expect("onlyoffice capability ids")
.iter()
.any(|id| id == "mnote-onlyoffice-live"));
let context_snapshot = tools
.iter()
.find(|tool| tool["name"] == "mnote.context.snapshot")
.expect("context snapshot tool");
assert!(
context_snapshot["capabilityIds"]
.as_array()
.expect("context capability ids")
.len()
> 1
);
}
#[tokio::test]
async fn hermes_tools_onlyoffice_session_current_reads_registered_bridge_session() {
let app = app();
@@ -5386,6 +5457,21 @@ mod tests {
)
.expect("markdown");
let root_uri = format!("file://{}", root.display());
crate::routes::local_search_index::write_local_index_settings(
&root,
&[String::from(".")],
None,
None,
None,
None,
)
.expect("settings");
crate::routes::local_search_index::refresh_local_search_index(
&root,
&root_uri,
"local-ws-docs-search",
)
.expect("refresh");
let response = app()
.oneshot(
@@ -5437,6 +5523,12 @@ mod tests {
result["source"]["ownerDocumentPath"].as_str(),
Some("README.md")
);
assert!(result["citationMarkdown"]
.as_str()
.is_some_and(|value| value.contains("](/documents/")));
assert!(result["citationUrl"]
.as_str()
.is_some_and(|value| value.contains("resourceTab=")));
let evidence_id = result["evidenceId"].as_str().expect("evidence id");
assert_eq!(
payload["result"]["evidenceIds"][0].as_str(),
@@ -5480,6 +5572,112 @@ mod tests {
let _ = fs::remove_dir_all(&root);
}
#[tokio::test]
async fn hermes_tools_local_index_update_and_status_manage_scope() {
let root = std::env::temp_dir().join(format!("mnote-index-tool-{}", std::process::id()));
let _ = fs::remove_dir_all(&root);
fs::create_dir_all(root.join(".mnote")).expect("metadata");
fs::create_dir_all(root.join("docs")).expect("docs");
fs::write(
root.join(".mnote").join("workspace.json"),
r#"{"workspaceId":"local-ws-index-tool","ownerId":"user_1","createdAt":"2026-06-04T00:00:00Z","capabilities":["local_files","search"]}"#,
)
.expect("manifest");
fs::write(
root.join("docs").join("indexed.md"),
"# Indexed\n\nindex-tool-token\n",
)
.expect("markdown");
let root_uri = format!("file://{}", root.display());
let app = app();
let update_response = app
.clone()
.oneshot(
Request::builder()
.method("POST")
.uri("/api/hermes/tools/mnote/call")
.header("content-type", "application/json")
.header("x-mnote-actor-id", "user_1")
.header("x-mnote-actor-type", "user")
.body(Body::from(
json!({
"toolName": "mnote.index.update_settings",
"workspaceId": "local-ws-index-tool",
"sourceKind": "local_folder",
"rootUri": root_uri,
"sessionId": "sess_index_tool",
"runId": "run_index_tool",
"toolCallId": "call_index_update",
"traceId": "trace_index_tool",
"dryRun": false,
"idempotencyKey": "idem_index_tool_update",
"args": {
"includePaths": ["docs"],
"scheduleMode": "manual",
"scheduleTime": "02:00",
"runOnChange": false
}
})
.to_string(),
))
.expect("update request"),
)
.await
.expect("update response");
assert_eq!(update_response.status(), StatusCode::OK);
let update_body = to_bytes(update_response.into_body(), usize::MAX)
.await
.expect("update body");
let update_payload: Value = serde_json::from_slice(&update_body).expect("update json");
assert_eq!(
update_payload["result"]["settings"]["includePaths"][0],
"docs"
);
assert_eq!(update_payload["result"]["index"]["documentCount"], 1);
assert!(root.join(".mnote/index/search-index.json").exists());
assert!(root.join(".mnote/index/evidence.sqlite").exists());
let status_response = app
.oneshot(
Request::builder()
.method("POST")
.uri("/api/hermes/tools/mnote/call")
.header("content-type", "application/json")
.header("x-mnote-actor-id", "user_1")
.header("x-mnote-actor-type", "user")
.body(Body::from(
json!({
"toolName": "mnote.index.status",
"workspaceId": "local-ws-index-tool",
"sourceKind": "local_folder",
"rootUri": root_uri,
"sessionId": "sess_index_tool",
"runId": "run_index_tool",
"toolCallId": "call_index_status",
"traceId": "trace_index_tool",
"args": {}
})
.to_string(),
))
.expect("status request"),
)
.await
.expect("status response");
assert_eq!(status_response.status(), StatusCode::OK);
let status_body = to_bytes(status_response.into_body(), usize::MAX)
.await
.expect("status body");
let status_payload: Value = serde_json::from_slice(&status_body).expect("status json");
assert_eq!(
status_payload["result"]["result"]["settings"]["includePaths"][0],
"docs"
);
assert_eq!(status_payload["audit"]["effect"], "read");
let _ = fs::remove_dir_all(&root);
}
#[tokio::test]
async fn hermes_tools_legacy_docs_read_returns_document_with_locator() {
let root =
@@ -295,6 +295,7 @@ struct LocalFolderRow {
capabilities: Vec<String>,
workspace_id: String,
root_source_uri: String,
index_status: Option<String>,
}
#[derive(Debug, Clone)]
@@ -303,6 +304,40 @@ struct LocalFolderScanResult {
watch_revision: LocalFolderWatchRevision,
}
#[derive(Debug, Clone, Default)]
struct LocalFileTreeIndexState {
indexed_paths: BTreeSet<String>,
failed_paths: BTreeSet<String>,
}
impl LocalFileTreeIndexState {
fn load(root: &Path) -> Self {
local_search_index::local_evidence_source_statuses(root)
.map(|statuses| Self {
indexed_paths: statuses.indexed_paths,
failed_paths: statuses.failed_paths,
})
.unwrap_or_default()
}
fn status_for_entry(&self, entry: &LocalFolderEntry) -> Option<String> {
if entry.is_dir || entry.is_symlink {
return None;
}
let path = entry.relative_path.trim();
if path.is_empty() {
return None;
}
if self.failed_paths.contains(path) {
return Some("failed".to_string());
}
if self.indexed_paths.contains(path) {
return Some("indexed".to_string());
}
None
}
}
#[derive(Debug, Clone)]
pub(crate) struct LocalUploadFile {
name: String,
@@ -2717,6 +2752,7 @@ fn load_local_folder_file_tree_scope_snapshot(
let workspace_id = local_workspace_id(&canonical_root);
let root_source_uri = file_uri_for_path(&canonical_root);
let metadata = load_local_folder_metadata(&canonical_root)?;
let index_state = LocalFileTreeIndexState::load(&canonical_root);
let parent_relative_path = parent_relative_path
.map(str::trim)
.filter(|value| !value.is_empty() && *value != ".")
@@ -2746,6 +2782,7 @@ fn load_local_folder_file_tree_scope_snapshot(
&root_source_uri,
&workspace_id,
&metadata,
&index_state,
)?;
append_file_tree_reveal_rows(
&canonical_root,
@@ -2754,6 +2791,7 @@ fn load_local_folder_file_tree_scope_snapshot(
&root_source_uri,
&workspace_id,
&metadata,
&index_state,
&mut scan_result.rows,
)?;
@@ -6648,6 +6686,7 @@ fn scan_directory(
root_source_uri: &str,
workspace_id: &str,
metadata: &LocalFolderMetadata,
index_state: &LocalFileTreeIndexState,
rows: &mut Vec<LocalFolderRow>,
) -> Result<(), WebError> {
let mut entries = read_sorted_entries(directory, root)?;
@@ -6698,6 +6737,7 @@ fn scan_directory(
capabilities: local_entry_capabilities(&entry),
workspace_id: workspace_id.to_string(),
root_source_uri: root_source_uri.to_string(),
index_status: index_state.status_for_entry(&entry),
});
if entry.is_dir && !entry.is_symlink && max_depth.map_or(true, |limit| depth < limit) {
scan_directory(
@@ -6709,6 +6749,7 @@ fn scan_directory(
root_source_uri,
workspace_id,
metadata,
index_state,
rows,
)?;
}
@@ -6725,6 +6766,7 @@ fn scan_directory_shallow_with_revision(
root_source_uri: &str,
workspace_id: &str,
metadata: &LocalFolderMetadata,
index_state: &LocalFileTreeIndexState,
) -> Result<LocalFolderScanResult, WebError> {
let mut entries = read_sorted_entries(directory, root)?;
let parent_key = file_order_parent_key_for_directory(root, directory)?;
@@ -6797,6 +6839,7 @@ fn scan_directory_shallow_with_revision(
capabilities: local_entry_capabilities(entry),
workspace_id: workspace_id.to_string(),
root_source_uri: root_source_uri.to_string(),
index_status: index_state.status_for_entry(entry),
});
}
@@ -6867,6 +6910,7 @@ fn append_file_tree_reveal_rows(
root_source_uri: &str,
workspace_id: &str,
metadata: &LocalFolderMetadata,
index_state: &LocalFileTreeIndexState,
rows: &mut Vec<LocalFolderRow>,
) -> Result<(), WebError> {
let Some(reveal_relative_path) = reveal_relative_path
@@ -6924,6 +6968,7 @@ fn append_file_tree_reveal_rows(
root_source_uri,
workspace_id,
metadata,
index_state,
&mut scoped_rows,
)?;
for row in scoped_rows {
@@ -7124,9 +7169,25 @@ fn should_ignore_entry(relative_path: &str, file_name: &str) -> bool {
}
relative_path == ".mnote/trash"
|| relative_path.starts_with(".mnote/trash/")
|| is_local_index_artifact_entry(relative_path)
|| is_local_ocr_intermediate_entry(relative_path, file_name)
}
fn is_local_index_artifact_entry(relative_path: &str) -> bool {
let normalized = relative_path.trim().trim_matches('/').replace('\\', "/");
if normalized.is_empty() {
return false;
}
let segments = normalized
.split('/')
.filter(|segment| !segment.trim().is_empty())
.collect::<Vec<_>>();
segments
.windows(2)
.any(|window| window[0] == ".mnote" && window[1] == "index")
|| segments.iter().any(|segment| segment.ends_with(".ocr"))
}
fn is_local_ocr_intermediate_entry(relative_path: &str, file_name: &str) -> bool {
let normalized = relative_path.trim().trim_matches('/').replace('\\', "/");
if normalized.is_empty() {
@@ -7435,6 +7496,7 @@ fn scan_markdown_page_tree(
capabilities: local_entry_capabilities(&entry),
workspace_id: workspace_id.to_string(),
root_source_uri: root_source_uri.to_string(),
index_status: None,
});
directory_rows.extend(child_rows);
contains_markdown = true;
@@ -7483,6 +7545,7 @@ fn scan_markdown_page_tree(
capabilities: local_entry_capabilities(&entry),
workspace_id: workspace_id.to_string(),
root_source_uri: root_source_uri.to_string(),
index_status: None,
});
}
directory_rows.extend(child_rows);
@@ -7510,6 +7573,7 @@ fn scan_markdown_page_tree(
capabilities: local_entry_capabilities(&entry),
workspace_id: workspace_id.to_string(),
root_source_uri: root_source_uri.to_string(),
index_status: None,
});
continue;
}
@@ -7546,6 +7610,7 @@ fn scan_markdown_page_tree(
capabilities: local_entry_capabilities(&entry),
workspace_id: workspace_id.to_string(),
root_source_uri: root_source_uri.to_string(),
index_status: None,
});
contains_markdown = true;
}
@@ -7691,6 +7756,10 @@ fn local_folder_row_to_projection_item(row: &LocalFolderRow) -> Value {
item["assetId"] = Value::String(asset_id.clone());
item["resourceMeta"]["assetId"] = Value::String(asset_id);
}
if let Some(index_status) = row.index_status.as_deref() {
item["indexStatus"] = Value::String(index_status.to_string());
item["resourceMeta"]["indexStatus"] = Value::String(index_status.to_string());
}
// Phase A1: 为 File/Page/Resource tree 输出统一 workspacePath
let object_kind = match row.row_kind.as_str() {
@@ -11648,6 +11717,15 @@ fn main() {}
.expect("write md");
let root_uri = format!("file://{}", root.display());
crate::routes::local_search_index::write_local_index_settings(
&root,
&[String::from(".")],
None,
None,
None,
Some(true),
)
.expect("settings");
crate::routes::local_search_index::refresh_local_search_index(
&root,
&root_uri,
@@ -12369,6 +12447,15 @@ fn main() {}
std::fs::write(root.join("README.md"), "# Before\nold-token\n").expect("write md");
let root_uri = format!("file://{}", root.display());
crate::routes::local_search_index::write_local_index_settings(
&root,
&[String::from(".")],
None,
None,
None,
Some(true),
)
.expect("settings");
crate::routes::local_search_index::refresh_local_search_index(
&root,
&root_uri,
@@ -13474,7 +13561,7 @@ fn main() {}
}
#[test]
fn local_ocr_sidecar_is_filetree_resource_not_page_tree_document() {
fn local_ocr_sidecar_artifacts_are_hidden_from_filetree_and_page_tree() {
let root = temp_root("mnote-local-ocr-sidecar-page-tree");
init_workspace(&root);
let root_uri = format!("file://{}", root.display());
@@ -13515,15 +13602,22 @@ fn main() {}
)
.expect("ocr image asset");
let docs_file_tree =
load_local_folder_file_tree_children_snapshot(&root_uri, "docs").expect("file tree");
let docs_file_items = docs_file_tree.projection["items"]
.as_array()
.expect("file items");
assert!(!docs_file_items
.iter()
.any(|item| item["title"].as_str() == Some("Page.ocr")));
let file_tree = load_local_folder_file_tree_children_snapshot(&root_uri, "docs/Page.ocr")
.expect("file tree");
.expect("direct hidden sidecar file tree");
let file_items = file_tree.projection["items"]
.as_array()
.expect("file items");
assert!(file_items
.iter()
.any(|item| item["title"].as_str() == Some("photo.png.ocr.md")));
for hidden_title in [
"photo.png.ocr.md",
"layout.json",
"abc_content_list.json",
"abc_model.json",
@@ -13534,7 +13628,7 @@ fn main() {}
!file_items
.iter()
.any(|item| item["title"].as_str() == Some(hidden_title)),
"OCR 中间产物不应出现在 FileTree: {hidden_title}"
"OCR / 索引产物不应出现在 FileTree: {hidden_title}"
);
}
@@ -13551,6 +13645,104 @@ fn main() {}
let _ = std::fs::remove_dir_all(&root);
}
#[test]
fn local_file_tree_marks_indexed_and_failed_source_files() {
let root = temp_root("mnote-local-filetree-index-status");
init_workspace(&root);
std::fs::write(root.join("ok.pdf"), b"%PDF-1.4\nok").expect("ok pdf");
std::fs::write(root.join("failed.pdf"), b"%PDF-1.4\nfailed").expect("failed pdf");
std::fs::write(root.join("draft.md"), "# Draft\n").expect("draft");
let index_dir = root.join(".mnote").join("index");
std::fs::create_dir_all(&index_dir).expect("index dir");
let evidence_path = index_dir.join("evidence.sqlite");
let connection = rusqlite::Connection::open(&evidence_path).expect("evidence sqlite");
connection
.execute_batch(
r#"
CREATE TABLE evidence_resource(
resource_id TEXT PRIMARY KEY,
owner_document_id TEXT NOT NULL,
owner_document_path TEXT NOT NULL,
source_root_relative_path TEXT NOT NULL,
provider TEXT NOT NULL,
source_hash TEXT NOT NULL,
artifact_root_relative_path TEXT NOT NULL,
source_map_root_relative_path TEXT NOT NULL,
updated_at_ms INTEGER NOT NULL
);
"#,
)
.expect("evidence schema");
connection
.execute(
"INSERT INTO evidence_resource(resource_id, owner_document_id, owner_document_path, source_root_relative_path, provider, source_hash, artifact_root_relative_path, source_map_root_relative_path, updated_at_ms) VALUES(?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9)",
rusqlite::params![
"local-resource:ok.pdf#parse",
"local-resource:ok.pdf",
"ok.pdf",
"ok.pdf",
"liteparse",
"hash",
"ok.ocr/ok.pdf.parse.md",
"ok.ocr/ok.pdf.source-map.json",
1_i64,
],
)
.expect("insert indexed evidence");
connection
.execute(
"INSERT INTO evidence_resource(resource_id, owner_document_id, owner_document_path, source_root_relative_path, provider, source_hash, artifact_root_relative_path, source_map_root_relative_path, updated_at_ms) VALUES(?1, ?2, ?3, ?4, ?5, ?6, ?7, ?8, ?9)",
rusqlite::params![
"local-md:draft.md",
"local-md:draft.md",
"draft.md",
"draft.md",
"markdown",
"hash",
"draft.md",
"draft.md.source-map.json",
1_i64,
],
)
.expect("insert markdown evidence");
let root_uri = format!("file://{}", root.display());
let search_index = json!({
"version": 1,
"builtAt": 1,
"rootUri": root_uri,
"workspaceId": "local-filetree-index-status",
"indexedPaths": ["."],
"documents": [],
"resources": [
{"resourceId": "local-resource:ok.pdf", "resourceType": "pdf", "title": "ok", "path": "ok.pdf", "updatedAt": 1},
{"resourceId": "local-resource:failed.pdf", "resourceType": "pdf", "title": "failed", "path": "failed.pdf", "updatedAt": 1}
]
});
std::fs::write(
index_dir.join("search-index.json"),
format!("{}\n", serde_json::to_string_pretty(&search_index).unwrap()),
)
.expect("search index");
let file_tree = load_local_folder_file_tree_snapshot(&format!("file://{}", root.display()))
.expect("file tree");
let items = file_tree.projection["items"].as_array().expect("items");
let status_for = |path: &str| {
items
.iter()
.find(|item| {
item["resourceMeta"]["extra"]["source"]["relativePath"].as_str() == Some(path)
})
.and_then(|item| item["indexStatus"].as_str())
};
assert_eq!(status_for("ok.pdf"), Some("indexed"));
assert_eq!(status_for("failed.pdf"), Some("failed"));
assert_eq!(status_for("draft.md"), None);
let _ = std::fs::remove_dir_all(&root);
}
#[test]
fn local_rename_markdown_page_renames_nested_bundle() {
let root = temp_root("mnote-local-rename-nested-bundle");
@@ -15,6 +15,7 @@ use core_protocol::{
use rusqlite::{params, Connection, OptionalExtension};
use serde::{Deserialize, Serialize};
use serde_json::{json, Value};
use std::collections::BTreeSet;
use std::fs;
use std::path::{Component, Path, PathBuf};
use std::time::{SystemTime, UNIX_EPOCH};
@@ -134,13 +135,19 @@ pub(crate) fn query_local_search_index_with_settings(
)?;
let normalized_query = normalize_search_text(query);
let page_id = page_id.map(str::trim).filter(|value| !value.is_empty());
let page_resource_path = page_id.and_then(local_resource_path_from_document_id);
let markdown_page_id = if page_resource_path.is_some() {
None
} else {
page_id
};
let recent_changes = local_recent_changes_projection(&index.documents, root_uri);
let mut results = Vec::new();
for document in index.documents.iter() {
if !index_relative_path_is_included(&document.path, &result_settings.include_paths) {
continue;
}
if let Some(page_id) = page_id {
if let Some(page_id) = markdown_page_id {
if document.document_id != page_id {
continue;
}
@@ -157,11 +164,16 @@ pub(crate) fn query_local_search_index_with_settings(
break;
}
}
if page_id.is_none() && results.len() < limit.max(1) as usize {
if markdown_page_id.is_none() && results.len() < limit.max(1) as usize {
for resource in index.resources.iter() {
if !index_relative_path_is_included(&resource.path, &result_settings.include_paths) {
continue;
}
if let Some(resource_path) = page_resource_path.as_deref() {
if resource.path != resource_path {
continue;
}
}
if !local_search_resource_matches(resource, &normalized_query, title_only, exact) {
continue;
}
@@ -180,7 +192,11 @@ pub(crate) fn query_local_search_index_with_settings(
continue;
}
if let Some(page_id) = page_id {
if entry.owner_document_id != page_id {
if let Some(resource_path) = page_resource_path.as_deref() {
if entry.source_root_relative_path != resource_path {
continue;
}
} else if entry.owner_document_id != page_id {
continue;
}
}
@@ -364,6 +380,43 @@ pub(crate) fn write_user_local_index_settings(
Ok(settings)
}
pub(crate) fn preview_user_local_index_settings(
store: &dyn ControlPlaneStore,
user_id: &str,
workspace_id: &str,
root_path: &Path,
include_paths: &[String],
schedule_mode: Option<&str>,
schedule_time: Option<&str>,
schedule_date: Option<&str>,
run_on_change: Option<bool>,
) -> Result<LocalIndexSettings, WebError> {
let user_id = user_id.trim();
if user_id.is_empty() || user_id == "anonymous" {
return Err(WebError::bad_request_code(
"local_index_settings_auth_required",
"本地索引设置需要登录用户",
));
}
let existing = read_user_local_index_settings(store, user_id, workspace_id, root_path)?;
let include_paths = normalize_index_include_paths(root_path, include_paths)?;
let schedule_mode =
normalize_index_schedule_mode(schedule_mode.unwrap_or(&existing.schedule_mode))?;
let schedule_time =
normalize_index_schedule_time(schedule_time.unwrap_or(&existing.schedule_time))?;
let schedule_date =
normalize_index_schedule_date(schedule_date.or(existing.schedule_date.as_deref()))?;
Ok(LocalIndexSettings {
schema: LOCAL_INDEX_SETTINGS_SCHEMA.to_string(),
include_paths,
schedule_mode,
schedule_time,
schedule_date,
run_on_change: run_on_change.unwrap_or(existing.run_on_change),
updated_at: now_ms(),
})
}
pub(crate) fn effective_local_index_settings_for_root(
store: &dyn ControlPlaneStore,
workspace_id: &str,
@@ -452,7 +505,7 @@ fn local_index_status_for_settings(
let mut document_count = 0usize;
let mut resource_count = 0usize;
let mut built_at = Value::Null;
let mut cache_matches_settings = false;
let cache_matches_settings;
let scheduled_due = if let Some(index) = index.as_ref() {
document_count = index.documents.len();
resource_count = index.resources.len();
@@ -463,6 +516,7 @@ fn local_index_status_for_settings(
&& normalized_indexed_paths(&index.indexed_paths) == settings.include_paths;
local_index_schedule_is_due(&settings, index.built_at)
} else {
cache_matches_settings = settings.include_paths.is_empty();
local_index_schedule_is_due(&settings, 0)
};
let evidence_block_count = if evidence_path.exists() {
@@ -622,6 +676,16 @@ pub(crate) fn query_evidence_sqlite_results(
query: &str,
owner_document_id: Option<&str>,
limit: u32,
) -> Result<Option<Vec<EvidenceSearchResult>>, WebError> {
query_evidence_sqlite_results_with_mode(root_path, query, owner_document_id, limit, true)
}
pub(crate) fn query_evidence_sqlite_results_with_mode(
root_path: &Path,
query: &str,
owner_document_id: Option<&str>,
limit: u32,
exact: bool,
) -> Result<Option<Vec<EvidenceSearchResult>>, WebError> {
let path = evidence_sqlite_path(root_path);
if !path.exists() {
@@ -631,24 +695,54 @@ pub(crate) fn query_evidence_sqlite_results(
if normalized_query.is_empty() {
return Ok(Some(Vec::new()));
}
let owner_document_id = owner_document_id
.map(str::trim)
.filter(|value| !value.is_empty());
let resource_path = owner_document_id.and_then(local_resource_path_from_document_id);
let owner_document_id = if resource_path.is_some() {
None
} else {
owner_document_id
};
let connection = Connection::open(&path).map_err(|error| {
WebError::bad_request_code(
"evidence_index_open_failed",
format!("无法打开 evidence 索引 {}: {error}", path.display()),
)
})?;
if !exact {
return Ok(Some(query_evidence_sqlite_fuzzy(
&connection,
normalized_query,
owner_document_id,
resource_path.as_deref(),
limit,
)?));
}
let fts_query = evidence_fts_phrase(normalized_query);
let results = match query_evidence_sqlite_fts(
&connection,
&fts_query,
normalized_query,
owner_document_id,
resource_path.as_deref(),
limit,
) {
Ok(results) if results.is_empty() => query_evidence_sqlite_like(
&connection,
normalized_query,
owner_document_id,
resource_path.as_deref(),
limit,
)?,
Ok(results) => results,
Err(_) => {
query_evidence_sqlite_like(&connection, normalized_query, owner_document_id, limit)?
}
Err(_) => query_evidence_sqlite_like(
&connection,
normalized_query,
owner_document_id,
resource_path.as_deref(),
limit,
)?,
};
Ok(Some(results))
}
@@ -718,6 +812,9 @@ pub(crate) fn read_evidence_sqlite_context(
0.8
},
source: source.clone(),
citation_url: None,
citation_label: None,
citation_markdown: None,
})
.collect::<Vec<_>>();
Ok(Some(results))
@@ -804,6 +901,9 @@ fn graph_edge_row(row: &rusqlite::Row<'_>) -> rusqlite::Result<EvidenceSearchRes
block_id: source.block_id.or(Some(source_block_id)),
..source
},
citation_url: None,
citation_label: None,
citation_markdown: None,
})
}
@@ -824,6 +924,7 @@ fn query_evidence_sqlite_fts(
fts_query: &str,
display_query: &str,
owner_document_id: Option<&str>,
resource_path: Option<&str>,
limit: u32,
) -> Result<Vec<EvidenceSearchResult>, WebError> {
let mut sql = String::from(
@@ -835,8 +936,10 @@ fn query_evidence_sqlite_fts(
);
if owner_document_id.is_some() {
sql.push_str(" AND r.owner_document_id = ?2");
} else if resource_path.is_some() {
sql.push_str(" AND r.source_root_relative_path = ?2");
}
sql.push_str(if owner_document_id.is_some() {
sql.push_str(if owner_document_id.is_some() || resource_path.is_some() {
" ORDER BY rank LIMIT ?3"
} else {
" ORDER BY rank LIMIT ?2"
@@ -850,6 +953,13 @@ fn query_evidence_sqlite_fts(
})
.map_err(sqlite_error)?
.collect::<Result<Vec<_>, _>>()
} else if let Some(resource_path) = resource_path {
statement
.query_map(params![fts_query, resource_path, limit], |row| {
evidence_result_from_sqlite_row(row, display_query)
})
.map_err(sqlite_error)?
.collect::<Result<Vec<_>, _>>()
} else {
statement
.query_map(params![fts_query, limit], |row| {
@@ -865,6 +975,7 @@ fn query_evidence_sqlite_like(
connection: &Connection,
query: &str,
owner_document_id: Option<&str>,
resource_path: Option<&str>,
limit: u32,
) -> Result<Vec<EvidenceSearchResult>, WebError> {
let mut sql = String::from(
@@ -875,8 +986,10 @@ fn query_evidence_sqlite_like(
);
if owner_document_id.is_some() {
sql.push_str(" AND r.owner_document_id = ?2");
} else if resource_path.is_some() {
sql.push_str(" AND r.source_root_relative_path = ?2");
}
sql.push_str(if owner_document_id.is_some() {
sql.push_str(if owner_document_id.is_some() || resource_path.is_some() {
" LIMIT ?3"
} else {
" LIMIT ?2"
@@ -891,6 +1004,13 @@ fn query_evidence_sqlite_like(
})
.map_err(sqlite_error)?
.collect::<Result<Vec<_>, _>>()
} else if let Some(resource_path) = resource_path {
statement
.query_map(params![like_query, resource_path, limit], |row| {
evidence_result_from_sqlite_row(row, query)
})
.map_err(sqlite_error)?
.collect::<Result<Vec<_>, _>>()
} else {
statement
.query_map(params![like_query, limit], |row| {
@@ -902,6 +1022,80 @@ fn query_evidence_sqlite_like(
rows.map_err(sqlite_error)
}
fn query_evidence_sqlite_fuzzy(
connection: &Connection,
query: &str,
owner_document_id: Option<&str>,
resource_path: Option<&str>,
limit: u32,
) -> Result<Vec<EvidenceSearchResult>, WebError> {
let mut sql = String::from(
"SELECT b.block_id, b.text, b.locator_json, 0.0 AS rank \
FROM evidence_block b \
JOIN evidence_resource r ON r.resource_id = b.resource_id",
);
let first_char_like = query
.chars()
.find(|ch| !ch.is_whitespace())
.map(|ch| format!("%{}%", ch));
match (
owner_document_id.is_some() || resource_path.is_some(),
first_char_like.is_some(),
resource_path.is_some(),
) {
(true, true, true) => {
sql.push_str(" WHERE r.source_root_relative_path = ?1 AND b.text LIKE ?2 LIMIT ?3")
}
(true, true, false) => {
sql.push_str(" WHERE r.owner_document_id = ?1 AND b.text LIKE ?2 LIMIT ?3")
}
(true, false, true) => sql.push_str(" WHERE r.source_root_relative_path = ?1 LIMIT ?2"),
(true, false, false) => sql.push_str(" WHERE r.owner_document_id = ?1 LIMIT ?2"),
(false, true, _) => sql.push_str(" WHERE b.text LIKE ?1 LIMIT ?2"),
(false, false, _) => sql.push_str(" LIMIT ?1"),
}
let mut statement = connection.prepare(&sql).map_err(sqlite_error)?;
let scan_limit = i64::from(limit.max(1)) * 200;
let scope_value = owner_document_id.or(resource_path);
let rows = match (scope_value, first_char_like.as_deref()) {
(Some(scope_value), Some(first_char_like)) => statement
.query_map(params![scope_value, first_char_like, scan_limit], |row| {
evidence_result_from_sqlite_row(row, query)
})
.map_err(sqlite_error)?
.collect::<Result<Vec<_>, _>>(),
(Some(scope_value), None) => statement
.query_map(params![scope_value, scan_limit], |row| {
evidence_result_from_sqlite_row(row, query)
})
.map_err(sqlite_error)?
.collect::<Result<Vec<_>, _>>(),
(None, Some(first_char_like)) => statement
.query_map(params![first_char_like, scan_limit], |row| {
evidence_result_from_sqlite_row(row, query)
})
.map_err(sqlite_error)?
.collect::<Result<Vec<_>, _>>(),
(None, None) => statement
.query_map(params![scan_limit], |row| {
evidence_result_from_sqlite_row(row, query)
})
.map_err(sqlite_error)?
.collect::<Result<Vec<_>, _>>(),
}
.map_err(sqlite_error)?;
Ok(rows
.into_iter()
.filter(|result| {
fuzzy_search_match(
&normalize_search_text(&result.quote),
&normalize_search_text(query),
)
})
.take(limit.max(1) as usize)
.collect())
}
fn evidence_result_from_sqlite_row(
row: &rusqlite::Row<'_>,
query: &str,
@@ -922,6 +1116,9 @@ fn evidence_result_from_sqlite_row(
1.0 / (1.0 + rank.abs())
},
source,
citation_url: None,
citation_label: None,
citation_markdown: None,
})
}
@@ -945,6 +1142,18 @@ pub(crate) fn refresh_local_search_index_with_settings(
workspace_id: &str,
settings: &LocalIndexSettings,
) -> Result<Value, WebError> {
if settings.include_paths.is_empty() {
clear_local_search_index(root_path)?;
return Ok(json!({
"version": LOCAL_SEARCH_INDEX_VERSION,
"rootUri": root_uri,
"workspaceId": workspace_id,
"indexedPaths": [],
"builtAt": now_ms(),
"documentCount": 0,
"resourceCount": 0
}));
}
let index =
rebuild_local_search_index_with_settings(root_path, root_uri, workspace_id, settings)?;
Ok(json!({
@@ -1427,7 +1636,7 @@ fn parse_local_index_settings_value(
fn default_local_index_settings() -> LocalIndexSettings {
LocalIndexSettings {
schema: LOCAL_INDEX_SETTINGS_SCHEMA.to_string(),
include_paths: default_indexed_paths(),
include_paths: Vec::new(),
schedule_mode: default_index_schedule_mode(),
schedule_time: default_index_schedule_time(),
schedule_date: None,
@@ -1605,11 +1814,46 @@ fn normalize_index_include_paths(
if trimmed.is_empty() {
continue;
}
let raw_path = PathBuf::from(trimmed);
let normalized = if trimmed == "." || trimmed == "/" {
".".to_string()
} else if raw_path.is_absolute() {
let canonical = raw_path.canonicalize().map_err(|error| {
WebError::bad_request_code(
"local_index_scope_not_found",
format!("本地索引范围不存在 {}: {error}", raw_path.display()),
)
})?;
if !canonical.starts_with(&root_canonical) {
return Err(WebError::bad_request_code(
"local_index_scope_escape",
"本地索引目录必须位于当前授权 root 内",
));
}
if !canonical.is_dir() {
return Err(WebError::bad_request_code(
"local_index_scope_not_directory",
"本地索引范围必须是目录",
));
}
canonical
.strip_prefix(&root_canonical)
.map_err(|_| {
WebError::bad_request_code(
"local_index_scope_escape",
"本地索引目录必须位于当前授权 root 内",
)
})?
.to_string_lossy()
.replace('\\', "/")
} else {
normalize_index_relative_path(trimmed.trim_start_matches("./"))?
};
let normalized = if normalized.is_empty() {
".".to_string()
} else {
normalized
};
if normalized.split('/').any(|part| part == ".mnote") {
return Err(WebError::bad_request_code(
"local_index_scope_reserved",
@@ -1640,9 +1884,6 @@ fn normalize_index_include_paths(
}
output.push(normalized);
}
if output.is_empty() {
output.push(".".to_string());
}
Ok(normalized_indexed_paths(&output))
}
@@ -1884,6 +2125,27 @@ fn write_local_search_index_json(
})
}
fn clear_local_search_index(root_path: &Path) -> Result<(), WebError> {
let index_path = root_path
.join(".mnote")
.join("index")
.join("search-index.json");
let evidence_path = evidence_sqlite_path(root_path);
for path in [index_path, evidence_path] {
match fs::remove_file(&path) {
Ok(()) => {}
Err(error) if error.kind() == std::io::ErrorKind::NotFound => {}
Err(error) => {
return Err(WebError::bad_request_code(
"local_search_index_delete_failed",
format!("无法删除本地搜索索引文件 {}: {error}", path.display()),
));
}
}
}
Ok(())
}
fn ensure_evidence_sqlite_index(
root_path: &Path,
index: &LocalSearchIndex,
@@ -3151,6 +3413,72 @@ fn count_evidence_blocks(path: &Path) -> Result<u64, WebError> {
.map_err(sqlite_error)
}
#[derive(Debug, Clone, Default)]
pub(crate) struct LocalEvidenceSourceStatuses {
pub(crate) indexed_paths: BTreeSet<String>,
pub(crate) failed_paths: BTreeSet<String>,
}
pub(crate) fn local_evidence_source_statuses(
root_path: &Path,
) -> Result<LocalEvidenceSourceStatuses, WebError> {
let mut statuses = LocalEvidenceSourceStatuses::default();
let evidence_path = evidence_sqlite_path(root_path);
if evidence_path.exists() {
let connection = Connection::open(&evidence_path).map_err(|error| {
WebError::bad_request_code(
"evidence_index_open_failed",
format!(
"无法打开 evidence 索引 {}: {error}",
evidence_path.display()
),
)
})?;
let mut statement = connection
.prepare(
"SELECT DISTINCT source_root_relative_path \
FROM evidence_resource \
WHERE provider NOT IN ('resource', 'markdown')",
)
.map_err(sqlite_error)?;
let rows = statement
.query_map([], |row| row.get::<_, String>(0))
.map_err(sqlite_error)?;
for row in rows {
if let Ok(path) = row {
if !path.trim().is_empty() {
statuses.indexed_paths.insert(path);
}
}
}
}
if let Some(index) = read_local_search_index(root_path)? {
for resource in index.resources {
if matches!(resource.resource_type.as_str(), "pdf" | "office")
&& !statuses.indexed_paths.contains(&resource.path)
{
statuses.failed_paths.insert(resource.path);
}
}
}
for entry in local_ocr::ocr_index_entries(root_path)? {
match entry.status.as_str() {
"done" => {
statuses
.indexed_paths
.insert(entry.source_root_relative_path);
}
"failed" | "interrupted" => {
statuses
.failed_paths
.insert(entry.source_root_relative_path);
}
_ => {}
}
}
Ok(statuses)
}
fn evidence_resource_kind_for_type(resource_type: &str) -> &'static str {
match resource_type {
"mindmap" => "mindmap",
@@ -3179,6 +3507,14 @@ fn evidence_resource_kind_for_path(path: &str) -> &'static str {
}
}
fn local_resource_path_from_document_id(document_id: &str) -> Option<String> {
document_id
.trim()
.strip_prefix("local-resource:")
.map(|value| value.replace("~2F", "/").replace("~2f", "/"))
.filter(|value| !value.trim().is_empty())
}
fn path_source_map_path(path: &str) -> Option<String> {
if let Some(stripped) = path.strip_suffix(".ocr.md") {
return Some(format!("{stripped}.source-map.json"));
@@ -3267,9 +3603,9 @@ fn local_search_document_matches(
))
};
if exact {
haystack == query
} else {
haystack.contains(query)
} else {
fuzzy_search_match(&haystack, query)
}
}
@@ -3291,9 +3627,9 @@ fn local_search_resource_matches(
))
};
if exact {
haystack == query
} else {
haystack.contains(query)
} else {
fuzzy_search_match(&haystack, query)
}
}
@@ -3319,9 +3655,9 @@ fn local_search_ocr_matches(
))
};
if exact {
haystack == query
} else {
haystack.contains(query)
} else {
fuzzy_search_match(&haystack, query)
}
}
@@ -3330,6 +3666,7 @@ fn local_search_document_projection(
root_uri: &str,
query: &str,
) -> Value {
let hit = search_document_hit(document, query);
json!({
"id": document.document_id,
"documentId": document.document_id,
@@ -3338,7 +3675,12 @@ fn local_search_document_projection(
"resourceType": "markdown",
"sourceKind": "local_folder",
"rootUri": root_uri,
"snippet": search_snippet(document, query),
"snippet": hit.snippet,
"blockId": hit.block_id,
"lineRange": {
"start": hit.line_number,
"end": hit.line_number,
},
"tags": document.tags,
"backlinks": document.backlinks,
"resourceRefs": document.resource_refs,
@@ -3548,16 +3890,43 @@ fn is_local_resource_reference(target: &str) -> bool {
}
fn search_snippet(document: &LocalSearchDocument, query: &str) -> String {
for line in document.raw_text.lines() {
search_document_hit(document, query).snippet
}
#[derive(Debug, Clone)]
struct SearchDocumentHit {
snippet: String,
line_number: usize,
block_id: String,
}
fn search_document_hit(document: &LocalSearchDocument, query: &str) -> SearchDocumentHit {
for (line_index, line) in document.raw_text.lines().enumerate() {
let trimmed = line.trim();
if trimmed.is_empty() {
continue;
}
if normalize_search_text(trimmed).contains(query) {
return trimmed.chars().take(180).collect();
let line_number = line_index + 1;
return SearchDocumentHit {
snippet: trimmed.chars().take(180).collect(),
line_number,
block_id: format!("{}#line{}", document.document_id, line_number),
};
}
}
document.raw_text.chars().take(180).collect()
let line_number = document
.raw_text
.lines()
.enumerate()
.find(|(_, line)| !line.trim().is_empty())
.map(|(line_index, _)| line_index + 1)
.unwrap_or(1);
SearchDocumentHit {
snippet: document.raw_text.chars().take(180).collect(),
line_number,
block_id: format!("{}#line{}", document.document_id, line_number),
}
}
fn ocr_search_snippet(body: &str, query: &str) -> String {
@@ -3576,11 +3945,73 @@ fn ocr_search_snippet(body: &str, query: &str) -> String {
.map(|(idx, _)| idx)
.unwrap_or(0);
normalized_body[start..].chars().take(160).collect()
} else if let Some(byte_index) = fuzzy_search_start_byte(&normalized_body, normalized_query) {
let start = normalized_body[..byte_index]
.char_indices()
.rev()
.nth(40)
.map(|(idx, _)| idx)
.unwrap_or(0);
normalized_body[start..].chars().take(180).collect()
} else {
normalized_body.chars().take(160).collect()
}
}
fn fuzzy_search_match(haystack: &str, query: &str) -> bool {
if query.is_empty() {
return false;
}
if haystack.contains(query) {
return true;
}
let mut query_chars = query.chars().filter(|ch| !ch.is_whitespace());
let Some(mut wanted) = query_chars.next() else {
return false;
};
for ch in haystack.chars().filter(|ch| !ch.is_whitespace()) {
if ch == wanted {
match query_chars.next() {
Some(next) => wanted = next,
None => return true,
}
}
}
false
}
fn fuzzy_search_start_byte(haystack: &str, query: &str) -> Option<usize> {
if query.is_empty() {
return None;
}
let query_chars = query
.chars()
.filter(|ch| !ch.is_whitespace())
.collect::<Vec<_>>();
if query_chars.is_empty() {
return None;
}
let haystack_chars = haystack.char_indices().collect::<Vec<_>>();
for (start_index, (byte_index, ch)) in haystack_chars.iter().enumerate() {
if ch != &query_chars[0] {
continue;
}
let mut query_index = 1usize;
for (_, next_ch) in haystack_chars.iter().skip(start_index + 1) {
if next_ch.is_whitespace() {
continue;
}
if query_index < query_chars.len() && next_ch == &query_chars[query_index] {
query_index += 1;
if query_index >= query_chars.len() {
return Some(*byte_index);
}
}
}
}
None
}
fn is_markdown_path(path: &Path) -> bool {
path.extension()
.and_then(|value| value.to_str())
@@ -3905,6 +4336,89 @@ mod tests {
let _ = fs::remove_dir_all(&root);
}
#[test]
fn local_index_settings_accepts_absolute_path_under_root_as_frozen_scope() {
let root = temp_root("mnote-local-index-settings-absolute");
fs::create_dir_all(root.join("docs").join("absolute")).expect("create absolute dir");
let absolute_scope = root.join("docs").join("absolute");
let settings = write_local_index_settings(
&root,
&[absolute_scope.to_string_lossy().to_string()],
None,
None,
None,
None,
)
.expect("absolute scope under root");
assert_eq!(settings.include_paths, vec![String::from("docs/absolute")]);
let _ = fs::remove_dir_all(&root);
}
#[test]
fn user_local_index_empty_scope_deletes_index_files() {
let root = temp_root("mnote-local-index-empty-delete");
fs::write(
root.join("docs").join("keep.md"),
"# Keep\nDeleteIndexToken\n",
)
.expect("write doc");
let root_uri = format!("file://{}", root.display());
let workspace_id = "local-ws-empty-delete";
let store = SqliteControlPlaneStore::in_memory().expect("init control plane");
write_user_local_index_settings(
&store,
"alice",
workspace_id,
&root,
&[String::from("docs")],
Some("manual"),
Some("02:00"),
None,
Some(false),
)
.expect("write indexed scope");
let effective = effective_local_index_settings_for_root(&store, workspace_id, &root)
.expect("effective indexed");
refresh_local_search_index_with_settings(&root, &root_uri, workspace_id, &effective)
.expect("refresh indexed");
assert!(root.join(".mnote/index/search-index.json").exists());
assert!(root.join(".mnote/index/evidence.sqlite").exists());
write_user_local_index_settings(
&store,
"alice",
workspace_id,
&root,
&[],
Some("manual"),
Some("02:00"),
None,
Some(false),
)
.expect("delete indexed scopes");
let empty_effective = effective_local_index_settings_for_root(&store, workspace_id, &root)
.expect("effective empty");
assert!(empty_effective.include_paths.is_empty());
refresh_local_search_index_with_settings(&root, &root_uri, workspace_id, &empty_effective)
.expect("clear index files");
assert!(!root.join(".mnote/index/search-index.json").exists());
assert!(!root.join(".mnote/index/evidence.sqlite").exists());
let status = local_index_status_with_settings(
&root,
&root_uri,
workspace_id,
&empty_effective,
&empty_effective,
)
.expect("status");
assert_eq!(status["cacheMatchesSettings"].as_bool(), Some(true));
let _ = fs::remove_dir_all(&root);
}
#[test]
fn local_index_settings_rejects_escape_and_internal_cache_scope() {
let root = temp_root("mnote-local-index-settings-escape");
@@ -3949,6 +4463,14 @@ mod tests {
initial_status["settings"]["runOnChange"].as_bool(),
Some(false)
);
assert_eq!(
initial_status["settings"]["includePaths"]
.as_array()
.map(Vec::len),
Some(0),
"默认不应索引工作区根目录;用户新增范围后才开始索引"
);
assert_eq!(initial_status["cacheMatchesSettings"].as_bool(), Some(true));
let settings = write_local_index_settings(
&root,
@@ -4356,6 +4878,8 @@ mod tests {
)
.expect("write child");
write_local_index_settings(&root, &[String::from(".")], None, None, None, None)
.expect("settings");
refresh_local_search_index(&root, &root_uri, workspace_id).expect("initial refresh");
fs::write(
root.join("docs").join("child.md"),
@@ -4643,6 +5167,8 @@ mod tests {
)
.expect("write child");
write_local_index_settings(&root, &[String::from(".")], None, None, None, None)
.expect("settings");
refresh_local_search_index(&root, &root_uri, workspace_id).expect("refresh");
let results = query_evidence_sqlite_results(&root, "EvidenceToken", None, 10)
@@ -4750,6 +5276,8 @@ mod tests {
)
.expect("write home");
write_local_index_settings(&root, &[String::from(".")], None, None, None, None)
.expect("settings");
refresh_local_search_index(&root, &root_uri, workspace_id).expect("refresh");
let results = query_evidence_sqlite_results(&root, "ReadContextToken", None, 10)
.expect("sqlite query")
@@ -4821,6 +5349,8 @@ JSON
let old_bin = std::env::var("MNOTE_LITEPARSE_BIN").ok();
std::env::set_var("MNOTE_LITEPARSE_BIN", &lit);
write_local_index_settings(&root, &[String::from(".")], None, None, None, None)
.expect("settings");
refresh_local_search_index(&root, &root_uri, workspace_id).expect("refresh");
if let Some(value) = old_bin {
@@ -4850,6 +5380,21 @@ JSON
hit.source.source_map_path.as_deref(),
Some("docs/Page.ocr/spec.pdf.source-map.json")
);
let resource_scoped = query_evidence_sqlite_results_with_mode(
&root,
"ResourceBodyToken",
Some("local-resource:docs~2FPage.assets~2Fspec.pdf"),
10,
true,
)
.expect("resource scoped sqlite query")
.expect("sqlite exists");
assert_eq!(resource_scoped.len(), 1);
assert_eq!(
resource_scoped[0].source.resource_path.as_deref(),
Some("docs/Page.assets/spec.pdf"),
"页面内搜索打开资源时应只限制到当前资源,而不是它的 owner Markdown"
);
assert!(root
.join("docs")
.join("Page.ocr")
+14 -1
View File
@@ -37,16 +37,21 @@ pub(crate) mod ui_preferences;
pub(crate) mod web_shell;
mod ws;
pub(crate) use gateway::current_actor_id;
pub(crate) use local_folder_source::{
decode_local_id_segment, ensure_local_path_read_access, ensure_local_workspace_access,
ensure_local_workspace_read_access_with_state, ensure_local_workspace_write_access_with_state,
local_markdown_conflict_detection_key, local_workspace_id_from_root_uri,
update_local_markdown_title, write_local_markdown_page_body,
};
#[cfg(test)]
pub(crate) use local_search_index::write_local_index_settings;
pub(crate) use local_search_index::{
refresh_local_search_index_for_change_path_with_store,
effective_local_index_settings_for_root, local_index_status_with_settings,
preview_user_local_index_settings, read_local_index_settings_or_default,
read_user_local_index_settings, refresh_local_search_index_for_change_path_with_store,
refresh_local_search_index_if_scheduled_due_with_store,
refresh_local_search_index_with_settings, write_user_local_index_settings, LocalIndexSettings,
};
use crate::app::AppState;
@@ -641,6 +646,14 @@ pub fn build_router(state: AppState) -> Router {
)
.route("/client/skills", get(hermes_client::list_skills))
.route("/client/skills/toggle", put(hermes_client::toggle_skill))
.route(
"/client/capabilities",
get(hermes_client::list_capabilities),
)
.route(
"/client/capabilities/toggle",
put(hermes_client::toggle_capability),
)
.route("/client/tools/toggle", put(hermes_client::toggle_tool))
.route("/client/runs", post(hermes_client::create_run))
.route(
+302
View File
@@ -229,6 +229,25 @@ pub async fn documents(
EvidenceSearchMode::Hybrid,
&normalized_query,
);
let limit = body.limit.unwrap_or(30).max(1) as usize;
let direct_evidence_results = if !filters.title_only.unwrap_or(false) {
local_search_index::query_evidence_sqlite_results_with_mode(
&root_path,
&normalized_query,
page_id.as_deref(),
body.limit.unwrap_or(30),
filters.exact.unwrap_or(false),
)?
.unwrap_or_default()
} else {
Vec::new()
};
let (result, evidence_results) = merge_local_search_with_evidence_results(
result,
evidence_results,
direct_evidence_results,
limit,
);
(result, evidence_results)
} else {
let result = load_search_results_with_filters(
@@ -419,6 +438,12 @@ pub async fn update_local_index_settings(
&effective_workspace_id,
&root_path,
)?;
let refreshed = local_search_index::refresh_local_search_index_with_settings(
&root_path,
root_uri,
&effective_workspace_id,
&effective_settings,
)?;
let status = local_search_index::local_index_status_with_settings(
&root_path,
root_uri,
@@ -434,6 +459,7 @@ pub async fn update_local_index_settings(
Json(json!({
"ok": true,
"settings": settings,
"index": refreshed,
"result": status,
"meta": {
"owner": "mnote-web",
@@ -446,6 +472,75 @@ pub async fn update_local_index_settings(
))
}
fn merge_local_search_with_evidence_results(
mut result: Value,
mut evidence_results: Vec<EvidenceSearchResult>,
direct_evidence_results: Vec<EvidenceSearchResult>,
limit: usize,
) -> (Value, Vec<EvidenceSearchResult>) {
if direct_evidence_results.is_empty() {
return (result, evidence_results);
}
let mut result_items = result
.get("results")
.and_then(Value::as_array)
.cloned()
.unwrap_or_default();
let mut seen_evidence_ids = evidence_results
.iter()
.map(|item| item.evidence_id.clone())
.collect::<std::collections::HashSet<_>>();
for evidence in direct_evidence_results {
if result_items.len() >= limit || !seen_evidence_ids.insert(evidence.evidence_id.clone()) {
continue;
}
result_items.push(search_result_from_evidence(&evidence));
evidence_results.push(evidence);
}
if let Some(map) = result.as_object_mut() {
map.insert("results".into(), Value::Array(result_items));
}
(result, evidence_results)
}
fn search_result_from_evidence(evidence: &EvidenceSearchResult) -> Value {
let source = &evidence.source;
let evidence_value = serde_json::to_value(evidence).unwrap_or(Value::Null);
let resource_path = source
.resource_path
.as_deref()
.unwrap_or(source.owner_document_path.as_str());
let title = std::path::Path::new(resource_path)
.file_name()
.and_then(|value| value.to_str())
.unwrap_or(resource_path)
.to_string();
let resource_type = match source.resource_kind {
core_protocol::EvidenceResourceKind::Markdown => "markdown",
core_protocol::EvidenceResourceKind::Pdf => "pdf",
core_protocol::EvidenceResourceKind::Image => "image",
core_protocol::EvidenceResourceKind::Office => "office",
core_protocol::EvidenceResourceKind::Mindmap => "mindmap",
core_protocol::EvidenceResourceKind::RawFile => "resource",
};
json!({
"id": format!("evidence:{}", evidence.evidence_id),
"documentId": source.owner_document_id,
"title": title,
"path": source.owner_document_path,
"resourceType": resource_type,
"sourceKind": "local_folder",
"rootUri": source.root_uri,
"snippet": evidence.quote,
"score": evidence.score,
"publicPath": source.open_action.url,
"evidence": evidence_value,
"source": {
"locator": source
},
})
}
fn attach_evidence_to_search_results(
results: Value,
evidence_results: &[EvidenceSearchResult],
@@ -463,6 +558,9 @@ fn attach_evidence_to_search_results(
};
let mut item = item;
if let Some(map) = item.as_object_mut() {
if map.get("evidence").is_some() {
return item;
}
let evidence_value = serde_json::to_value(evidence).unwrap_or(Value::Null);
map.insert("evidence".into(), evidence_value);
let source = map.entry("source").or_insert_with(|| json!({}));
@@ -793,6 +891,7 @@ fn stamp_search_headers(headers: &mut HeaderMap) {
#[cfg(test)]
mod tests {
use crate::app::{build_app, AppConfig, AppState};
use crate::routes::local_search_index;
use axum::body::{to_bytes, Body};
use axum::http::{Request, StatusCode};
use control_plane::{DirectoryGrantInput, UpsertUserInput};
@@ -998,6 +1097,15 @@ mod tests {
.expect("readme");
fs::write(root.join("docs").join("child.md"), "# Child\nalpha child\n").expect("child");
let root_uri = format!("file://{}", root.display());
local_search_index::write_local_index_settings(
&root,
&[String::from(".")],
None,
None,
None,
None,
)
.expect("settings");
let response = app()
.oneshot(
@@ -1102,6 +1210,104 @@ mod tests {
let _ = fs::remove_dir_all(&root);
}
#[tokio::test]
async fn search_documents_local_folder_includes_evidence_sqlite_body_hits() {
let root = std::env::temp_dir().join(format!(
"mnote-local-search-evidence-route-{}",
std::process::id()
));
let _ = fs::remove_dir_all(&root);
fs::create_dir_all(root.join(".mnote")).expect("metadata");
fs::write(
root.join(".mnote").join("workspace.json"),
r#"{"workspaceId":"local-ws-search-evidence","ownerId":"user_test","createdAt":"2026-05-19T00:00:00Z","capabilities":["local_files"]}"#,
)
.expect("manifest");
fs::write(
root.join("README.md"),
"# Search Home\nBodyOnlyEvidenceToken only exists inside evidence.sqlite after refresh.\n",
)
.expect("readme");
let root_uri = format!("file://{}", root.display());
local_search_index::write_local_index_settings(
&root,
&[String::from(".")],
None,
None,
None,
None,
)
.expect("settings");
local_search_index::refresh_local_search_index(
&root,
&root_uri,
"local-ws-search-evidence",
)
.expect("refresh");
fs::write(
root.join(".mnote").join("index").join("search-index.json"),
serde_json::to_string_pretty(&json!({
"version": 1,
"builtAt": 1,
"rootUri": root_uri,
"workspaceId": "local-ws-search-evidence",
"documents": [],
"resources": []
}))
.expect("stale search index"),
)
.expect("overwrite search index");
let response = app()
.oneshot(
Request::builder()
.method("POST")
.uri("/api/search/documents")
.header("content-type", "application/json")
.header("x-mnote-actor-id", "user_test")
.header("x-mnote-actor-type", "user")
.body(Body::from(
json!({
"workspaceId": "local-ws-search-evidence",
"sourceKind": "local_folder",
"rootUri": root_uri,
"query": "BodyOnlyEvidenceToken",
"limit": 10,
"filters": {
"titleOnly": false,
"includeOcr": true
}
})
.to_string(),
))
.expect("request"),
)
.await
.expect("response");
assert_eq!(response.status(), StatusCode::OK);
let body = to_bytes(response.into_body(), usize::MAX)
.await
.expect("body");
let payload: Value = serde_json::from_slice(&body).expect("json");
let results = payload["results"].as_array().expect("results");
let hit = results
.iter()
.find(|item| {
item["snippet"]
.as_str()
.unwrap_or("")
.contains("BodyOnlyEvidenceToken")
})
.expect("evidence sqlite body hit should be promoted to search result");
assert_eq!(hit["resourceType"].as_str(), Some("markdown"));
assert_eq!(
hit["evidence"]["source"]["schema"].as_str(),
Some("mnote.evidence_locator.v1")
);
let _ = fs::remove_dir_all(&root);
}
#[tokio::test]
async fn search_local_index_refresh_rebuilds_authorized_root() {
let root = std::env::temp_dir().join(format!(
@@ -1156,6 +1362,102 @@ mod tests {
let _ = fs::remove_dir_all(&root);
}
#[tokio::test]
async fn local_index_settings_empty_scope_deletes_index_files() {
let root = std::env::temp_dir().join(format!(
"mnote-local-search-settings-delete-route-{}",
std::process::id()
));
let _ = fs::remove_dir_all(&root);
fs::create_dir_all(root.join(".mnote")).expect("metadata");
fs::create_dir_all(root.join("docs")).expect("docs");
fs::write(
root.join(".mnote").join("workspace.json"),
r#"{"workspaceId":"local-ws-settings-delete","ownerId":"user_test","createdAt":"2026-05-19T00:00:00Z","capabilities":["local_files","search"]}"#,
)
.expect("manifest");
fs::write(
root.join("docs").join("keep.md"),
"# Keep\nRouteDeleteToken\n",
)
.expect("doc");
let root_uri = format!("file://{}", root.display());
let app = app();
let create_response = app
.clone()
.oneshot(
Request::builder()
.method("PUT")
.uri("/api/search/local-index/settings")
.header("content-type", "application/json")
.header("x-mnote-actor-id", "user_test")
.header("x-mnote-actor-type", "user")
.body(Body::from(
json!({
"workspaceId": "local-ws-settings-delete",
"rootUri": root_uri,
"includePaths": ["docs"],
"scheduleMode": "manual",
"scheduleTime": "02:00",
"runOnChange": false
})
.to_string(),
))
.expect("create settings request"),
)
.await
.expect("create settings response");
assert_eq!(create_response.status(), StatusCode::OK);
assert!(root.join(".mnote/index/search-index.json").exists());
assert!(root.join(".mnote/index/evidence.sqlite").exists());
let delete_response = app
.oneshot(
Request::builder()
.method("PUT")
.uri("/api/search/local-index/settings")
.header("content-type", "application/json")
.header("x-mnote-actor-id", "user_test")
.header("x-mnote-actor-type", "user")
.body(Body::from(
json!({
"workspaceId": "local-ws-settings-delete",
"rootUri": root_uri,
"includePaths": [],
"scheduleMode": "manual",
"scheduleTime": "02:00",
"runOnChange": false
})
.to_string(),
))
.expect("delete settings request"),
)
.await
.expect("delete settings response");
assert_eq!(delete_response.status(), StatusCode::OK);
let delete_body = to_bytes(delete_response.into_body(), usize::MAX)
.await
.expect("delete body");
let delete_payload: Value = serde_json::from_slice(&delete_body).expect("delete json");
assert_eq!(
delete_payload["index"]["indexedPaths"]
.as_array()
.map(Vec::len),
Some(0)
);
assert_eq!(
delete_payload["result"]["settings"]["includePaths"]
.as_array()
.map(Vec::len),
Some(0)
);
assert!(!root.join(".mnote/index/search-index.json").exists());
assert!(!root.join(".mnote/index/evidence.sqlite").exists());
let _ = fs::remove_dir_all(&root);
}
#[tokio::test]
async fn search_local_index_backlinks_and_tags_read_authorized_root() {
let root = std::env::temp_dir().join(format!(
+11
View File
@@ -660,6 +660,17 @@ pub(crate) fn collect_filetree_render_rows(
object_identity: resource_meta
.and_then(|meta| meta.get("objectIdentity"))
.and_then(|value| serde_json::to_string(value).ok()),
index_status: item
.get("indexStatus")
.and_then(Value::as_str)
.or_else(|| {
resource_meta
.and_then(|meta| meta.get("indexStatus"))
.and_then(Value::as_str)
})
.map(str::trim)
.filter(|value| *value == "indexed" || *value == "failed")
.map(ToOwned::to_owned),
selected,
})
})
+330 -1
View File
@@ -986,6 +986,14 @@ pub struct OfficePreviewQuery {
source_kind: Option<String>,
root_uri: Option<String>,
document_id: Option<String>,
#[serde(default)]
page: Option<u32>,
#[serde(default)]
bbox: Option<String>,
#[serde(default, alias = "sourceMapPath")]
source_map_path: Option<String>,
#[serde(default, alias = "blockId")]
block_id: Option<String>,
}
pub async fn office_preview_page(Query(query): Query<OfficePreviewQuery>) -> Response {
@@ -1013,6 +1021,13 @@ pub async fn office_preview_page(Query(query): Query<OfficePreviewQuery>) -> Res
let source_kind = query.source_kind.unwrap_or_default();
let root_uri = query.root_uri.unwrap_or_default();
let document_id = query.document_id.unwrap_or_default();
let target_page = query
.page
.map(|value| value.to_string())
.unwrap_or_default();
let target_bbox = query.bbox.unwrap_or_default();
let target_source_map_path = query.source_map_path.unwrap_or_default();
let target_block_id = query.block_id.unwrap_or_default();
let html = format!(
r#"<!doctype html>
<html lang="zh-CN">
@@ -1047,6 +1062,10 @@ pub async fn office_preview_page(Query(query): Query<OfficePreviewQuery>) -> Res
.mnote-office-viewer .mnote-docx-wrapper {{ width: 100%; background: transparent !important; padding: 0 !important; align-items: stretch !important; }}
.mnote-office-viewer .docx-wrapper > section.docx,
.mnote-office-viewer .mnote-docx-wrapper > section.mnote-docx {{ width: 100% !important; max-width: none !important; margin: 0 0 14px !important; box-shadow: 0 2px 10px rgba(25, 25, 22, .08); }}
.mnote-office-viewer {{ position: relative; }}
.mnote-office-viewer [data-mnote-office-evidence-target="true"] {{ outline: 0; border-radius: 2px; background: #FFE9E6; color: #D83A32; box-shadow: 0 0 0 1px rgba(216, 58, 50, .22); }}
.mnote-office-evidence-marker {{ position: absolute; z-index: 3; left: 24px; max-width: min(720px, calc(100% - 48px)); padding: 6px 10px; border: 1px solid rgba(216, 58, 50, .45); border-radius: 6px; background: rgba(255, 249, 248, .96); color: #D83A32; font-size: 13px; line-height: 1.5; box-shadow: 0 2px 10px rgba(15, 23, 42, .12); }}
.mnote-office-evidence-marker[data-mnote-office-evidence-marker-mode="range"] {{ pointer-events: none; background: rgba(255, 233, 230, .72); box-shadow: 0 0 0 1px rgba(216, 58, 50, .28); }}
.mnote-office-pptx-stage {{ width: 100%; overflow: auto; }}
.mnote-office-pptx-stage .pptx-preview-wrapper {{ max-width: 100%; background: transparent !important; }}
.mnote-office-pptx-stage .pptx-preview-slide-wrapper {{ box-shadow: 0 2px 10px rgba(25, 25, 22, .08); }}
@@ -1055,7 +1074,7 @@ pub async fn office_preview_page(Query(query): Query<OfficePreviewQuery>) -> Res
}}
</style>
</head>
<body data-file-url="{file_url}" data-file-name="{file_name}" data-file-type="{file_type}" data-page-width-content-type="{page_width_content_type}" data-workspace-id="{workspace_id}" data-mnote-source-kind="{source_kind}" data-mnote-root-uri="{root_uri}" data-document-id="{document_id}">
<body data-file-url="{file_url}" data-file-name="{file_name}" data-file-type="{file_type}" data-page-width-content-type="{page_width_content_type}" data-workspace-id="{workspace_id}" data-mnote-source-kind="{source_kind}" data-mnote-root-uri="{root_uri}" data-document-id="{document_id}" data-evidence-page="{target_page}" data-evidence-bbox="{target_bbox}" data-evidence-source-map-path="{target_source_map_path}" data-evidence-block-id="{target_block_id}">
<main class="mnote-office-preview">
<section class="mnote-office-viewer" id="mnote-office-viewer"></section>
</main>
@@ -1070,6 +1089,10 @@ pub async fn office_preview_page(Query(query): Query<OfficePreviewQuery>) -> Res
const fileName = body.dataset.fileName || '';
const fileType = (body.dataset.fileType || '').toLowerCase();
const pageWidthContentType = body.dataset.pageWidthContentType || 'word';
let evidencePage = Number(body.dataset.evidencePage || 0);
let evidenceBbox = body.dataset.evidenceBbox || '';
let evidenceSourceMapPath = body.dataset.evidenceSourceMapPath || '';
let evidenceBlockId = body.dataset.evidenceBlockId || '';
let currentPptxBuffer = null;
let pptxRenderToken = 0;
let pptxResizeTimer = 0;
@@ -1136,6 +1159,299 @@ pub async fn office_preview_page(Query(query): Query<OfficePreviewQuery>) -> Res
viewer.append(message);
}}
function normalizeEvidenceText(value) {{
return String(value || '').replace(/\s+/g, ' ').trim();
}}
function markEvidenceTarget(target) {{
if (!(target instanceof HTMLElement)) return false;
viewer.querySelectorAll('[data-mnote-office-evidence-target="true"]').forEach(node => {{
if (node instanceof HTMLElement) node.removeAttribute('data-mnote-office-evidence-target');
}});
target.setAttribute('data-mnote-office-evidence-target', 'true');
window.setTimeout(() => target.scrollIntoView({{ block: 'center', inline: 'nearest' }}), 0);
document.documentElement.setAttribute('data-mnote-office-evidence-applied', 'true');
return true;
}}
function normalizedTextWithRawOffsets(value) {{
const raw = String(value || '');
let text = '';
const offsets = [];
let previousWhitespace = true;
for (let index = 0; index < raw.length; index += 1) {{
const ch = raw[index];
if (/\s/.test(ch)) {{
if (text && !previousWhitespace) {{
text += ' ';
offsets.push(index);
}}
previousWhitespace = true;
}} else {{
text += ch;
offsets.push(index);
previousWhitespace = false;
}}
}}
if (text.endsWith(' ')) {{
text = text.slice(0, -1);
offsets.pop();
}}
return {{ text, offsets }};
}}
function compactTextWithRawOffsets(value) {{
const raw = String(value || '');
let text = '';
const offsets = [];
for (let index = 0; index < raw.length; index += 1) {{
const ch = raw[index];
if (/\s/.test(ch)) continue;
text += ch;
offsets.push(index);
}}
return {{ text, offsets }};
}}
function wrapEvidenceTextNode(node, needle) {{
if (!(node instanceof Text)) return null;
const raw = String(node.textContent || '');
let start = raw.indexOf(needle);
let end = start >= 0 ? start + needle.length : -1;
if (start < 0) {{
const compact = compactTextWithRawOffsets(raw);
const compactNeedle = normalizeEvidenceText(needle).replace(/\s+/g, '');
let normalizedStart = compact.text.indexOf(compactNeedle);
let sourceOffsets = compact.offsets;
if (normalizedStart < 0) {{
const mapped = normalizedTextWithRawOffsets(raw);
const mappedNeedle = normalizeEvidenceText(needle);
normalizedStart = mapped.text.indexOf(mappedNeedle);
sourceOffsets = mapped.offsets;
if (normalizedStart < 0) return null;
start = sourceOffsets[normalizedStart];
end = sourceOffsets[normalizedStart + mappedNeedle.length - 1] + 1;
if (!Number.isFinite(start) || !Number.isFinite(end) || start < 0 || end <= start) return null;
const range = document.createRange();
range.setStart(node, start);
range.setEnd(node, end);
const span = document.createElement('span');
span.setAttribute('data-mnote-office-evidence-target', 'true');
try {{
range.surroundContents(span);
return span;
}} catch (_) {{
return null;
}}
}}
if (normalizedStart < 0) return null;
start = sourceOffsets[normalizedStart];
end = sourceOffsets[normalizedStart + compactNeedle.length - 1] + 1;
}}
if (!Number.isFinite(start) || !Number.isFinite(end) || start < 0 || end <= start) return null;
const range = document.createRange();
range.setStart(node, start);
range.setEnd(node, end);
const span = document.createElement('span');
span.setAttribute('data-mnote-office-evidence-target', 'true');
try {{
range.surroundContents(span);
return span;
}} catch (_) {{
return null;
}}
}}
function markEvidenceRangeAcrossTextNodes(needle) {{
if (!viewer) return false;
const compactNeedle = normalizeEvidenceText(needle).replace(/\s+/g, '');
if (!compactNeedle) return false;
const refs = [];
let compactText = '';
const walker = document.createTreeWalker(viewer, NodeFilter.SHOW_TEXT);
let node = walker.nextNode();
while (node) {{
const raw = String(node.textContent || '');
for (let offset = 0; offset < raw.length; offset += 1) {{
const ch = raw[offset];
if (/\s/.test(ch)) continue;
compactText += ch;
refs.push({{ node, offset }});
}}
node = walker.nextNode();
}}
const startIndex = compactText.indexOf(compactNeedle);
if (startIndex < 0) return false;
const endIndex = startIndex + compactNeedle.length - 1;
const startRef = refs[startIndex];
const endRef = refs[endIndex];
if (!startRef || !endRef) return false;
const range = document.createRange();
range.setStart(startRef.node, startRef.offset);
range.setEnd(endRef.node, endRef.offset + 1);
const rect = range.getBoundingClientRect();
if (!rect || rect.width <= 0 || rect.height <= 0) return false;
let marker = viewer.querySelector('[data-mnote-office-evidence-marker="true"]');
if (!(marker instanceof HTMLElement)) {{
marker = document.createElement('div');
marker.className = 'mnote-office-evidence-marker';
marker.setAttribute('data-mnote-office-evidence-marker', 'true');
viewer.append(marker);
}}
const viewerRect = viewer.getBoundingClientRect();
marker.textContent = '';
marker.setAttribute('data-mnote-office-evidence-target-text', normalizeEvidenceText(needle));
marker.setAttribute('data-mnote-office-evidence-marker-mode', 'range');
marker.style.left = Math.max(0, Math.round(rect.left - viewerRect.left + viewer.scrollLeft)).toString() + 'px';
marker.style.top = Math.max(0, Math.round(rect.top - viewerRect.top + viewer.scrollTop)).toString() + 'px';
marker.style.width = Math.max(8, Math.round(rect.width)).toString() + 'px';
marker.style.height = Math.max(8, Math.round(rect.height)).toString() + 'px';
marker.style.maxWidth = 'none';
marker.style.padding = '0';
return markEvidenceTarget(marker);
}}
function pageForEvidenceBlock(sourceMap, block) {{
if (!sourceMap || typeof sourceMap !== 'object' || !block) return null;
const pages = Array.isArray(sourceMap.pages) ? sourceMap.pages : [];
for (const page of pages) {{
const blocks = Array.isArray(page && page.blocks) ? page.blocks : [];
if (blocks.includes(block)) return page;
}}
return null;
}}
function findEvidenceBlockInSourceMap(sourceMap) {{
if (!sourceMap || typeof sourceMap !== 'object' || !evidenceBlockId) return null;
const pages = Array.isArray(sourceMap.pages) ? sourceMap.pages : [];
for (const page of pages) {{
const blocks = Array.isArray(page && page.blocks) ? page.blocks : [];
const block = blocks.find(item => String(item && (item.id || item.blockId || item.block_id) || '') === evidenceBlockId);
if (block) return block;
}}
return null;
}}
async function fetchEvidenceSourceMap() {{
if (!evidenceSourceMapPath || !body.dataset.mnoteRootUri) return null;
const params = new URLSearchParams();
params.set('rootUri', body.dataset.mnoteRootUri);
params.set('path', evidenceSourceMapPath);
const response = await fetch('/api/local-folder/files/open?' + params.toString(), {{
credentials: 'same-origin',
headers: {{ accept: 'application/json, text/plain, */*' }}
}});
if (!response.ok) return null;
return response.json().catch(() => null);
}}
function scrollToEvidenceText(text) {{
const needle = normalizeEvidenceText(text);
if (!needle || !viewer) return false;
viewer.querySelectorAll('[data-mnote-office-evidence-target="true"]').forEach(node => {{
if (node instanceof HTMLElement) {{
if (node.tagName === 'SPAN' && node.childNodes.length === 1 && node.firstChild instanceof Text) {{
node.replaceWith(node.firstChild);
}} else {{
node.removeAttribute('data-mnote-office-evidence-target');
}}
}}
}});
const walker = document.createTreeWalker(viewer, NodeFilter.SHOW_TEXT);
let node = walker.nextNode();
while (node) {{
if (normalizeEvidenceText(node.textContent).includes(needle)) {{
const inlineTarget = wrapEvidenceTextNode(node, needle);
if (inlineTarget) return markEvidenceTarget(inlineTarget);
const target = node.parentElement && node.parentElement.closest('p, div, span, table, section') || node.parentElement;
if (target instanceof HTMLElement) {{
const rect = target.getBoundingClientRect();
if (rect.height > window.innerHeight * 1.8 || normalizeEvidenceText(target.textContent).length > 4000) break;
}}
return markEvidenceTarget(target);
}}
node = walker.nextNode();
}}
if (markEvidenceRangeAcrossTextNodes(needle)) return true;
return false;
}}
function scrollToEvidenceCoordinate(sourceMap, block) {{
if (!viewer || !sourceMap || !block) return false;
const page = pageForEvidenceBlock(sourceMap, block);
const bbox = block.bbox && typeof block.bbox === 'object' ? block.bbox : null;
const pageNumber = Number(page && page.page || evidencePage || 0);
const pageCount = Math.max(1, Number(sourceMap.pageCount || (Array.isArray(sourceMap.pages) ? sourceMap.pages.length : 0)) || 1);
if (!Number.isFinite(pageNumber) || pageNumber <= 0) return false;
const renderedPages = Array.from(viewer.querySelectorAll('section.docx, section.mnote-docx, .pptx-preview-slide-wrapper'))
.filter(node => node instanceof HTMLElement);
const pageElement = renderedPages[pageNumber - 1];
let top = 0;
if (pageElement instanceof HTMLElement) {{
const pageHeight = Math.max(1, pageElement.scrollHeight || pageElement.getBoundingClientRect().height || 1);
const sourcePageHeight = Math.max(1, Number(page && page.height || 0) || pageHeight);
top = pageElement.offsetTop + (bbox ? (Number(bbox.y0) / sourcePageHeight) * pageHeight : pageHeight / 2);
}} else {{
const contentHeight = Math.max(1, viewer.scrollHeight || document.documentElement.scrollHeight || 1);
const estimatedPageHeight = contentHeight / pageCount;
const sourcePageHeight = Math.max(1, Number(page && page.height || 0) || estimatedPageHeight);
top = estimatedPageHeight * (pageNumber - 1) + (bbox ? (Number(bbox.y0) / sourcePageHeight) * estimatedPageHeight : estimatedPageHeight / 2);
}}
let marker = viewer.querySelector('[data-mnote-office-evidence-marker="true"]');
if (!(marker instanceof HTMLElement)) {{
marker = document.createElement('div');
marker.className = 'mnote-office-evidence-marker';
marker.setAttribute('data-mnote-office-evidence-marker', 'true');
viewer.append(marker);
}}
marker.textContent = normalizeEvidenceText(block.text || evidenceBlockId || '');
marker.removeAttribute('data-mnote-office-evidence-target-text');
marker.setAttribute('data-mnote-office-evidence-marker-mode', 'estimated');
marker.style.width = '';
marker.style.height = '';
marker.style.padding = '';
marker.style.top = Math.max(0, Math.round(top)).toString() + 'px';
return markEvidenceTarget(marker);
}}
function scrollToEvidencePageFallback() {{
if (!viewer || !Number.isFinite(evidencePage) || evidencePage <= 0) return false;
const pages = Array.from(viewer.querySelectorAll('section.docx, section.mnote-docx, .pptx-preview-slide-wrapper'));
const target = pages[Math.max(0, Math.min(pages.length - 1, evidencePage - 1))];
return markEvidenceTarget(target);
}}
async function applyEvidenceLocator() {{
if (!viewer || (!evidencePage && !evidenceBlockId && !evidenceBbox)) return;
try {{
const sourceMap = await fetchEvidenceSourceMap();
const block = findEvidenceBlockInSourceMap(sourceMap);
if (block && scrollToEvidenceText(block.text)) return;
if (scrollToEvidenceCoordinate(sourceMap, block)) return;
}} catch (_) {{}}
scrollToEvidencePageFallback();
}}
function updateEvidenceLocator(locator) {{
const next = locator && typeof locator === 'object' ? locator : {{}};
evidencePage = Number(next.page || 0);
evidenceBbox = String(next.bbox || '');
evidenceSourceMapPath = String(next.sourceMapPath || '');
evidenceBlockId = String(next.blockId || '');
body.dataset.evidencePage = evidencePage ? String(evidencePage) : '';
body.dataset.evidenceBbox = evidenceBbox;
body.dataset.evidenceSourceMapPath = evidenceSourceMapPath;
body.dataset.evidenceBlockId = evidenceBlockId;
document.documentElement.removeAttribute('data-mnote-office-evidence-applied');
void applyEvidenceLocator();
}}
window.addEventListener('message', (event) => {{
if (event.origin !== window.location.origin) return;
const data = event.data && typeof event.data === 'object' ? event.data : {{}};
if (data.type === 'mnote:office-evidence-locator') updateEvidenceLocator(data);
}});
async function fetchArrayBuffer() {{
const response = await fetch(fileUrl, {{ credentials: fileUrl.startsWith('/') ? 'same-origin' : 'include' }});
if (!response.ok) throw new Error(fileReadErrorMessage(response.status));
@@ -1321,6 +1637,7 @@ pub async fn office_preview_page(Query(query): Query<OfficePreviewQuery>) -> Res
else if (fileType === 'csv') await renderCsv();
else if (fileType === 'pptx') await renderPptx();
else showMessage(' POC .' + fileType + ' OnlyOffice ');
await applyEvidenceLocator();
setStatus('');
}} catch (error) {{
console.warn('[mnote office preview] render failed', error);
@@ -1342,6 +1659,10 @@ pub async fn office_preview_page(Query(query): Query<OfficePreviewQuery>) -> Res
source_kind = escape_html(&source_kind),
root_uri = escape_html(&root_uri),
document_id = escape_html(&document_id),
target_page = escape_html(&target_page),
target_bbox = escape_html(&target_bbox),
target_source_map_path = escape_html(&target_source_map_path),
target_block_id = escape_html(&target_block_id),
);
let mut response = Html(html).into_response();
stamp_shell_headers(response.headers_mut(), "office-preview");
@@ -3310,6 +3631,12 @@ mod tests {
let resource_runtime = DOCUMENT_RESOURCE_TAB_RUNTIME_JS;
assert!(runtime.contains("document-resource-tab-runtime.js"));
assert!(resource_runtime.contains("resourceTabBadgeKind(input, kind)"));
assert!(resource_runtime.contains("后台任务"));
assert!(resource_runtime.contains("data-mnote-local-ocr-task-tab"));
assert!(resource_runtime.contains("data-mnote-local-ocr-task-clear-completed"));
assert!(resource_runtime.contains("role=\"progressbar\""));
assert!(resource_runtime.contains("localOcrTaskCategory(job)"));
assert!(resource_runtime.contains("localOcrTaskProgress(job)"));
assert!(resource_runtime.contains("mnote.open_editors_snapshot.v1"));
assert!(runtime.contains("getOpenEditorsSnapshot"));
assert!(resource_runtime.contains("bindMainEditorTabStrip"));
@@ -3361,6 +3688,8 @@ mod tests {
assert!(resource_runtime.contains("data-mnote-evidence-bbox"));
assert!(resource_runtime.contains("data-mnote-evidence-text-highlight"));
assert!(runtime.contains("applyDocumentEvidenceLocatorFromUrl"));
assert!(runtime.contains("sourceMapPath: String(url.searchParams.get('sourceMapPath')"));
assert!(runtime.contains("blockId: String(url.searchParams.get('blockId')"));
let sidebar_runtime = SIDEBAR_TREE_RUNTIME_JS;
assert!(sidebar_runtime.contains("openEvidenceSearchResult"));
assert!(sidebar_runtime.contains("data-evidence-locator"));