Struct KnowledgeBaseIndex 

pub struct KnowledgeBaseIndex {
    pub pages: HashMap<PageId, PageMeta>,
    pub sorted_ids: Vec<PageId>,
    pub index_issues: Vec<ParseIssue>,
    pub duplicate_ids: Vec<DuplicateId>,
    /* private fields */
}

Indexed knowledge base metadata with lazy page loading.

Fields

pages: HashMap<PageId, PageMeta>

Metadata map keyed by canonical page id.

sorted_ids: Vec<PageId>

Page ids in case-insensitive title sort order, computed at open time, maintained by Self::register_page.

index_issues: Vec<ParseIssue>

Non-fatal issues encountered while scanning metadata.

duplicate_ids: Vec<DuplicateId>

Ids claimed by more than one .lepiter file, captured during KnowledgeBase::open before the collision is lost to pages.

Implementations

impl KnowledgeBaseIndex

pub fn register_page(&mut self, meta: PageMeta)

Registers a page in the index, inserting it at the correct sorted position via binary search. If the page already exists, its old sort position is removed first so re-registration never creates duplicates (and handles title changes correctly).

The exact-title index is kept in sync as well: on a title change the page id is moved from its stale title bucket to the new one, and now-empty buckets are dropped.

pub fn load_page(&self, id: &str) -> Result<Page>

Loads and parses a single page by canonical id.

Returns an error if the id is missing from the index or if JSON parsing fails.

pub fn sorted_pages(&self) -> Vec<&PageMeta>

Returns metadata entries in cached title-sorted order.

pub fn filter_page_ids(&self, query: &str) -> Vec<PageId>

Returns page ids filtered by metadata query (title/id/tags), sorted by title.

pub fn filter_page_ids_scored( &self, query: &str, ) -> Vec<(PageId, SearchMatchKind)>

Returns page ids with their match kinds, filtered by metadata query.

pub fn search_hits(&self, query: &str, include_content: bool) -> Vec<SearchHit>

Searches pages by metadata and optionally content, returning hits ranked by relevance (title > tag > content), with ties broken alphabetically by title.

pub fn resolve_page_id_by_title(&self, title: &str) -> TitleResolution

Resolves a page id from title using case-insensitive exact match, then partial match.

pub fn resolve_page_id_by_title_exact(&self, title: &str) -> TitleResolution

Resolves a page id from title using a case-insensitive exact match only, never falling back to a substring like Self::resolve_page_id_by_title. Used for wikilink/page:/title: resolution, where a substring hit would fabricate a graph edge and hide a genuinely broken link.

pub fn classify_link_target(&self, raw: &str) -> LinkTargetKind

Classifies a raw link target for navigation/open behavior.

pub fn build_backlinks(&mut self)

Builds the reverse link index by loading every page, extracting link targets, classifying them, and recording which pages link to which.

Call this once after KnowledgeBase::open when backlink data is needed.

pub fn update_backlinks_for(&mut self, page_id: &str)

Incrementally updates the backlinks index for a single page.

Removes any existing outgoing links from page_id, then re-extracts and classifies its current links. Much cheaper than a full Self::build_backlinks call when only one page changed.

pub fn backlinks_for(&self, id: &str) -> &[PageId]

Returns the page ids that link to the given page, sorted by title.

pub fn build_link_graph(&self) -> LinkGraph

Builds the full directed link graph across all pages.

Each edge represents one internal page-to-page link (deduplicated per source/target pair). Self-links are excluded.

pub fn root(&self) -> &Path

Returns the root path used to build this index.

pub fn attachment_resolver(&self) -> AttachmentResolver

Returns an attachment resolver rooted at this knowledge base.

pub fn scan_all_pages(&self) -> LinkAnalysisResult

Scans all pages in a single pass, collecting broken links, linked pages, link graph edges, missing attachments, and load errors.

pub fn analyze_all(&self) -> LinkAnalysisResult

alias for Self::scan_all_pages.

pub fn analyze_links(&self) -> LinkAnalysisResult

alias for Self::scan_all_pages.

pub fn orphan_ids( &self, linked_pages: &HashSet<PageId>, toc_page_id: Option<&str>, ) -> Vec<PageId>

Returns page ids that are not linked to by any other page.

The toc_page_id (table-of-contents), when present, is excluded from the result since it serves as the root entry point and is not expected to be linked to.

pub fn find_duplicate_titles(&self) -> Vec<DuplicateTitle>

Finds page titles shared by more than one page (case-insensitive).

pub fn find_duplicate_ids(&self) -> Vec<DuplicateId>

Returns page ids claimed by more than one file, captured at open time.

pub fn find_missing_attachments(&self) -> Vec<MissingAttachment>

Finds attachment references whose files are missing from disk; see LinkAnalysisResult::missing_attachments.

Trait Implementations

impl Clone for KnowledgeBaseIndex

fn clone(&self) -> KnowledgeBaseIndex

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for KnowledgeBaseIndex

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.