Files
deerflow-code/frontend-web/src/components/wiki/wikiUtils.ts
T

317 lines
10 KiB
TypeScript
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
import type { KnowledgeNote } from "@/core/knowledge/types";
import type { TocItem, TreeNode } from "./wikiTypes";
export const SOURCE_TYPE_LABELS: Record<string, string> = {
thread: "对话沉淀",
search: "检索沉淀",
tool: "工具沉淀",
file: "文件导入",
manual: "手动创建",
report: "研究报告",
};
export const STATUS_LABELS: Record<string, string> = {
draft: "草稿",
approved: "已批准",
archived: "已归档",
};
const FALLBACK_FOLDER = "未分类";
/**
* Display labels for known English directory segments (the wiki-capture
* categories that historically drove the vault path). Keyed by the lowercased
* segment. User-created folders are arbitrary (usually Chinese) and simply don't
* match, so they pass through unchanged.
*/
const FOLDER_LABELS: Record<string, string> = {
synthesis: "分析综合",
concepts: "概念框架",
concept: "概念框架",
references: "外部来源",
reference: "外部来源",
decision: "决策记录",
journal: "会话纪要",
session: "会话纪要",
};
/** Map a known English directory segment to its Chinese display label (else unchanged). */
export function localizeFolderName(name: string): string {
return FOLDER_LABELS[name.trim().toLowerCase()] ?? name;
}
/**
* Canonical English wiki-capture section headings → Chinese display labels.
* Keyed by the lowercased, symbol-stripped heading text.
*/
const HEADING_LABELS: Record<string, string> = {
context: "背景",
"finding / decision": "结论 / 决策",
"finding/decision": "结论 / 决策",
finding: "结论",
decision: "决策",
reasoning: "推理依据",
implications: "影响 / 后续",
sources: "来源",
related: "相关知识",
merged: "已合并",
};
/**
* Replace the canonical English wiki-capture section headings (`## Context`,
* `## Finding / Decision`, …) with Chinese labels at render time. Operates on
* heading lines only (fenced code blocks are skipped) and matches the heading
* text case-insensitively, so **all historical notes are localized with no data
* migration**. Heading levels are preserved; because `extractToc` and the
* markdown viewer both receive this same transformed text, the slug/anchor ids
* stay consistent.
*/
export function localizeHeadings(markdown: string): string {
let inFence = false;
return markdown
.split("\n")
.map((line) => {
if (/^\s*(```|~~~)/.test(line)) {
inFence = !inFence;
return line;
}
if (inFence) return line;
const match = /^(#{1,4})\s+(.+?)\s*$/.exec(line);
if (!match) return line;
const key = match[2].replace(/[#`*_~]/g, "").trim().toLowerCase();
const zh = HEADING_LABELS[key];
return zh ? `${match[1]} ${zh}` : line;
})
.join("\n");
}
/** Stable slug for a heading: keeps CJK characters, collapses whitespace to dashes. */
export function slugify(text: string): string {
return text
.trim()
.toLowerCase()
.replace(/[#`*_~[\]()]/g, "")
.replace(/\s+/g, "-")
.replace(/[^\w一-龥-]/g, "");
}
/**
* Parse h1–h4 headings out of raw markdown (skipping fenced code blocks) and
* assign each a stable id; duplicate titles get a numeric suffix. The same id
* sequence is mounted onto the rendered headings by MarkdownViewer.
*/
export function extractToc(markdown: string): TocItem[] {
const usedIds = new Map<string, number>();
const items: TocItem[] = [];
let inFence = false;
for (const line of markdown.split("\n")) {
if (/^\s*(```|~~~)/.test(line)) {
inFence = !inFence;
continue;
}
if (inFence) continue;
const match = /^(#{1,4})\s+(.+)$/.exec(line);
if (!match) continue;
const level = match[1].length;
const text = match[2].replace(/[#`*_~]/g, "").trim();
const baseId = slugify(text) || "heading";
const count = usedIds.get(baseId) ?? 0;
usedIds.set(baseId, count + 1);
items.push({ id: count === 0 ? baseId : `${baseId}-${count + 1}`, text, level });
}
return items;
}
/** One entry parsed from the note's trailing `## Related` wikilink list. */
export interface RelatedLink {
target: string;
alias?: string;
}
/**
* Split the backend-injected `## Related` section (a plain list of
* `- [[Title]]` bullets, see knowledge/service.py `_inject_related_notes`)
* out of the markdown body so it can be rendered as a structured module
* instead of a loose bullet list. If the section contains anything other
* than wikilink bullets, the body is returned untouched (no data loss).
*/
export function splitRelatedSection(markdown: string): {
body: string;
related: RelatedLink[];
} {
const lines = markdown.split("\n");
let inFence = false;
let start = -1;
for (let i = 0; i < lines.length; i++) {
if (/^\s*(```|~~~)/.test(lines[i])) {
inFence = !inFence;
continue;
}
if (inFence) continue;
if (/^##\s+(Related|相关文档|相关)\s*$/i.test(lines[i].trim())) {
start = i;
break;
}
}
if (start === -1) return { body: markdown, related: [] };
let end = lines.length;
for (let i = start + 1; i < lines.length; i++) {
if (/^#{1,2}\s+/.test(lines[i])) {
end = i;
break;
}
}
const related: RelatedLink[] = [];
for (const raw of lines.slice(start + 1, end)) {
const line = raw.trim();
if (!line) continue;
const match = /^[-*]\s*\[\[([^\]\n|]+)(?:\|([^\]\n]+))?\]\]\s*$/.exec(line);
if (!match) return { body: markdown, related: [] };
const target = match[1].trim();
if (!related.some((r) => r.target === target)) {
related.push({ target, alias: match[2]?.trim() || undefined });
}
}
if (related.length === 0) return { body: markdown, related: [] };
const body = [...lines.slice(0, start), ...lines.slice(end)]
.join("\n")
.replace(/\n{3,}/g, "\n\n")
.trimEnd();
return { body, related };
}
/**
* Folder path segments for a note. The user-assigned ``folder`` (a slash-joined
* path) is authoritative; unfiled notes fall back to a source-type group so the
* tree is never empty. (The old behavior keyed off ``vault_path`` — the fixed
* ``projects/<project>/<category>`` scaffold — which the editable directory now
* replaces.)
*/
export function noteFolderSegments(note: KnowledgeNote): string[] {
const folder = (note.folder || "").replace(/^\/+|\/+$/g, "").trim();
if (folder) {
const segments = folder.split("/").map((s) => s.trim()).filter(Boolean);
if (segments.length > 0) return segments;
}
return [SOURCE_TYPE_LABELS[note.source_type] ?? FALLBACK_FOLDER];
}
/** Breadcrumb segments for a note (folders only; the title is appended by the caller). */
export function noteBreadcrumbSegments(note: KnowledgeNote): string[] {
return ["知识库管理", ...noteFolderSegments(note).map(localizeFolderName)];
}
function sortTree(nodes: TreeNode[]): void {
nodes.sort((a, b) => {
if (a.type !== b.type) return a.type === "folder" ? -1 : 1;
return a.name.localeCompare(b.name, "zh-CN");
});
for (const node of nodes) {
if (node.children) sortTree(node.children);
}
}
/**
* Build the directory tree from the flat note list. Notes with a vault_path
* become nested folders mirroring that path; others are grouped by source type.
*/
export function buildKnowledgeTree(
notes: KnowledgeNote[],
folderPaths: string[] = [],
): TreeNode[] {
const root: TreeNode[] = [];
const folderMap = new Map<string, TreeNode>();
const ensureFolder = (segments: string[]): TreeNode[] => {
let children = root;
let path = "";
for (const segment of segments) {
path = path ? `${path}/${segment}` : segment;
let node = folderMap.get(path);
if (!node) {
// Display name is localized (known English category → 中文); the id /
// folderPath keep the raw segment so filtering/matching is unaffected.
node = { id: `folder:${path}`, name: localizeFolderName(segment), type: "folder", children: [], folderPath: path };
folderMap.set(path, node);
children.push(node);
}
children = node.children!;
}
return children;
};
// Pre-create managed (possibly empty) directories so a freshly-made folder
// shows up in the tree before any note is filed into it.
for (const path of folderPaths) {
const segments = path.split("/").map((s) => s.trim()).filter(Boolean);
if (segments.length) ensureFolder(segments);
}
for (const note of notes) {
ensureFolder(noteFolderSegments(note)).push({
id: note.id,
name: note.title || "未命名文档",
type: "file",
item: note,
});
}
sortTree(root);
return root;
}
/** Ids of every folder in the tree (used by "expand all" while searching). */
export function collectFolderIds(nodes: TreeNode[]): string[] {
const ids: string[] = [];
const walk = (list: TreeNode[]) => {
for (const node of list) {
if (node.type === "folder") {
ids.push(node.id);
if (node.children) walk(node.children);
}
}
};
walk(nodes);
return ids;
}
/** Folder ids on the path from the root down to the given note, or null if absent. */
export function findAncestorFolderIds(
nodes: TreeNode[],
noteId: string,
): string[] | null {
for (const node of nodes) {
if (node.type === "file" && node.id === noteId) return [];
if (node.type === "folder" && node.children) {
const sub = findAncestorFolderIds(node.children, noteId);
if (sub) return [node.id, ...sub];
}
}
return null;
}
export function formatDateTime(value?: string | null): string {
return value ? value.slice(0, 19).replace("T", " ") : "—";
}
export function formatDate(value?: string | null): string {
return value ? value.slice(0, 10) : "—";
}
/** Trigger a client-side download of the note body as a .md file. */
export function downloadMarkdown(note: KnowledgeNote): void {
const blob = new Blob([note.content_md], { type: "text/markdown;charset=utf-8" });
const url = URL.createObjectURL(blob);
const anchor = document.createElement("a");
anchor.href = url;
anchor.download = `${(note.title || "knowledge").replace(/[\\/:*?"<>|]/g, "_")}.md`;
anchor.click();
URL.revokeObjectURL(url);
}