Files
docmost/apps/server/src/integrations/import/services/file-import-task.service.ts
T
Philip Okugbe 520c07a0bc fix: generic page import hierarchy (#1747)
* fix page hierarchy

* fix
2025-11-29 11:50:02 +00:00

513 lines
16 KiB
TypeScript

import { Injectable, Logger } from '@nestjs/common';
import * as path from 'path';
import { jsonToText } from '../../../collaboration/collaboration.util';
import { InjectKysely } from 'nestjs-kysely';
import { KyselyDB } from '@docmost/db/types/kysely.types';
import {
extractZip,
FileImportSource,
FileTaskStatus,
} from '../utils/file.utils';
import { StorageService } from '../../storage/storage.service';
import * as tmp from 'tmp-promise';
import { pipeline } from 'node:stream/promises';
import { createWriteStream } from 'node:fs';
import { ImportService } from './import.service';
import { promises as fs } from 'fs';
import { generateSlugId } from '../../../common/helpers';
import { v7 } from 'uuid';
import { generateJitteredKeyBetween } from 'fractional-indexing-jittered';
import { FileTask, InsertablePage } from '@docmost/db/types/entity.types';
import { markdownToHtml } from '@docmost/editor-ext';
import { getProsemirrorContent } from '../../../common/helpers/prosemirror/utils';
import { formatImportHtml } from '../utils/import-formatter';
import {
buildAttachmentCandidates,
collectMarkdownAndHtmlFiles,
stripNotionID,
} from '../utils/import.utils';
import { executeTx } from '@docmost/db/utils';
import { BacklinkRepo } from '@docmost/db/repos/backlink/backlink.repo';
import { ImportAttachmentService } from './import-attachment.service';
import { ModuleRef } from '@nestjs/core';
import { PageService } from '../../../core/page/services/page.service';
import { ImportPageNode } from '../dto/file-task-dto';
import { EventEmitter2 } from '@nestjs/event-emitter';
import { EventName } from '../../../common/events/event.contants';
@Injectable()
export class FileImportTaskService {
private readonly logger = new Logger(FileImportTaskService.name);
constructor(
private readonly storageService: StorageService,
private readonly importService: ImportService,
private readonly pageService: PageService,
private readonly backlinkRepo: BacklinkRepo,
@InjectKysely() private readonly db: KyselyDB,
private readonly importAttachmentService: ImportAttachmentService,
private moduleRef: ModuleRef,
private eventEmitter: EventEmitter2,
) {}
async processZIpImport(fileTaskId: string): Promise<void> {
const fileTask = await this.db
.selectFrom('fileTasks')
.selectAll()
.where('id', '=', fileTaskId)
.executeTakeFirst();
if (!fileTask) {
this.logger.log(`Import file task with ID ${fileTaskId} not found`);
return;
}
if (fileTask.status === FileTaskStatus.Failed) {
return;
}
if (fileTask.status === FileTaskStatus.Success) {
this.logger.log('Imported task already processed.');
return;
}
const { path: tmpZipPath, cleanup: cleanupTmpFile } = await tmp.file({
prefix: 'docmost-import',
postfix: '.zip',
discardDescriptor: true,
});
const { path: tmpExtractDir, cleanup: cleanupTmpDir } = await tmp.dir({
prefix: 'docmost-extract-',
unsafeCleanup: true,
});
try {
const fileStream = await this.storageService.readStream(
fileTask.filePath,
);
await pipeline(fileStream, createWriteStream(tmpZipPath));
await extractZip(tmpZipPath, tmpExtractDir);
} catch (err) {
await cleanupTmpFile();
await cleanupTmpDir();
throw err;
}
try {
if (
fileTask.source === FileImportSource.Generic ||
fileTask.source === FileImportSource.Notion
) {
await this.processGenericImport({
extractDir: tmpExtractDir,
fileTask,
});
}
if (fileTask.source === FileImportSource.Confluence) {
let ConfluenceModule: any;
try {
// eslint-disable-next-line @typescript-eslint/no-require-imports
ConfluenceModule = require('./../../../ee/confluence-import/confluence-import.service');
} catch (err) {
this.logger.error(
'Confluence import requested but EE module not bundled in this build',
);
return;
}
const confluenceImportService = this.moduleRef.get(
ConfluenceModule.ConfluenceImportService,
{ strict: false },
);
await confluenceImportService.processConfluenceImport({
extractDir: tmpExtractDir,
fileTask,
});
}
try {
await this.updateTaskStatus(fileTaskId, FileTaskStatus.Success, null);
await cleanupTmpFile();
await cleanupTmpDir();
// delete stored file on success
await this.storageService.delete(fileTask.filePath);
} catch (err) {
this.logger.error(
`Failed to delete import file from storage. Task ID: ${fileTaskId}`,
err,
);
}
} catch (err) {
await cleanupTmpFile();
await cleanupTmpDir();
throw err;
}
}
async processGenericImport(opts: {
extractDir: string;
fileTask: FileTask;
}): Promise<void> {
const { extractDir, fileTask } = opts;
const allFiles = await collectMarkdownAndHtmlFiles(extractDir);
const attachmentCandidates = await buildAttachmentCandidates(extractDir);
const pagesMap = new Map<string, ImportPageNode>();
for (const absPath of allFiles) {
const relPath = path
.relative(extractDir, absPath)
.split(path.sep)
.join('/'); // normalize to forward-slashes
const ext = path.extname(relPath).toLowerCase();
pagesMap.set(relPath, {
id: v7(),
slugId: generateSlugId(),
name: stripNotionID(path.basename(relPath, ext)),
content: '',
parentPageId: null,
fileExtension: ext,
filePath: relPath,
});
}
// Create placeholder pages for folders without corresponding files
const foldersWithContent = new Set<string>();
pagesMap.forEach((page) => {
const segments = page.filePath.split('/');
segments.pop(); // remove filename
// Build up all folder paths and mark them as having content
let currentPath = '';
for (const segment of segments) {
currentPath = currentPath ? `${currentPath}/${segment}` : segment;
foldersWithContent.add(currentPath); // All ancestor folders have content
}
});
// Determine if there's a single root container folder
const rootLevelItems = new Set<string>();
pagesMap.forEach((page) => {
const firstSegment = page.filePath.split('/')[0];
rootLevelItems.add(firstSegment);
});
// If all files are in a single root folder and no files at root level exist
let skipRootFolder: string | null = null;
if (rootLevelItems.size === 1) {
const onlyRootItem = Array.from(rootLevelItems)[0];
// Check if this is a folder (not a file at root)
const hasRootFiles = Array.from(pagesMap.keys()).some(
(filePath) => !filePath.includes('/'),
);
if (!hasRootFiles) {
skipRootFolder = onlyRootItem;
}
}
// For each folder with content, create a placeholder page if no corresponding .md or .html exists
foldersWithContent.forEach((folderPath) => {
if (folderPath.toLowerCase() === skipRootFolder.toLowerCase()) {
return;
}
const mdPath = `${folderPath}.md`;
const htmlPath = `${folderPath}.html`;
if (!pagesMap.has(mdPath) && !pagesMap.has(htmlPath)) {
const folderName = path.basename(folderPath);
pagesMap.set(mdPath, {
id: v7(),
slugId: generateSlugId(),
name: stripNotionID(folderName),
content: '',
parentPageId: null,
fileExtension: '.md',
filePath: mdPath,
});
}
});
// parent/child linking
pagesMap.forEach((page, filePath) => {
const segments = filePath.split('/');
segments.pop();
let parentPage = null;
while (segments.length) {
const tryMd = segments.join('/') + '.md';
const tryHtml = segments.join('/') + '.html';
if (pagesMap.has(tryMd)) {
parentPage = pagesMap.get(tryMd)!;
break;
}
if (pagesMap.has(tryHtml)) {
parentPage = pagesMap.get(tryHtml)!;
break;
}
segments.pop();
}
if (parentPage) page.parentPageId = parentPage.id;
});
// generate position keys
const siblingsMap = new Map<string | null, ImportPageNode[]>();
pagesMap.forEach((page) => {
const group = siblingsMap.get(page.parentPageId) ?? [];
group.push(page);
siblingsMap.set(page.parentPageId, group);
});
// get root pages
const rootSibs = siblingsMap.get(null);
if (rootSibs?.length) {
rootSibs.sort((a, b) => a.name.localeCompare(b.name));
// get first position key from the server
const nextPosition = await this.pageService.nextPagePosition(
fileTask.spaceId,
);
let prevPos: string | null = null;
rootSibs.forEach((page, idx) => {
if (idx === 0) {
page.position = nextPosition;
} else {
page.position = generateJitteredKeyBetween(prevPos, null);
}
prevPos = page.position;
});
}
// non-root buckets (children & deeper levels)
siblingsMap.forEach((sibs, parentId) => {
if (parentId === null) return; // root already done
sibs.sort((a, b) => a.name.localeCompare(b.name));
let prevPos: string | null = null;
for (const page of sibs) {
page.position = generateJitteredKeyBetween(prevPos, null);
prevPos = page.position;
}
});
// internal page links
const filePathToPageMetaMap = new Map<
string,
{ id: string; title: string; slugId: string }
>();
pagesMap.forEach((page) => {
filePathToPageMetaMap.set(page.filePath, {
id: page.id,
title: page.name,
slugId: page.slugId,
});
});
// Group pages by level (topological sort for parent-child relationships)
const pagesByLevel = new Map<number, Array<[string, ImportPageNode]>>();
const pageLevel = new Map<string, number>();
// Calculate levels using BFS
const calculateLevels = () => {
const queue: Array<{ filePath: string; level: number }> = [];
// Start with root pages (no parent)
for (const [filePath, page] of pagesMap.entries()) {
if (!page.parentPageId) {
queue.push({ filePath, level: 0 });
pageLevel.set(filePath, 0);
}
}
// BFS to assign levels
while (queue.length > 0) {
const { filePath, level } = queue.shift()!;
const currentPage = pagesMap.get(filePath)!;
// Find children of current page
for (const [childFilePath, childPage] of pagesMap.entries()) {
if (
childPage.parentPageId === currentPage.id &&
!pageLevel.has(childFilePath)
) {
pageLevel.set(childFilePath, level + 1);
queue.push({ filePath: childFilePath, level: level + 1 });
}
}
}
// Group pages by level
for (const [filePath, page] of pagesMap.entries()) {
const level = pageLevel.get(filePath) || 0;
if (!pagesByLevel.has(level)) {
pagesByLevel.set(level, []);
}
pagesByLevel.get(level)!.push([filePath, page]);
}
};
calculateLevels();
if (pagesMap.size < 1) return;
// Process pages level by level sequentially to respect foreign key constraints
const allBacklinks: any[] = [];
const validPageIds = new Set<string>();
let totalPagesProcessed = 0;
// Sort levels to process in order
const sortedLevels = Array.from(pagesByLevel.keys()).sort((a, b) => a - b);
try {
await executeTx(this.db, async (trx) => {
// Process pages level by level sequentially within the transaction
for (const level of sortedLevels) {
const levelPages = pagesByLevel.get(level)!;
for (const [filePath, page] of levelPages) {
const absPath = path.join(extractDir, filePath);
let content = '';
// Check if file exists (placeholder pages won't have physical files)
try {
await fs.access(absPath);
content = await fs.readFile(absPath, 'utf-8');
if (page.fileExtension.toLowerCase() === '.md') {
content = await markdownToHtml(content);
}
} catch (err: any) {
if (err?.code === 'ENOENT') {
// Use empty content, title will be the folder name
content = '';
} else {
throw err;
}
}
const htmlContent =
await this.importAttachmentService.processAttachments({
html: content,
pageRelativePath: page.filePath,
extractDir,
pageId: page.id,
fileTask,
attachmentCandidates,
});
const { html, backlinks, pageIcon } = await formatImportHtml({
html: htmlContent,
currentFilePath: page.filePath,
filePathToPageMetaMap: filePathToPageMetaMap,
creatorId: fileTask.creatorId,
sourcePageId: page.id,
workspaceId: fileTask.workspaceId,
});
const pmState = getProsemirrorContent(
await this.importService.processHTML(html),
);
const { title, prosemirrorJson } =
this.importService.extractTitleAndRemoveHeading(pmState);
const insertablePage: InsertablePage = {
id: page.id,
slugId: page.slugId,
title: title || page.name,
icon: pageIcon || null,
content: prosemirrorJson,
textContent: jsonToText(prosemirrorJson),
ydoc: await this.importService.createYdoc(prosemirrorJson),
position: page.position!,
spaceId: fileTask.spaceId,
workspaceId: fileTask.workspaceId,
creatorId: fileTask.creatorId,
lastUpdatedById: fileTask.creatorId,
parentPageId: page.parentPageId,
};
await trx.insertInto('pages').values(insertablePage).execute();
// Track valid page IDs and collect backlinks
validPageIds.add(insertablePage.id);
allBacklinks.push(...backlinks);
totalPagesProcessed++;
// Log progress periodically
if (totalPagesProcessed % 50 === 0) {
this.logger.debug(`Processed ${totalPagesProcessed} pages...`);
}
}
}
const filteredBacklinks = allBacklinks.filter(
({ sourcePageId, targetPageId }) =>
validPageIds.has(sourcePageId) && validPageIds.has(targetPageId),
);
// Insert backlinks in batches
if (filteredBacklinks.length > 0) {
const BACKLINK_BATCH_SIZE = 100;
for (
let i = 0;
i < filteredBacklinks.length;
i += BACKLINK_BATCH_SIZE
) {
const backlinkChunk = filteredBacklinks.slice(
i,
Math.min(i + BACKLINK_BATCH_SIZE, filteredBacklinks.length),
);
await this.backlinkRepo.insertBacklink(backlinkChunk, trx);
}
}
if (validPageIds.size > 0) {
this.eventEmitter.emit(EventName.PAGE_CREATED, {
pageIds: Array.from(validPageIds),
});
}
this.logger.log(
`Successfully imported ${totalPagesProcessed} pages with ${filteredBacklinks.length} backlinks`,
);
});
} catch (error) {
this.logger.error('Failed to import files:', error);
throw new Error(`File import failed: ${error?.['message']}`);
}
}
async getFileTask(fileTaskId: string) {
return this.db
.selectFrom('fileTasks')
.selectAll()
.where('id', '=', fileTaskId)
.executeTakeFirst();
}
async updateTaskStatus(
fileTaskId: string,
status: FileTaskStatus,
errorMessage?: string,
) {
try {
await this.db
.updateTable('fileTasks')
.set({ status: status, errorMessage, updatedAt: new Date() })
.where('id', '=', fileTaskId)
.execute();
} catch (err) {
this.logger.error(err);
}
}
}