@struktoai/mirage-core/dist/core/gdrive/search.js

A Unified Virtual Filesystem For AI Agents

// ========= Copyright 2026 @ Strukto.AI All Rights Reserved. =========
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
// ========= Copyright 2026 @ Strukto.AI All Rights Reserved. =========
import { isFsError } from "../../errors/fs.js";
import { FileType } from "../../types.js";
import { mountedPath } from "../../utils/key_prefix.js";
import { stripSlash } from "../../utils/slash.js";
import { GoogleApiError } from "../google/client.js";
import { rootContext } from "./resolve.js";
import { PARENTS_PER_QUERY, iterTree } from "./tree.js";
/**
 * The files under `under` Drive full-text search names. The folder is walked
 * by id and Drive is asked which files in the walked folders hold the text.
 * Null when a listing or the search fails or times out, nothing is found,
 * Drive skips a corpus, or the scope is not a folder.
 *
 * Drive searches the text it extracted, not the bytes grep reads, so the
 * answer can leave out a file that holds the text: a Docs, Sheets or Slides
 * file (its JSON keys and ids), an office file, PDF or archive, the end of a
 * large file, a word Drive splits differently (`foo.bar`), or a recent write.
 * Mirrors Python's `files_containing`.
 */
export async function filesContaining(accessor, text, under, entry) {
    let root;
    let driveId;
    if (stripSlash(under.vfsPath) === '') {
        ;
        [root, driveId] = await rootContext(accessor);
    }
    else if (entry?.type !== FileType.DIRECTORY) {
        return null;
    }
    else {
        root = entry.extra.file_id;
        const entryDriveId = entry.extra.drive_id;
        driveId = typeof entryDriveId === 'string' ? entryDriveId : null;
    }
    const drive = accessor.drive;
    const found = [];
    try {
        const folders = [root];
        const files = new Map();
        for await (const [rel, item, isDir] of iterTree(accessor, under, root, driveId)) {
            if (isDir)
                folders.push(item.id);
            else
                files.set(item.id, [...(files.get(item.id) ?? []), mountedPath(under, `/${rel}`)]);
        }
        for (let start = 0; start < folders.length; start += PARENTS_PER_QUERY) {
            const parents = folders.slice(start, start + PARENTS_PER_QUERY);
            const answer = await drive.searchFiles(text, parents);
            if (!answer.complete)
                return null;
            for (const id of answer.ids)
                found.push(...(files.get(id) ?? []));
        }
    }
    catch (err) {
        // fetch rejects with a TypeError when the connection fails and a
        // DOMException when its timeout aborts it
        const failed = isFsError(err) ||
            err instanceof GoogleApiError ||
            err instanceof TypeError ||
            err instanceof DOMException;
        if (!failed)
            throw err;
        console.debug(`drive search failed (${String(err)}); reading every file`);
        return null;
    }
    return found.length > 0 ? found : null;
}
//# sourceMappingURL=search.js.map