@struktoai/mirage-core/dist/core/qdrant/naming.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 { md5Hex } from "../../utils/hash.js";
import { fitIdName, parseIdName } from "../../utils/naming.js";
import { byteLength, NAME_MAX_BYTES, pathSafeName } from "../../utils/sanitize.js";
import { PATH_SAFE } from "../hierarchy/codec.js";
import { valueText } from "../render/json.js";
import { fieldValue } from "./payload.js";
const ENC = new TextEncoder();
/**
 * Render one payload value as a VFS directory segment.
 *
 * A non-string value spells the way the point's `.json` spells it, so
 * TypeScript and Python render one tree. Every level then renders through
 * `PATH_SAFE`, so the scope table decodes a segment back to the exact value it
 * stands for. A basename level first drops the value's URL or path parents,
 * which is lossy, so the lister resolves such a segment against the payload
 * instead of decoding it. A leaf longer than NAME_MAX, which ext4 and APFS
 * refuse, is cut to fit and keeps the md5 of the whole segment as its id,
 * the `<label>__<id>` shape every long name takes, so two leaves the cut
 * would merge stay two directories.
 */
export function groupName(value, basename = false) {
    const name = valueText(value);
    if (!basename)
        return PATH_SAFE.encode(name);
    const withoutFragment = name.split('#', 1)[0] ?? name;
    const withoutQuery = withoutFragment.split('?', 1)[0] ?? withoutFragment;
    const trimmed = withoutQuery.replace(/[\\/]+$/, '');
    const parts = trimmed.replace(/\\/g, '/').split('/');
    const leaf = parts[parts.length - 1] ?? '';
    const segment = PATH_SAFE.encode(leaf === '' ? name : leaf);
    if (byteLength(segment) <= NAME_MAX_BYTES)
        return segment;
    return fitIdName(segment, md5Hex(ENC.encode(segment)));
}
/** Return the stable, human-readable stem for a point's files. */
export function rowStem(row, config) {
    // The point id is synthetic rather than payload data: pointToRow stores it
    // under the configured key verbatim, even when that key contains dots.
    const pointId = String(row[config.idField]);
    const label = fieldValue(row, config.nameField);
    if (label === undefined || label === null)
        return pointId;
    const suffixes = ['.json'];
    if (config.textField !== null && config.textField !== '')
        suffixes.push('.txt');
    if (config.blobField !== null && config.blobField !== '')
        suffixes.push(`.${config.blobExt}`);
    const longestSuffix = suffixes.reduce((a, b) => (byteLength(a) >= byteLength(b) ? a : b));
    const fitted = fitIdName(pathSafeName(valueText(label)), pointId, longestSuffix);
    return fitted.slice(0, -longestSuffix.length);
}
/** Recover the opaque Qdrant point id from a VFS file stem. */
export function pointIdFromStem(stem, config) {
    if (config.nameField === null || config.nameField === '')
        return stem;
    try {
        return parseIdName(stem)[1];
    }
    catch {
        // A point missing the optional naming payload still lists by id.
        return stem;
    }
}
//# sourceMappingURL=naming.js.map