kamp/anything-llm
1
0
Fork 0
mirror of https://github.com/Mintplex-Labs/anything-llm.git synced 2024-11-11 01:10:11 +01:00
Code Issues Packages Projects Releases Wiki Activity

update normalization function with more clarity

Browse Source
This commit is contained in:
timothycarambat 2024-06-06 15:37:18 -07:00
parent 44b40b0209
commit ad50e886bc
1 changed files with 30 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
server/utils/vectorDbProviders/chroma/index.js
View File

@ -6,6 +6,9 @@ const { v4: uuidv4 } = require("uuid");
const { toChunks, getEmbeddingEngineSelection } = require("../../helpers"); const { toChunks, getEmbeddingEngineSelection } = require("../../helpers");
const { parseAuthHeader } = require("../../http"); const { parseAuthHeader } = require("../../http");
const { sourceIdentifier } = require("../../chats"); const { sourceIdentifier } = require("../../chats");
const COLLECTION_REGEX = new RegExp(
/^(?!\d+\.\d+\.\d+\.\d+$)(?!.*\.\.)(?=^[a-zA-Z0-9][a-zA-Z0-9_-]{1,61}[a-zA-Z0-9]$).{3,63}$/
);
const Chroma = { const Chroma = {
name: "Chroma", name: "Chroma",
@ -18,19 +21,38 @@ const Chroma = {
// We need to enforce these rules by normalizing the collection names // We need to enforce these rules by normalizing the collection names
// before communicating with the Chroma DB. // before communicating with the Chroma DB.
normalize: function (inputString) { normalize: function (inputString) {
let normalized = inputString.replace(/[^a-zA-Z0-9_-]/g, "_"); if (COLLECTION_REGEX.test(inputString)) return inputString;
if (!/^[a-zA-Z0-9]/.test(normalized)) { let normalized = inputString.replace(/[^a-zA-Z0-9_-]/g, "-");
normalized = `anythingllm_${normalized}`;
// Replace consecutive periods with a single period (if any)
normalized = normalized.replace(/\.\.+/g, ".");
// Ensure the name doesn't start with a non-alphanumeric character
if (normalized[0] && !/^[a-zA-Z0-9]$/.test(normalized[0])) {
normalized = "anythingllm-" + normalized.slice(1);
} }
if (!/[a-zA-Z0-9]$/.test(normalized)) {
normalized = `${normalized}_anythingllm`; // Ensure the name doesn't end with a non-alphanumeric character
if (
normalized[normalized.length - 1] &&
!/^[a-zA-Z0-9]$/.test(normalized[normalized.length - 1])
) {
normalized = normalized.slice(0, -1);
} }
// Ensure the length is between 3 and 63 characters
if (normalized.length < 3) { if (normalized.length < 3) {
normalized = `anythingllm_${normalized}`; normalized = `anythingllm-${normalized}`;
} else if (normalized.length > 63) { } else if (normalized.length > 63) {
normalized = normalized.slice(0, 63); // Recheck the norm'd name if sliced since its ending can still be invalid.
normalized = this.normalize(normalized.slice(0, 63));
} }
normalized = normalized.replace(/\.{2,}/g, "_");
// Ensure the name is not an IPv4 address
if (/^\d+\.\d+\.\d+\.\d+$/.test(normalized)) {
normalized = "-" + normalized.slice(1);
}
return normalized; return normalized;
}, },
connect: async function () { connect: async function () {