"Out of order word insertion" in lunr.js due to differences in string comparison between Python and JavaScript?

[abdnh]

I'm using lunr.py to create a search index for use with lunr.js, and I hit an issue that I believe is due to differences in how Python and JavaScript compare strings.

One example to reproduce the issue:

"🔥" < "\uf0ae" 

This evaluates to false in Python but true in JavaScript.

What's the best way to handle this?

[yeraydiazdiaz]

Can you provide a more complete example the interaction between the indices from lunr.py and lunr.js?

[abdnh]

Here's an example to reproduce the error.

1. Create a search index using lunr.py:

import json

from lunr import lunr

documents = [
    {
        "id": "a",
        "body": "🔥",
    },
    {
        "id": "b",
        "body": "\uf0ae",
    },
]

idx = lunr(ref="id", fields=("body",), documents=documents)
with open("index.json", "w", encoding="utf-8") as file:
    json.dump(idx.serialize(), file)

2. Load it into lunr.js (Example in node.js):

const lunr = require("lunr");

async function loadIndex() {
    const serializedIndex = require("./index.json");
    lunr.Index.load(serializedIndex);
}

loadIndex();