analysis-stempel incorrect tokens generation for numbers [LUCENE-10290]

{**}Actual{**}:
I observed unexpected behaviour. Some numbers are affected by stemmer. It causes wrong search results.
For example "2021" -> "20ć".

{**}Expected{**}:
string numbers should not be changed.

{**}Reproduce{**}:

Issue can be reproduced with elasticsearch:

request:

```json
POST _analyze
{
  "tokenizer": "standard",
  "filter": ["polish_stem"],
  "text": "2021"
}
```

response:

```json
{
  "tokens": [
    {
      "token": "20ć",
      "start_offset": 0,
      "end_offset": 4,
      "type": "",
      "position": 0
    }
  ]
}
```

I suspect the newer versions are also affected, but I don't have possibility to verify it.

---
Migrated from [LUCENE-10290](https://issues.apache.org/jira/browse/LUCENE-10290) by Dominik
Environment:
```
**Elasticsearch version** 7.11.2:

**Plugins installed**: [analysis-stempel]

**OS version** CentOS
```

Reproduce the issue with Elasticsearch's _analyze request using the standard tokenizer and the polish_stem filter, starting from the analysis-stempel component. Verify that numeric input such as "2021" remains unchanged and add a regression test covering the expected token output.

Written by the indexing model from the issue text.