You asked BeautifulSoup for the html5lib parser (`BeautifulSoup(html, 'html5lib')`) without having html5lib installed. Run `pip install html5lib` and the same line works. If you cannot install it, switch to `'html.parser'`, which ships with Python.

## The error

```text
bs4.FeatureNotFound: Couldn't find a tree builder with the features you requested: html5lib
```

## Fix it

1. **Install html5lib**

```
python -m pip install html5lib
```
Expected: `Successfully installed html5lib-1.1`.

2. **Re-run your parse line unchanged**

```
python -c "from bs4 import BeautifulSoup; print(BeautifulSoup('hi', 'html5lib').get_text())"
```
Expected: prints `hi`.

3. **Alternative without new installs**

```python
soup = BeautifulSoup(html, 'html.parser')
```
Expected: parses with the stdlib parser. Behavior differs slightly on broken markup, which is the whole reason people pick html5lib.

## When this applies

- the FeatureNotFound message names `html5lib`
- you passed `'html5lib'` as the parser, or a library did on your behalf

## When it does NOT apply

- FeatureNotFound naming `lxml` means install lxml instead
- `No module named 'html5lib'` from a direct import is the same missing package without bs4 in the middle

## Compatibility

All bs4 4.x with html5lib 1.1. html5lib is the most lenient parser bs4 supports, which is why guides recommend it.

## Why it happens

BeautifulSoup is a tree API over pluggable parsers. Naming `'html5lib'` selects that parser, and if it is not installed bs4 raises FeatureNotFound instead of guessing. The message names the missing feature set, so the fix is mechanical.

## Edge cases

- html5lib is slower than lxml. Pick it for leniency on broken markup, not for speed.
- If html5lib IS installed but the error persists, it is the two-interpreter problem. `python -m pip install html5lib` targets the right one.