Markdown forces an empty line at the end

[hynek]

Hi, I've finally come around to try it!

Looks promising, but the following doesn't work:

```python
>>> import svc_reg
>>> import uuid

>>> reg = svc_reg.Registry()

>>> reg.register_factory(uuid.UUID, uuid.uuid4)
>>> reg.register_value(str, "Hello World")
```

Gives me:

Example at /Users/hynek/FOSS/svc-reg/README.md, line 41, column 1 did not evaluate as expected:
Expected:
   ```
Got nothing

README.md:41: SybilFailure

While the following does (empty last line):

```python
>>> import svc_reg
>>> import uuid

>>> reg = svc_reg.Registry()

>>> reg.register_factory(uuid.UUID, uuid.uuid4)
>>> reg.register_value(str, "Hello World")

```

This is not just about None results.

```python
>>> container = svc_reg.Container(reg)

>>> u = container.get(uuid.UUID)
>>> u
UUID('...')
>>> # Calling get() again returns the SAME UUID instance!
>>> # Good for DB connections, bad for UUIDs.
>>> u is container.get(uuid.UUID)
True
>>> container.get(str)
'Hello World'
```

gives me:

Example at /Users/hynek/FOSS/svc-reg/README.md, line 62, column 1 did not evaluate as expected:
Expected:
    'Hello World'
    ```
Got:
    'Hello World'

README.md:62: SybilFailure

[cjw296]

Hmm, smells like another version of #52.
Parsing is hard 😭

[cjw296]

Can you share your sybil setup from conftest.py?

[hynek]

https://github.com/hynek/svc-reg/blob/main/conftest.py 🤓

[cjw296]

How come you're using the ReST parsers rather than sybil.parsers.myst.PythonCodeBlockParser?

[cjw296]

Added a couple of tests which appear to show this working:
ed027e8

Also chucked in a PR to what I hope will fix things up:
hynek/svcs#2

[hynek]

How come you're using the ReST parsers rather than sybil.parsers.myst.PythonCodeBlockParser?

the answer is, as always: copy and paste :D

[cjw296]

But from where?

[hynek]

I think maybe https://sybil.readthedocs.io/en/latest/patterns.html – I was a bit in a panic mode doing three things at once after fighting my own stuff for a whole day.

[cjw296]

Okay, let's move this over to hynek/svcs#2, if there are bugs, we can re-open this one, or others