Multiline codeblocks seen as "superscript" instead of "code" block in confluence

[nado]

Pretty much what the title says, I upload a README file with the following content being rendered incorrectly. Inline code renders fine.

` ` `ps1
git init
echo * > .gitignore
git remote add origin git@server
git fetch
git checkout origin/master -b master
` ` `

[szn]

There is a temporary hack to bypass the problem but I am not happy with the results.

Remember that only bare code blocks work (without language specification in the first line).

$ pip install --upgrade confluence.md

[mxmo0rhuhn]

@szn A CI job that runs on an unchanged README.md file of me fails now (since using the new version):

requests.exceptions.HTTPError: com.atlassian.confluence.api.service.exceptions.BadRequestException: Content body cannot be converted to new editor format

Is there any way that I can extract the HTML for debugging?

[szn]

Sad to hear that @mxmo0rhuhn.

Please install a dev version to see a raw HTML log:

$ pip install --upgrade --index-url https://test.pypi.org/simple/ confluence.md==0.4.7.dev1

And run a command with -v option.

[mxmo0rhuhn]

Hello,
I was able to reproduce the issue with the code block.

Working

a file like this:

# Hello world

resulting in a page like this:

    <p><strong>Automatic content</strong> - Do not edit this page in Confluence.
    Page automatically generated from: <code>test.md</code></p><hr />
    <h1>Hello world</h1>

Not working

a file like this:

    # Hello world

    a file like this should be converted to a confluence page

    ```markdown
    # Hello world
    ```

resulting in a page like this:

    <p><strong>Automatic content</strong> - Do not edit this page in Confluence.
    Page automatically generated from: <code>test_code.md</code></p><hr />
    <h1>Hello world</h1>

<p>a file like this should be converted to a confluence page</p>

<div class="codehilite">
<pre><span></span><code><span class="gh"># Hello world</span>
</code>
</div>

which is subsequently not accepted by the confluence API:

requests.exceptions.HTTPError: com.atlassian.confluence.api.service.exceptions.BadRequestException: Content body cannot be converted to new editor format

At first glance, it looks to me like there is a <pre> tag generated that is not used and not closed.

[szn]

Thanks @mxmo0rhuhn

I am not proud of this fix but it seems to work:

pip install --upgrade --index-url https://test.pypi.org/simple/ confluence.md==0.4.7.dev2

I believe I reached the limits of markdown2 package. For code blocks to render properly I would need to either replace the md2html translator or perform more intense HTML post-processing. Marked as TODO.

[mxmo0rhuhn]

Thank you for the super fast fix! There is a special case that does not work now:

```bash
echo "hello, world!"
```

```python
print("hello, world!")
```

resulting in a page like this:

    <p><strong>Automatic content</strong> - Do not edit this page in Confluence.
    Page automatically generated from: <code>test_code.md</code></p><hr />
    <div class="codehilite">
<code><span class="nb">echo</span><span class="w"> </span><span class="s2">&quot;Hello, World!&quot;</span>
</code>
</div>

<div class="codehilite">
<pre><span></span><code><span class="nb">print</span><span class="p">(</span><span class="s2">&quot;Hello, World!&quot;</span><span class="p">)</span>
</code>
</div>

from a first look it seems like in this setup the regex is not able to match the opening <pre> tag. interestingly, if there is only one code block in the file, the <pre> tag is matched correctly and the upload works as expected.

[szn]

Aaaand:

pip install --upgrade --index-url https://test.pypi.org/simple/ confluence.md==0.4.7.dev3

:)

[mxmo0rhuhn]

Yay it works! All files that I tested with were correctly uploaded to Confluence 👍

Thanks a lot for your continued effort!

[szn]

Thanks for testing!

[szn]

@mxmo0rhuhn 0.4.7.dev3 was published as 0.4.8 (except debug html log). You can:

$ pip install --upgrade confluence.md