pelican-import from Wordpress produces a lot of noise

[mart-e]

Hello,

I am trying to import a Wordpress XML file into pelican website but it seems to produce a lot of noise so I am wondering if I does something wrong or manually correcting articles is inevitable.

So I tried the following command

$ pelican-import --wpfile -o ./content/ Worpress-export.xml -m markdown --wp-attach 

Which will extract the following article from XML to HTML:

<p><!-- wp:paragraph --></p>
<p>Paragraph content</p>
<p><!-- /wp:paragraph --></p>
<p><!-- wp:image {"align":"center","id":3747,"sizeSlug":"full"} --></p>
<div class="wp-block-image">
<figure class="aligncenter size-full"><img src="https://test.com/test.jpg" alt="" class="wp-image-3747" title="Some title"/><br />
<figcaption><em>Some caption</em></figcaption>
</figure>
</div>
<p><!-- /wp:image --></p>

and call the convert command

pandoc -f html+raw_html --to=markdown-smart --wrap=none -o "content/article.md" "content/article.html"

which produces

`<!-- wp:paragraph -->`{=html}

Paragraph content

`<!-- /wp:paragraph -->`{=html}

`<!-- wp:image {"align":"center","id":3747,"sizeSlug":"full"} -->`{=html}

::: wp-block-image
<figure class="aligncenter size-full">
<img src="https://test.com/test.jpg" title="Some title" class="wp-image-3747" /><br />

<figcaption><em>Some caption</em></figcaption>
</figure>
:::

`<!-- /wp:image -->`{=html}

which, once rendered back to HTML will end to

<article>
...
<p><code>&lt;!-- wp:paragraph --&gt;</code></p>
<p>Paragraph content</p>
<p><code>&lt;!-- /wp:paragraph --&gt;</code></p>
<p><code>&lt;!-- wp:image {"align":"center","id":3747,"sizeSlug":"full"} --&gt;</code></p>
<p>::: wp-block-image</p>
<figure class="aligncenter size-full">
<img src="https://test.com/test.jpg" title="Some title" class="wp-image-3747" /><br />

<figcaption><em>Some caption</em></figcaption>
</figure>
<p>:::</p>
<p><code>&lt;!-- /wp:image --&gt;</code></p>
    </div><!-- /.entry-content -->
</artlcle>

I found I can get rid of the <!-- wp:paragraph --> by using the --strip-raw option but that does not seem right, the docstring says

--strip-raw Strip raw HTML code that can't be converted to markup such as flash embeds or iframes (wordpress import only) (default: False)

which is not really my case. Is it expected that pandoc renders HTML comments?

But even assuming I should use the --strip-raw command, what about the ::: wp-block-image ? I end up now with:

Paragraph content

::: wp-block-image
<figure class="aligncenter size-full">
<img src="https://test.com/test.jpg" title="Some title" class="wp-image-3747" /><br />

<figcaption><em>Some caption</em></figcaption>
</figure>
:::

Looks like a <div class="something">...</div> will be rendered to ::: which does not seem to be a known by Python-Markdown.

[mart-e]

Think this is actually an issue with the usage of pandoc with the wrong converter. Trying a fix at #3114