utilities/snippets_cli_go#6: Fetch item by ID and print

The commit above introduces the matching logic.

What we need now is a function which can print the item. I think it makes sense to do it under this ticket so I'll update the title

This is where we're going to diverge, quite significantly, from the original script.

The old script fetched JSON which had an attribute for each of the sections. We're going to be fetching HTML and parsing that out.

Obviously, this is also going to mean that we need to define the structure that the site should use once migrated.

The old script's output looks like this

$ sbt_cli 175
175: Calculating the difference between subsequent values in SQL (SQL)

-------------
Details
-------------

Language: SQL

-------------
Description
-------------

It's sometimes desirable to caculate the difference between subsequent points.
This might be necessary, for example, if a record's value is drawn from a cumulative total (such as packet counter on a switch) - we're not really interested in the total value so much as the change over time.
The LAG function can be used to look at the previous value, whilst LEAD can be used to look at the next
If the source is something like a packet counter, it might also be desirable to generate a non-negative difference so that a counter reset doesn't lead to large negative figures being reported.

Similar To
-------------

InfluxQL difference
    Flux difference()


-------------
Snippet
-------------

-- Compare to the previous value
SELECT
   time
   _value - LAG (_value) OVER (order by time) AS _value,
FROM foo


-- Compare to the next value
SELECT
   time
   _value - LEAD (_value) OVER (order by time) AS _value,
FROM foo


-- Non-negative difference
SELECT
   time
   CASE
     WHEN _value - LAG (_value) OVER (order by time) < 0 THEN NULL,
     ELSE _value - LAG (_value) OVER (order by time)
   END as _value
FROM foo


HTML Link
----------
https://snippets.bentasker.co.uk/page-2406152213-Calculating-the-difference-between-subsequent-values-SQL.html

On some snippets there's also an example section.

This, not unexpectedly, is quite complicated and will rely entirely on the structure we end up using for snippets.

With the current setup, each section will need to be wrapped in a div:

<div id="snippet">
<pre><code>
foo
</code></pre>
</div>

etc.

Although it's workable, that doesn't lend itself particularly well to the markdown template that I was hoping to use.

One option might be to swap from the html.Parser to html.Tokeniser so that I can look for an attribute with an id (the section heading) and then scoop up everything between it and the next known heading.

I think what we probably need to do next is stand up a local Nikola site so there's something to play around with.

Note: there's a big TODO here - the code currently looks for h4's because I was testing it against a blog post and looking for a heading (because those have IDs).

Actually, thinking about it, do we actually want to make it reliant on that level of granularity?

It might be better to look at how things render if we were to grab a single div (which could be put into the template) and text-ise everything under it. That way minor changes to layout don't lead to us having to update/change this tool.

Re-opening: I forgot to add a print of the full link at the end