Remaining tasks for release

[rocky]

• Add doc title links which include possible wikipedia, nltk, spacy, pyenchant or pycountry, and WMA links (see below for format)
• Split code into WMA sections. If there is more in WMA listed than below please add a #TODO item with the name:
  • Text-Level Normalization (TextSentences, TextWords, DeleleteStopWords)
  • Mophological & Lingusitics Normalization (WordStem, DictionaryLookup, DictionaryWordQ, SpellingCorrectionList)
  • Word List Normalization (WordWordFrequency, WordFrequencyData)
  • Word Manipulation (Pluralize, WordData)
  • Language Translation (LanguageIdentity)
  • TextAnalysis (WordCount)
• Add summary text for builtins

Doc title and summary guide

We can use use pymathics.graph as an example to compare against.

If there is a wikipedia entry that goes first. See AdjacencyList for an example.

It may be that only a part of the Wikipedia entry is available. Fill in other text outside of the URL. See DirectedEdge for an example.

If there is no wikipedia mention, it is okay to give some free title. EdgeDelete is an example.

Or you can omit the title altogether. RandomGraph is an Example.

In general we go with the Wikipedia name rather than the WMA for the title. And this includes symbolic parameter names. CompleteKaryTree is an example.

When the only thing we have is a WMA link we add "link" to the title. EdgeList is an example .

Remember that line breaks are significant. \ can be used to wrap a long line.
Start the url name on a new line after <url>. For example:

<url>
:WMA:
https://reference.wolfram.com/language/ref/EdgeList.html</url>

Note that there is no line break at the end before or after </url>.

Please don't get too creative in formatting. There are many other areas in the selection of words to describe what is need may require care. But here it shouldn't require much thought for the formatting aspects.

If the URL is too long, of course, you can split it up in a way that the URL tag understands. Please inspect the URLs in a browser for change. Ideally you would click the link, but if not or before, look at the URL that appears when the link is hovered over.

For Summary text, we start with an active verb with the word in lowercase, e.g. "retrieve" as opposed to "Retrieve". If you look at the section that the summary appears it is nice to use the same verb for similar kinds of things. For example we may "list" builtins that end with "List" (EdgeList, VertexList" but, "find" builtins with "Index" at the end of the name "EdgeIndex", "VertexIndex".

For variables, and options, we do not start with an active verb.

There should be at least one doc example for each function in that is focused on describing what the function does (not how it can be tested). Examples for tests should be added as pytests.

[mmatera]

Thanks for the detailed guidelines!

[rocky]

Thanks for the detailed guidelines!

You are most welcome, and I regret not having done this sooner. Eventually it will make its way into the developer guide. (Feel free to add it if you want.)

[mmatera]

I started with the splitting. I hope to have a PR by the evening.

[mmatera]

If someone wants to work in parallel, the best would be to organize urls and tests in a separated file, and then put all together after the splitting be completed.

[rocky]

I am not going to have much time before the next weekend.

[mmatera]

I guess we are ready with this. Is there something else to do before the release?

[rocky]

I don't think so for pymathics.natlang. And the issues for pymathics.graph I think can be deferred. For a 6.0.1 release.

Any reduction of the warnings such as in cross references in building the LaTeX PDF would be awesome though!