Create a template that incorporates general information about the dataset (title, abstract, creators, keywords, etc.)

Since I'm going to need it at work anyway.

Write a quick summary of what we accomplished Thursday, share this with @jhollist et al

https://www.w3schools.com/xml/xsl_intro.asp

Write an R function to do this so we can do fancy things like add leaflet maps. This relates to #15 -- we could use this function to add a map to visualize geographic coverage.

Coverage includes:

• geographicCoverage
• temporalCoverage
• taxonomicCoverage

First step: turn the code in inst/notebook/testing.R into a function.

Next step: add ability to configure e.g. CSS based on YAML

Create a template with information about dataTables. Subcomponents of this include:

• entityName
• entityDescription
• physical
• attributeList

🙏

safe would be FALSE by default. When it is TRUE EML::eml_validate is run before building the website.

Looking at https://eml.ecoinformatics.org/whats-new-in-eml-2-2-0.html#data-paper-support it seems of interest to add these data papers oriented EML elements to this package isn't it ?

Educational = if the EML is missing something that'd be nice to have a warning is written.

When the publish mode is used this kind of weaknesses of the EML are just ignored.

Similar to what pkgdown offers.

Could we create EML from markdown?

This was a suggestion from @mbjones in response to examples of the extensive metadata folks seem to be documenting using markdown and .Rmd. (e.g. see this awesome example from @jmlondon http://jmlondon.github.io/kotzeb0912/kotzebue-bearded-seal-telemetry-2009-2012.html)

Sounds like this would require quite a different approach, the XSLT transform strategy is unlikely to be the best route. Might involve exploiting a prescribed header structure or keywords or .Rmd template, and maybe is just a crazy idea. Just wanted to signpost it here for posterity / further discussion before I forgot.

🥂

E.g. if the authors list is really long.

I think we discussed this and @carlesmila suggested it again.

Sometimes, instead of building an entire website, authors might want just a partial one. This would come in handy if, for example, you wanted to include the rendered EML website into another, more complete, website.

This might be done by plugging the essential xslt templates into a simpler, partial ("snippit") HTML file. Then, the function render_eml() would apply one or the other, depending on the value of an argument (something like partial = TRUE)

• remove "Download" button

• Title to top

• Involved partners to top

• Change the order of the sections of the template so that they match the order in the EML

I don't really know how to approach testing our output, but we should think about it...

@maelle @karawoo @aammd @jhollist and others -- I think emldown is awesome 🎉:tada: :tada: in the way it aims to create a standalone HTML representation of an EML document. I hope the work continues! I've been lurking as you develop the tool, and it seems that work has paused on trying to develop out a more complete XSLT for the complex EML structure.

We have developed an extensive XSLT for transforming EML to HTML, and we reuse that XSLT in multiple places, including our MetacatUI search application, and the Morpho metadata editor, among others. Have you considered re-using this XSLT to generate the base HTML, and then restyle that using CSS and other transforms that make it look how you would like in emldown?

At DataONE, we provide a view service that can run a set of XSLT transforms on a document, and produce an HTML document for further processing. We then add styles to that, and rearrange it with various Javascript DOM manipulations to produce a final product. So, you can see the result of running our XSLT styes on a document by running the view service to produce raw html. The same styles could definitely be run locally in R to produce inital HTML to be further styled and modified in emldown.

Here is an example:

• A raw EML document
  https://arcticdata.io/metacat/d1/mn/v2/object/doi%3A10.18739%2FA2556Q

• The unstyled HTML document from our view service using the metacatui XSLT (see view source)
  https://arcticdata.io/metacat/d1/mn/v2/views/metacatui/doi:10.18739/A2556Q

• The final styled HTML document after we apply CSS and Javascript
  https://arcticdata.io/catalog/#view/doi:10.18739/A2556Q

I see you have a bunch of our XSLT files checked into the emldown repo. We (@mbjones, @csjx, @mobb, @amoeba, and many others) been developing XSLT for EML for a long time, and would be happy to help consult on how to get past the significant bottleneck of getting the initial EML to HTML conversion working. Please let us know how we can help. And thanks for the awesome work!

Try to find a way to take advantage of https://github.com/rich-iannone/DiagrammeR

@carlesmila suggested making it collapsable only when there are more than X authors or depending on an option of the render_eml function.

For the geographic region description.

• [ ]

https://github.com/DataONEorg/rdataone

With a bunch of EML downloaded via that package, stress test emldown.