url = 'https://lotus.naturalproducts.net/api/search/simple?query=';
sfx = 'LTS0120864';
raw = webread(strcat(url,sfx))
"However, I don't known the names of subfields to access ..."
You can use FIELDNAMES() to get a cell array of the fieldnames:
fnr = fieldnames(raw)
You can loop over those and use dynamic field names to access to the field content:
For example the 3rd field contains this, another structure:
raw.(fnr{3})
"Inside of this structure (html_lotus.naturalProducts.taxonomyReferenceObjects) there are 12 [1x1 struct]. So, I can't use an indexing because all them are [1x1 struct];"
There are indeed 12 fields, so you can easily use the method I just showed with FIELDNAMES() and dynamic field names. Lets try it now:
sub = raw.(fnr{3}).taxonomyReferenceObjects
fns = fieldnames(sub);
for k = 1:numel(fns)
tmp = sub.(fns{k});
display(tmp)
%... do whatever you want with that data...
end
Note that there is no simple, single answer for how to process a nested structure, because how it needs to be processed depends on the structure content and its meaning. MATLAB cannot know this. For example, those 12 fields shown above each consist of a scalar sructure. But those scalar structures have different fieldnames: MATLAB cannot decide for you, if you want to process all of those scalar structures or perhaps you might need to ignore the ones that that are missing a particular field. This is something that only you know, based on your needs. This means that you need to write the code that checks and processes the data according to your needs.
