Improve rezeptwelt.de recipe parsing

[wummel]

This change improves the parser of recipes at rezeptwelt.de:

• detect ingredient groups
• support HTML layout for newer recipes, especially for instruction parsing
• add prep time
• add equipment entries

[jayaddison]

Suggested change

	return "Thermomix Rezeptwelt"
	raise StaticValueException(return_value="Thermomix Rezeptwelt")

I admit this is a slightly unusual pattern that we use; it is used so that the interface of the library can indicate whether values were retrieved from the source HTML or whether they are static/constant values returned by the code.

[jayaddison]

BeautifulSoup (bs4 / self.soup) allows non-empty content filtering by passing a boolean True value, so I think we can simplify these methods slightly:

Suggested change

	def prep_time(self):
	tag = self.soup.find(itemprop="performTime", content=nonempty)
	return get_minutes(tag['content']) if tag else None
	def equipment(self):
	return [tag['content'] for tag in self.soup.find_all("meta", itemprop="tool", content=nonempty)]
	def prep_time(self):
	tag = self.soup.find(itemprop="performTime", content=True)
	return get_minutes(tag['content']) if tag else None
	def equipment(self):
	return [tag['content'] for tag in self.soup.find_all("meta", itemprop="tool", content=True)]

[jayaddison]

Some observations here:

• The retrieval from an itemprop="author" attribute is essentially schema.org metadata retrieval; we have an existing helper method to implement that, so let's re-use them here.
• The information contained in the viewRecipeAuthor element seems more-specific than the schema metadata, which is sometimes generic. So let's prefer viewRecipeAuthor when mentioned.

What this leads me to when adapting the code locally is:

Suggested change

	tag = self.soup.find("div", itemprop="author")
	if tag:
	return normalize_string(tag.get_text())
	tag = self.soup.find("span", {"id": "viewRecipeAuthor"})
	return normalize_string(tag.get_text())
	name_from_schema = self.schema.author()
	name_from_hyperlink = None
	tag = self.soup.find("span", {"id": "viewRecipeAuthor"})
	if tag:
	name_from_hyperlink = tag.get_text()
	return normalize_string(name_from_hyperlink or name_from_schema)

Note: the word von in some of the test data seems redundant, so we can remove that (these changes affect that).