Analysis across documents loaded into a single analyzer

[chklauser]

After reading the book and the examples in the repository, it's not clear to me, how I would implement an analysis that spans multiple documents.

• I see how the AttrContext::read_doc method lets me fetch a document given its ID, but the only doc ID I have inside a Computable impl is the doc ID I get from context.node_ref().id (the "current" document). I don't see a method to "enumerate" docs (other than trying out all possible IDs with AttrContext::contains_doc)
• I could theoretically try to maintain my own index of documents outside the analyzer. But then I'm not clear on how to Access to external state from Computable impls #12

[Eliah-Lakhin]

I hope that my answer in #12 addressed both questions. Regarding the document enumeration, I want to add that introducing this enumeration function is possible and not too difficult. I'm just not sure if this feature should be a part of the Analyzer's API. My vision is that it is essentially a part of the external state. However, if you have specific use cases where document enumeration is necessary, I will be happy to implement the corresponding function.

[chklauser]

My gut feeling is that you are right. The document alone is probably never going to be "enough". I'd always need additional information that puts the document into context. I guess you could attach information to an Attr on the root node, but that seems hacky.