Hi Jernej, Sorry, the path should be /instances/instance/name. I do know the specifications about how to resolve XPATH of RFC7950, and I wrote a YANG parser which comply with these specifications. When I compile YANG modules like the example, I get an error.
Yes, if the author write path as "/a:instances/a:instance/a:name", it's correct. But if you are the author, the path "/instances/instance/name" also should point to the schema node defined in module a, not point to the module b where the typedef is used. The specification of RFC7950 force author to write prefixed names in xpath if the xpath is used in groupings and typedefs. I think it may be unreasonable. -----邮件原件----- 发件人: Jernej Tuljak [mailto:[email protected]] 发送时间: 2023年1月12日 23:02 收件人: Fengchong (frank) <[email protected]>; Ladislav Lhotka <[email protected]>; [email protected] 主题: Re: [netmod] 答复: question about name without prefix in XPath expression On 12/01/2023 02:24, Fengchong (frank) wrote: > Lada, > For example: > Module a { > namespace "http://example.com/a";; > typedef instance-ref { > type leafref { > path "instances/instance/name"; > } > } > container instances { > list instance { > key name; > leaf name { > type string; > } > .... > } > } > } > > module b import module a, and use the instance-ref defined by module a > > module b { > namespace "http://example.com/b";; > import a { > prefix a; > } > list interface { > key name; > leaf name { > ... > } > leaf instance { > type a:instance-ref; > } > } > } > > According section 6.4.1, the namespace of names without prefix belong > to the same namespace where the typedefs are used. So, The path > instances/instance/name，the namespace of these 3 names should be > http://example.com/b, but module b have no these definitions. I think it's > unreasonable. > > The meaning of author of module a should be reference the instance name > defined in module a, not module b. > > So, IMO, the namespace of un-prefixed names in absolute location XPATH should > be the same with the namespace of the module where they are defined. The > context node of absolute path should be the module where they are defined. Your path is relative, however - making your example not valid YANG with that specific expression. You probably meant "/instances/instance/name"? The context node is always the leaf/leaf-list for which the type is defined, regardless of whether the path expression is relative or absolute (Section 9.9.2.): The "path" XPath expression is conceptually evaluated in the following context, in addition to the definition in Section 6.4.1: o If the "path" statement is defined within a typedef, the context node is the leaf or leaf-list node in the data tree that references the typedef. o Otherwise, the context node is the node in the data tree for which the "path" statement is defined. A YANG author that writes their YANG with accordance to best practices would write that expression as "/a:instances/a:instance/a:name" (assuming prefix "a"), since RFC 8407 instructs to do so (Section 4.2.): The following guidelines apply to prefix usage of the current (local) module: o The local module prefix SHOULD be used instead of no prefix in all path expressions. Jernej > > -----邮件原件----- > 发件人: Ladislav Lhotka [mailto:[email protected]] > 发送时间: 2023年1月11日 15:42 > 收件人: Fengchong (frank) <[email protected]>; [email protected] > 主题: Re: [netmod] question about name without prefix in XPath > expression > > > > Dne 11. 01. 23 v 4:16 Fengchong (frank) napsal(a): >> Hi all, >> >> In RFC 7950 section 6.4.1: >> >> o Names without a namespace prefix belong to the same namespace >> as >> >> the identifier of the current node. Inside a grouping, that >> >> namespace is affected by where the grouping is used (see >> >> Section 7.13 <https://www.rfc-editor.org/rfc/rfc7950#section-7.13>). >> Inside a typedef, that namespace is affected by >> >> where the typedef is referenced. If a typedef is defined and >> >> referenced within a grouping, the namespace is affected by >> where >> >> the grouping is used (see Section 7.13 >> <https://www.rfc-editor.org/rfc/rfc7950#section-7.13>). >> >> and section 5.4: >> >> Grouping, type, and identity names are resolved in the context >> in >> >> which they are defined, rather than the context in which they >> are >> >> used. Users of groupings, typedefs, and identities are not >> required >> >> to import modules or include submodules to satisfy all >> references >> >> made by the original definition. This behaves like static >> scoping in >> >> a conventional programming language. >> >> I think these two section have some conflicts. If groupings and >> typedefs are resolved in the context where they are defined, why >> names without prefix belongs to the namespace where they are used? > I don't think there is any conflict. The quoted text from 5.4 can actually be > expanded as "Grouping names, type names and identity names are resolved ..." > > I may be missing something because I also don't see the XPath expression part > that you have in the subject. > > Lada > >> The authors who defined the groupings and typedefs don’t know where >> these definitions are used, they only know the place where these >> definitions are defined. >> >> >> _______________________________________________ >> netmod mailing list >> [email protected] >> https://www.ietf.org/mailman/listinfo/netmod > -- > Ladislav Lhotka, CZ.NIC > PGP Key ID: 0xB8F92B08A9F76C67 > > _______________________________________________ > netmod mailing list > [email protected] > https://www.ietf.org/mailman/listinfo/netmod _______________________________________________ netmod mailing list [email protected] https://www.ietf.org/mailman/listinfo/netmod
