QT4 CG Meeting 064 Minutes 2024-02-06

Regrets: MSM.

• [X] Reece Dunn (RD)
• [X] Sasha Firsov (SF) [:06-]
• [X] Christian Grün (CG)
• [X] Joel Kalvesmaki (JK) [:08-]
• [X] Michael Kay (MK)
• [X] Juri Leino (JLO)
• [X] John Lumley (JLY)
• [X] Dimitre Novatchev (DN)
• [X] Wendell Piez (WP)
• [X] Ed Porter (EP)
• [ ] Adam Retter (AR) [:10-]
• [ ] C. M. Sperberg-McQueen (MSM)
• [X] Norm Tovey-Walsh (NW). Scribe. Chair.

Proposal: Accept the agenda.

Accepted.

Proposal: Accept the minutes of the previous meeting.

Accepted.

The next meeting is scheduled for Tuesday, 13 February 2024.

Any regrets for the next meeting? MSM.

• [ ] QT4CG-052-02: NW to consider how to schedule an “editor’s meeting”
• [ ] QT4CG-056-04: MK to write a proposal for adding a select attribute to xsl:text
• [ ] QT4CG-058-02: MK to consider providing more advice about the pitfalls of mixing decimal and double when sorting
• [ ] QT4CG-062-03: JK to revise the fn:hash function along the lines discussed at the meeting
• [ ] QT4CG-063-01: MK to revise #956 especially with respect to the options parameter
• [ ] QT4CG-063-02: JK to consider whether the roman numeral example is appropriate for the spec.
• [ ] QT4CG-063-03: DN to address the editorial remarks from meeting 063 in a new PR.
• [ ] QT4CG-063-04: NW to try to add test review to the editorial meeting.
• [ ] QT4CG-063-05: MK to revise PR #953 to take account of CG’s comments
• [ ] QT4CG-063-06: MK to consider refactoring the declare item type syntax to something like declare record

See PR #993

MK reviews the PR.

• RD: Does this mention errors if you use an invalid Uniecode codepoint value?
• MK: Yes.
• RD: Should we add more detail to the error?
• MK: We didn’t before.
• JLO: If we drop the ability to use “#x20” then we always have to convert to integer in XQuery.
• CG: You can now specify hex integer literals.
• DN: I’m confused it says that it returns something that is a character or more than one characters. We don’t have character strictly defined. The second question is what is the difference between this and just using numeric character references.
• MK: This isn’t related to the change we’re proposing, but the question of multiple characters is that you can use HTML names and some of those return to multiple characters.
• DN: Then it’s not a very accurate name.
• MK: True, but the HTML spec isn’t very clear. We don’t need to change the function because there are edge cases.
• DN: What about numeric character references?
• MK: Two reasons: for clarity and because the argument can be a variable. And XQuery doesn’t define &-entity references.
• RD: For reference, character is defined as a reference to the XML char which is a predefined range of valid Unicode characters.
• CG: Regarding the question, I think we should drop the string variants.

Anyone want to keep the numeric forms “#…” and “#x…”?

No one does.

ACTION QT4CG-064-01: MK to remove “#…” and “#x…” arguments from fn:char

Proposal: Accept this PR.

Accepted.

See PR #984

MK explains that this is designed to make it easier to compute a number of seconds.

Proposal: Accept this PR.

Accepted.

See PR #975

CG explains the PR.

• CG: This is a revision to a change that’s already been accepted. When numbers are specified in the input JSON, how should they be converted to the XDM representation?
  • … This is a few tweaks on the number-parser function in parsing JSON.
  • … Added a default type for the return value.
• CG: Also a few similar changes to the fallback function.

ACTION QT4CG-064-02: CG to update the fallback function to use the new char argument.

ACTION QT4CG-064-03: CG to add examples to show how JSON parsing changes are useful.

The reasons for the changes are along the lines of fn:replace where we want to avoid explicit string conversions.

• MK: Why is this necessary? The input will always be a string.
• CG: In the body of the function, you don’t have to do explicit string conversions.
• MK: It now returns item()? which allows you to return a string if the JSON has a number in it.
• CG: This is already the case because of things like num.
• RD: Or if you want to return null as an empty value.
• MK: Are there constraints on this? What if you returned a map?
• CG: For number parsing, I think that’s okay. But for fallback maybe it needs to be xs:untypedAtomic?

ACTION QT4CG-064-04: CG to change the return type of fallback to xs:untypedAtomic?

• JLY: In the text, the function is called to parse a string. From reading the function name, I thought it would only be involved with numbers.
• CG: That’s right. It converts JSON numbers.

Some discussion of how the input is passed to the string. It’s always a sequence of characters, but it can be things like INF.

• JLY: So I could even do weird things and output a record fo type complex number?
• CG: Basically, yes.
• MK: It’s intended for cases where you want to change out-of-range values into INF for example.

Also: recognizing that there’s no decimal point and returning an integer instead of a double.

JLY observes that the complex number cases are hard to support because there’s no context.

• DN: I think this definintion is good; what JLY asked about untyped atomic. It would survive even if JSON changed what numbers are.

CG continues with changes.

• CG: These options are also added to the json-to-xml function.

Proposal: Accept this PR.

See PR #969

CG reviews the PR.

• CG: This is an attempt to apply some of the ideas that arose out of the discussion.
  • … Add map:empty
  • … Made map:keys-where isntead of having a second argument on map:keys

ACTION QT4CG-064-05: CG to amend the names in the keys-where example to be more inclusive

• JK: Are there places where keys-where is an improvement over a predicate following keys?
• CG: You can also compare values in keys-where
• MK: Another example is, give me the keys where the entry isn’t an empty sequence.

ACTION QT4CG-064-06: Add a keys-where example that uses the value in a more interesting way.

• DN: I think values-where would also be useful here.
• MK: It’s very easily implemented with a filter.
• JLY: There is an example that looks at the value; it’s the birthday example.

Some discussion of whether additonal examples are needed. Leave it up to CG.

Proposal: Accept this PR.

Accepted.

See PR #968

CG introduces the PR.

• CG: This is like the fn:index function except it operates on arrays.
  • … There are repeated use cases where you want something simpler than searching.
  • … The main challenge is how to compare the values.
  • … For the fn:index, there’s only a single value. It’s more complicated here.

CG describes the equality expression chosen.

• CG: This makes the solution closest to what fn:index does.

CG describes the examples.

• JLO: I’d like to see at least one example that uses an explicit collation.
• MK: It’s easy to add an example that uses the HTML collation.

ACTION QT4CG-064-07: CG to add an array:index-of example that uses a collation

• DN: Is xs:anyAtomicType a reasonable search type?
• CG: We made this like fn:index, you can use other functions for more complex searches.
• DN: I think “search” implies some sort of action, like it should be a function. We should use a name that’s more noun-like, perhaps $target.
• MK: I guess target would work.
• JLO: What about term or terms?

ACTION QT4CG-064-08: NW to open an issue to try to resolve $search to $target consistently.

Proposal: Accept this PR.

Accepted.

See PR #943

• MK: This is quite a simple change. I put it in XQuery only but so is the where clause.
  • … It parallels there where clause. It’s described as a filter over the tuples.
  • … Identical to the where clause except that it drops all subsequent tuples.

Proposal: Accept this PR.

Accepted.

See PR #795

• CG: We discussed this a while ago. When I asked if folks thought this would be useful, they said yes.
  • … This implements the sort-with function like you find in other languages.
  • … It uses fn:compare which we introduced recently.
  • … The user must ensure that the comparitor function provided is transitive.
• JLO: I’m using the old sort function a lot and I love the flexibility to just return the sequence. Why only a single value?
• CG: It helps keep the comparitor function tractable.
• JLO: Why is that different in the sort function? I wouldn’t use it.
• CG: The implementations for sort and sort-with are quite different, I’d have to think about it.
  • … I’d like to see an example.
• JLO: It’s not a sequence of comparitors, it’s a sequence of return values. You could return -1, 2, etc. It would be sorted by those values.
• CG: Right…I’ll have to think about that a bit more.
• MK: You just use the first non-zero value.
• DN: Why is this limited only to sequences and items? Such a function would be useful to sort the values of a map or the members of an array.
• CG: The rational is the same as fn:sort. It’s hard to know what to do with an item if it’s an array.
• DN: I’d like to see something more unified.
• CG: We already have the collection issue, but I think it would require a lot of thought.
• MK: A couple of small comments:
  • … The formal equivalence is defined normatively where the bit abou transitivity is non-normative. The equivalence should only apply if the function is transitive and we should say that normatively.
• CG: Suggestion?
• MK: Move the second paragraph of the note up into the normative text.
• MK: In the XML sorting example, the age sort should be numeric.

We’ve run out of time, we’ll return to this next week.