More accurate definition of the Group operator #132
Conversation
hartig
added
the
spec:editorial
Co-authored-by: Ted Thibodeau Jr <tthibodeau@openlinksw.com>
| <p>ListEval((expr<sub>1</sub>, ..., expr<sub>n</sub>), μ) returns a list | ||
| (e<sub>1</sub>, ..., e<sub>n</sub>), where e<sub>i</sub> = expr<sub>i</sub>(μ) or | ||
| error.</p> | ||
| <p>ListEval retains errors resulting from the evaluation of the list elements.</p> | ||
| </div> | ||
| <p>Note that, although the result of a ListEval can be an error, and errors may be used | ||
| <p>Note that, although the result of <a href="#defn_ListEval">ListEval</a> | ||
| may contain errors, and errors may be used | ||
| to group, solutions containing error values are removed at projection time.</p> |
There was a problem hiding this comment.
"projection" seems ambiguous here.
| to group, solutions containing error values are removed at projection time.</p> | |
| to group, solutions containing error values are removed at the end of evaluating the group and any aggregation functions.</p> |
The errors do not appear because the overall grouping expression does not let errors out (it may be used within an expression) which is earlier than "project"
(treat this as an observation more than a necessary change)
There was a problem hiding this comment.
@afs Thanks for pointing this out! In fact, I think there are more issues with this sentence. In particular, it talks about "solutions containing error values" which is inaccurate because solutions (i.e., solution mappings) cannot contain error values. However, I considered fixing/improving this sentence as out of scope of this PR, whose purpose is to improve the definition of the Group operator. The only reason why I touched this sentence in this PR was to add the link to the definition of ListEval for which the PR introduces a linkable anchor ID to be used in the improved definition of the Group operator.
There was a problem hiding this comment.
out of scope of this PR
Please create a new issue for the part that's out of scope of this PR, so we don't lose track of it.
... as discussed in #131.
Preview | Diff