Joe Wicentowski

Here is an XQSuite test that further simplifies the test supplied above. It demonstrates that each time we reindex using `xmldb:reindex#2`, the number of facet hits is incremented. However, when...

@alanpaxton Thank you for this PR and the truly Herculean effort that went into implementing this complex function!

In eXist 5.2.0, no error is raised in the `index-of` examples at the beginning of this issue. But I assume the general problem of "eXist not enforcing return types on...

I have updated the referenced gists with eXist 6.1.0-SNAPSHOT 65d3d5e6968ae8e281cc9926604b055254022529 20220512042037. The updated signature mismatches detected by the script are https://gist.github.com/joewiz/78b1827971a3c1c5dec8fc2d13dec8b2#file-signature-mismatches_results-xml. Some cases are not necessarily bugs - but are...

Leaving out the cases where the function signature mismatches are due to eXist's less verbose signature reporting, here are functions that I think contain real issues: ```xml collection#1 xs:string? xs:string*...

The many CI failures seem to be due to owasp:dependency-check-maven having hit its limit (I think).

I can confirm the report using 5.3.0-SNAPSHOT. A slightly reduced test: ```xquery ()/concat(., "foo") ``` ... returns "foo" instead of empty-sequence(). Similarly, (h/t to Jean-Paul Rehr): ```xquery () => concat("foo")...

BaseX also returns the correct result. Its compiling note for the query: > rewrite mixed path to empty sequence: `()/concat(., "foo")` -> `()`

@amclark42 I can confirm the bug affects eXist 5.3.0. In the meantime, escaping the trailing hyphen will work around the issue: ```xquery matches("aww he--", "[l\-]") ``` ... returns `true()`.

@kermitt2 Thanks so much! I tried making your suggested change to CMakeLists.txt, but the same error appeared, even though I have the XCode Command Line Tools installed (via `xcode-select --install`)....