Reader docs say foo.bar is a qualified symbol but qualified-symbol? is false

Question

The reader reference on clojure.org says "Symbols containing / or . are said to be 'qualified'." but qualified-symbol? is defined to return true if the argument is a "symbol with a namespace".

Almost every reference to a "qualified symbol" in the docs (on clojure.org) refers to a "namespace-qualified symbol" or a "fully-qualified symbol" (also with a namespace).

The only other type of "qualified symbol" mentioned is on the evaluation page where it refers to a "package-qualified symbol" for a Java class name.

I suggest changing that original sentence to read "Symbols containing / are said to be 'namespace-qualified' or 'fully-qualified'. Symbols containing . are said to be 'package-qualified' if they identify a Java class name." or something similar.

Comments

Recently I ran into an interesting behavior of "symbols that contain periods but not slashes":
```
(defn page.html [])  ;=> fine
(page.html) ;=> syntax error
((resolve 'my.ns/page.html)) ;=> fine
(my.ns/page.html) ;=> fine
```
They can be used to define functions, but require ns qualification to call. Hehehehe.

---

Syntax error (ClassNotFoundException) compiling at (REPL:2:1).
page.html

I was not expecting that :)

Answer 1

"Qualified" means more than one thing depending on context. Java classes are "qualified" by their package name. symbols are "qualified" by their namespace. class names are syntactically represented by a symbol which has overlapping senses - the class package is a dotted prefix that lives in the "name" part of the symbol.

qualified-symbol? is specifically referring to namespace qualification.

My guess is that Rich would say that the original sentence is correct and written as intended.