Error in cljr-clean-ns when js literal is present
#443
Comments
|
My personal workaround for this is to use the |
|
I found 2 fixes, but I'm not familiar enough with the rest of the code to know if it's acceptable. 1Similarly to how it's done here, the reader could be augmented with the built-in cljs tag literals. However I'm not sure it's okay to impose a Clojurescript requirement in this namespace. 2Alternatively, we could add a default data reader that would simply drop the tag and return the form. If it's anything else than valid Clojure it could throw an error, but without a default reader we're already throwing anyway. [*ns* file-ns
reader/*data-readers* *data-readers*
reader/*default-data-reader-fn* (fn [_tag form] form) ;; <---- added
mranderson048.toolsreader.v1v1v1.clojure.tools.reader/*alias-map* ns-aliases]Should I open an issue in refactor-nrepl instead? |
|
I like alternative 2) here. You're right that the discussion belongs to the refactor-nrepl so feel free to open a PR over there. Thanks for investigating, @Frozenlock! |
As described in clojure-emacs/clj-refactor.el#443, it was impossible to use clj-clean-ns when a js literal was present: 'No reader function for tag js' One possible solution was to use *default-data-reader-fn* with the function `(fn [_tag value] value)`. However this had the potential of swallowing useful error messages when dealing with other undefined tags. The chosen solution is to provide a dummy reader function specifically for the #js tag (and only for the :cljs dialect). Rather than transforming the provided form into a JS object, it returns the clojure form, which can then be analyzed normally.
As described in clojure-emacs/clj-refactor.el#443, it was impossible to use clj-clean-ns when a js literal was present: 'No reader function for tag js' One possible solution was to use *default-data-reader-fn* with the function `(fn [_tag value] value)`. However this had the potential of swallowing useful error messages when dealing with other undefined tags. The chosen solution is to provide a dummy reader function specifically for the #js tag (and only for the :cljs dialect). Rather than transforming the provided form into a JS object, it returns the clojure form, which can then be analyzed normally.
|
PR submitted clojure-emacs/refactor-nrepl#280. |
As described in clojure-emacs/clj-refactor.el#443, it was impossible to use clj-clean-ns when a js literal was present: 'No reader function for tag js' One possible solution was to use *default-data-reader-fn* with the function `(fn [_tag value] value)`. However this had the potential of swallowing useful error messages when dealing with other undefined tags. The chosen solution is to provide a dummy reader function specifically for the #js tag (and only for the :cljs dialect). Rather than transforming the provided form into a JS object, it returns the clojure form, which can then be analyzed normally.
Expected behavior
Cleaning the namespace with
cljr-clean-ns.Actual behavior
Error when encountering
#js.No error with other tag literal such as
#inst "1985-04-12".Steps to reproduce the problem
Create a clojurescript file with a js literal such as
#js {:a 1}.Try to run
cljr-clean-ns.Environment & Version information
clj-refactor.el version information
clj-refactor 2.4.0, refactor-nrepl 2.4.0
CIDER version information
CIDER 0.21.0 (New York), nREPL 0.6.0
Clojure 1.10.0, Java 1.8.0_212
Leiningen or Boot version
Leiningen 2.9.1 on Java 1.8.0_212 OpenJDK 64-Bit Server VM
Emacs version
Emacs 26.2
Operating system
Ubuntu 18.04.2 LTS
The text was updated successfully, but these errors were encountered: