<< Back to previous view

[CLJS-1318] Fix typo in documentation of `specify` Created: 22/Jun/15  Updated: 22/Jun/15

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: 0.0-3308
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Yehonathan Sharvit Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None

Attachments: Text File patch.txt    
Patch: Code

 Description   

Fix typo in documentation of `specify`



 Comments   
Comment by Yehonathan Sharvit [ 22/Jun/15 6:27 AM ]

here is a patch that fixes the doc of `specify`





[CLJS-1295] Improve doc for instance Created: 01/Jun/15  Updated: 01/Jun/15

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Julien Eluard Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None

Attachments: Text File CLJS-1295.patch    

 Description   

Doc for instance? does not reflect argument names.






[CLJS-1251] Missing semicolons when emitting deftype and defrecord Created: 07/May/15  Updated: 07/May/15

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Defect Priority: Trivial
Reporter: Antonin Hildebrand Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None

Attachments: Text File cljs_1251.patch    

 Description   

Emitted code for deftype and defrecord has to separate the var assignment from body content by a semicolon (do not rely on automatic javascript semicolon insertion).

The bad case:
If body has the first statement wrapped in (...), javascript will treat it as a function call on the constructor definition.






[CLJS-1185] Support preambles when our target is node.js and we are compiler a main file Created: 02/Apr/15  Updated: 02/Apr/15

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Ivan Willig Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None

Attachments: Text File CLJS-1185.patch    




[CLJS-1009] Allow deps.cljs to declare a foreign lib as remote Created: 05/Feb/15  Updated: 21/Feb/15

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Martin Klepsch Assignee: Unassigned
Resolution: Unresolved Votes: 1
Labels: None


 Description   

When using things like the Google Maps JS API [1] the Javascript that is required can't be bundled inside a jar as it depends on the used API key.

To be able to provide externs for those kind of libraries there should be a way to declare them as "remote" in deps.cljs.

[1] https://developers.google.com/maps/documentation/javascript/tutorial






[CLJS-970] assert - generate Error string when compiling Created: 10/Jan/15  Updated: 12/Jan/15

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Defect Priority: Trivial
Reporter: Thomas Heller Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None

Attachments: File cljs-generate-assert-message.diff    
Patch: Code

 Description   

Currently when using assert in CLJS the assert macro will generate the assert error message through CLJS. We can do that in CLJ which reduces the amount of code generated.

(let [bar 1]
  (assert (string? bar)))

Currently generates

if(typeof bar_11332 === 'string'){
} else {
throw (new Error([cljs.core.str("Assert failed: "),cljs.core.str(cljs.core.pr_str.call(null,cljs.core.list(new cljs.core.Symbol(null,"string?","string?",-1129175764,null),new cljs.core.Symbol(null,"bar","bar",254284943,null))))].join('')));
}

But could generate

if(typeof bar_23183 === 'string'){
} else {
throw (new Error("Assert failed: (string? bar)"));
}

Attached patch does that.



 Comments   
Comment by Thomas Heller [ 12/Jan/15 9:16 AM ]

Forgot to account for the fact that

(assert (string? something) "this might not be a string"))

New patch only generates the string of the form when compiling.





[CLJS-924] Better error message for mistaken use of 'def' Created: 24/Dec/14  Updated: 29/Apr/15

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Alex Dowad Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None

Attachments: Text File 0001-Better-error-message-if-def-is-mistakenly-substitute.patch    
Patch: Code

 Description   

ClojureScript shares what is IMHO one of the biggest weaknesses of the current JVM Clojure implementation: those (in)famous stack traces going down into the innards of the compiler if you get your syntax wrong. An easy mistake for noobs is to use 'def' in place of 'defn'; this patch makes that mistake a lot less painful to debug.

Feedback please!






[CLJS-884] with-redefs don't restores previous values when used inside go block Created: 09/Nov/14  Updated: 09/Nov/14

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Defect Priority: Trivial
Reporter: Vladimir Iakovlev Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None
Environment:

clojure 1.6, clojurescript 0.0-2371, core.async 0.1.346.0-17112a-alpha


Attachments: Text File cljs_884.patch    

 Description   

For example I have code and test:

(defn test-fn [] :original)

(deftest ^:async test-with-redefs-async
  (go (is (= (test-fn) :original))
      (with-redefs [test-fn (fn [] :mocked)]
        (is (= (test-fn) :mocked)))
      (is (= (test-fn) :original))
      (done)))

And test failed in last `is`:

FAIL in (test-with-redefs-async) (:)
expected: (= (test-fn) :original)
  actual: (not (= :mocked :original))


 Comments   
Comment by Vladimir Iakovlev [ 09/Nov/14 1:26 PM ]

I attached little patch with fix. With it `with-redefs` forced to use previous value for restoring bindings.

But I don't know how to add test for this issue, because it requires core.async. So I tested with:

(defn test-fn [] :original)

(deftest test-with-redefs-sync
  (is (= (test-fn) :original))
  (with-redefs [test-fn (fn [] :mocked)]
    (is (= (test-fn) :mocked)))
  (is (= (test-fn) :original)))

(deftest ^:async test-with-redefs-async
  (go (is (= (test-fn) :original))
      (with-redefs [test-fn (fn [] :mocked)]
        (is (= (test-fn) :mocked)))
      (is (= (test-fn) :original))
      (done)))




[CLJS-872] Ordered collections printed as if they were unordered at the REPL Created: 13/Oct/14  Updated: 14/Oct/14

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Defect Priority: Trivial
Reporter: Michał Marczyk Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None


 Description   

This is due to cljs.repl's read-then-print processing of string representations of return values that come back from the JS env. As of release 2371, the relevant code fragment lives here:

https://github.com/clojure/clojurescript/blob/r2371/src/clj/cljs/repl.clj#L156

A larger snippet to demonstrate this:

ClojureScript:cljs.user> (array-map 1 2 3 4 5 6 7 8 9 10 11 12 13 14)
{1 2, 3 4, 5 6, 7 8, 9 10, 11 12, 13 14}
ClojureScript:cljs.user> (array-map 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18)
{7 8, 1 2, 15 16, 13 14, 17 18, 3 4, 11 12, 9 10, 5 6}
ClojureScript:cljs.user> (seq (array-map 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18))
([1 2] [3 4] [5 6] [7 8] [9 10] [11 12] [13 14] [15 16] [17 18])
ClojureScript:cljs.user> (hash-map 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18)
{7 8, 1 2, 15 16, 13 14, 17 18, 3 4, 11 12, 9 10, 5 6}

This issue seems to be the most likely cause of the problem described in this StackOverflow question. It would be nice to print ordered collections in the "expected" way to prevent user confusion.



 Comments   
Comment by David Nolen [ 14/Oct/14 5:23 AM ]

How is this handled in Clojure?

Comment by Michał Marczyk [ 14/Oct/14 7:40 AM ]

The built-in REPL simply prints string representations of values to *out* without attempting to read them back in first.

Comment by David Nolen [ 14/Oct/14 7:54 AM ]

Well the result of REPL evaluation is a string unlike Clojure, in order to get a proper print at the ClojureScript REPL it seems necessary to read back the result. I will say it's not clear to me array-map promises anything about order anyway.

Comment by Michał Marczyk [ 14/Oct/14 8:02 AM ]

It does – see http://clojure.org/data_structures, where it says

ArrayMaps
When doing code form manipulation it is often desirable to have a map which maintains key order. An array map is such a map - it is simply implemented as an array of key val key val... As such, it has linear lookup performance, and is only suitable for very small maps. It implements the full map interface. New ArrayMaps can be created with the array-map function. Note that an array map will only maintain sort order when un-'modified'. Subsequent assoc-ing will eventually cause it to 'become' a hash-map.

This snippet has been there for as long as I can remember; and this particular feature comes up in various online discussions from time to time. The docstrings for array-map (the core function) are unambiguous in their promise to construct array maps in both Clojure and ClojureScript.

As for the issue at hand, I do think it's a very minor one (hence the "Trivial" priority), but it is real – the representation printed at the REPL is out of line with the string returned from pr-str on the same value. In fact, one way to fix this would be to use pr-str representations computed on the CLJS side. (There may be some complications here, though, which we'd have to work out.)

Comment by David Nolen [ 14/Oct/14 8:07 AM ]

Thanks for pointing out the API promise. OK, yeah I don't really have a good idea for how do this but a patch that isn't too ugly is welcome





[CLJS-844] Optimize js->clj by switching to transducers Created: 22/Aug/14  Updated: 26/Aug/14

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Darrick Wiebe Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None

Attachments: Text File use-transducers-in-js-to-clj.patch     Text File use-transducers-in-js-to-clj.patch    

 Description   

This simple patch just switches a couple of lines to use transducers. I made this change after finding that using transducers for these types of mappings is significantly faster in (at least in Chrome).

I've checked that this code is being actually tested and both pass under V8 and JavaScriptCore.



 Comments   
Comment by David Nolen [ 23/Aug/14 1:19 PM ]

Did you any benchmarks on other JS engines?

Comment by Darrick Wiebe [ 23/Aug/14 2:14 PM ]

No. My main concern was to check if there was any startup overhead that might offset the basic efficiency improvements for processing small collections. Using into with a transducer turned out to be faster in every scenario, significantly in most.

Comment by David Nolen [ 23/Aug/14 2:19 PM ]

It would be nice to see a jsperf of before and after on this ticket. Thanks.

Comment by Darrick Wiebe [ 23/Aug/14 2:23 PM ]

Is there a existing one that I can work from?

Comment by David Nolen [ 23/Aug/14 2:35 PM ]

There is not. I generally just make a simple project, create an advanced compiled artifact and copy and paste it into jsperf.

Comment by Darrick Wiebe [ 24/Aug/14 7:24 PM ]

Turns out reducing into a transient is considerably better than using a transducer (which was itself a little faster) for this.

http://jsperf.com/js-clj-transducer-test

The code is at:

https://gist.github.com/pangloss/591d77231fda460c2fbe

Let me know if you want me to prepare an updated patch.

Comment by David Nolen [ 25/Aug/14 7:34 AM ]

Thanks for putting this together. Yes please provide an updated patch.

Comment by Darrick Wiebe [ 26/Aug/14 11:19 AM ]

Not sure whether the convention is to comment that I've uploaded a new patch. Regardless, I uploaded it yesterday.





[CLJS-741] seq error message readability is not optimal Created: 02/Jan/14  Updated: 02/Jan/14

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Julien Eluard Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None

Attachments: Text File cljs_741.patch    

 Description   

When calling seq on an invalid type an errr with following message is thrown: :keywordis not ISeqable.

Adding a space between the argument and the message improves the readability.



 Comments   
Comment by David Nolen [ 02/Jan/14 12:58 PM ]

Excellent thanks!





[CLJS-633] Missing Java type hints in closure.clj Created: 22/Oct/13  Updated: 22/Oct/13

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Julien Eluard Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None

Attachments: Text File cljs_ticket_633.patch    

 Description   

A number of reflection warnings are generated from closure.clj. Fixing those probably would not improved performance but it would be cleaner anyway.



 Comments   
Comment by Julien Eluard [ 22/Oct/13 8:03 AM ]

Attached patch.

There are a couple warnings left, similar to:
Reflection warning, cljs/closure.clj:232:1 - reference to field url can't be resolved.
They appear to be due to some unexpected protocol method name? It looks harmless.





[CLJS-555] CLONE - Implement ratios Created: 27/Jul/13  Updated: 02/Dec/14

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Stuart Halloway Assignee: Unassigned
Resolution: Unresolved Votes: 2
Labels: None


 Description   

Clojure.java contains support for Ratio types. It would be nice to have them in Clojurescript as well, but as far as I can tell this would be new development (please comment if I'm wrong). That is, there is no implementation of Ratio types in GClosure so this feature would need to be implemented from the ground up. In additional to the Ratio type, the following corresponding functions would also need implementations:

  • `ratio?`
  • `numerator`
  • `denominator`
  • `rationalize`

Plus the ratio type would need to be rolled into the existing mathematical frameworkings.

Imported from github issue #66



 Comments   
Comment by David Nolen [ 02/Dec/14 6:14 AM ]

Lower priority, unsolvable without numerics overhaul.





[CLJS-507] Persistent Data Structure Benchmark Created: 20/May/13  Updated: 02/Dec/14

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Task Priority: Trivial
Reporter: David Nolen Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None


 Description   

Put together an easy to run series of persistent data structure benchmarks that can be easily run. For example we would like to submit to this Mozilla ticket http://bugzilla.mozilla.org/show_bug.cgi?id=874174






[CLJS-450] (ns) within (do) inconsistent with Clojure behaviour Created: 27/Dec/12  Updated: 21/Feb/15

Status: Open
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Defect Priority: Trivial
Reporter: Stuart Campbell Assignee: Unassigned
Resolution: Unresolved Votes: 0
Labels: None


 Description   

An (ns) within a (do) doesn't quite work as expected at the REPL:

ClojureScript:cljs.user> (do (ns foo) (def x 42))   
nil
ClojureScript:foo> x
nil
ClojureScript:cljs.user> cljs.user/x
42

The Clojure equivalent:

user=> (do (ns foo) (def x 42))
#'foo/x
foo=> x
42


 Comments   
Comment by David Nolen [ 05/Jan/13 2:05 PM ]

Looks like we need to do something similar to what is done in Clojure with top level do - http://github.com/frenchy64/typed-clojure/pull/4





[CLJS-324] cljs.core/format Created: 24/Jun/12  Updated: 11/Mar/15

Status: Reopened
Project: ClojureScript
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Enhancement Priority: Trivial
Reporter: Gary Fredericks Assignee: Unassigned
Resolution: Unresolved Votes: 1
Labels: None

Attachments: Text File 0001-CLJS-324-implement-cljs.core-format-as-wrapper-for-g.patch     File goog.string.format-0.0-1913.tgz    

 Description   

Implement cljs.core/format. Make sure there is nothing in gclosure for this. I suppose we should try to emulate the jvm version as much as that makes sense?



 Comments   
Comment by Michał Marczyk [ 28/Jun/12 11:29 AM ]

GClosure does actually include a string formatter – goog.string.format. Note that that's both a non-ctor function name and the GClosure "namespace" name, so in order to use it one must require goog.string.format on top of gstring and say gstring/format (or perhaps leave out the gstring require spec and use goog.string/format – not tested, but should work).

The patch to introduce format and printf as wrappers for goog.string.format is naturally extremely simple, so here it is. Note that this particular patch must be applied on top of CLJS-328.

I have no idea how goog.string.format compares to the JVM's String/format (basic number formatting seems to work as it should in sprintf-like functions, but other than that I haven't tested it much).

Comment by David Nolen [ 29/Jun/12 10:44 AM ]

The tests fail for me with this patch applied.

Comment by David Nolen [ 29/Jun/12 11:07 AM ]

fixed, http://github.com/clojure/clojurescript/commit/8f518760a3df8b351208e97bb70270856623bb0a

Comment by David Nolen [ 11/Sep/13 5:05 PM ]

Backing this one out, goog.string.format defies advanced optimization and it provides few of the capabilities of Clojure's format - which does a lot because of java.util.Formatter. Apologies for the churn, but this is a simple thing for people to stub in themselves for the little bit of functionality it actually delivers.

Comment by Lars Bohl [ 12/Oct/13 6:33 AM ]

Uploading a slighly modified version lein-cljsbuild's "advanced" demo, to demonstrate that using goog.string.format produces a runtime error with clojurescript 0.0.1913. Run "lein ring server" and navigate to http://localhost:3000/

The code in hello.cljs shows that goog.string.toTitleCase works, but not goog.string.format.

Comment by Julien Eluard [ 12/Oct/13 6:43 AM ]

It looks like you are not requiring correctly format. See a working example here.

Comment by Lars Bohl [ 12/Oct/13 6:58 AM ]

Julent, thanks! It needs another [goog.string.format] after [goog.string :as gstring] before you can use gstring/format. hello.cljs now looks like this, and throws no exceptions: https://www.refheap.com/19693





Generated at Tue Jun 30 14:51:09 CDT 2015 using JIRA 4.4#649-r158309.