<< Back to previous view

[CLJ-700] contains? broken for transient collections Created: 01/Jan/11  Updated: 06/Oct/14

Status: Open
Project: Clojure
Component/s: None
Affects Version/s: Release 1.2
Fix Version/s: Release 1.8

Type: Defect Priority: Major
Reporter: Herwig Hochleitner Assignee: Unassigned
Resolution: Unresolved Votes: 13
Labels: None

Attachments: Java Source File 0001-Refactor-of-some-of-the-clojure-.java-code-to-fix-CL.patch     File clj-700-7.diff     File clj-700-8.diff     File clj-700.diff     Text File clj-700-patch4.txt     Text File clj-700-patch6.txt    
Patch: Code and Test
Approval: Vetted

 Description   

Behavior with Clojure 1.6.0:

user=> (contains? (transient {:x "fine"}) :x)
IllegalArgumentException contains? not supported on type: clojure.lang.PersistentArrayMap$TransientArrayMap  clojure.lang.RT.contains (RT.java:724)
;; expected: true

user=> (contains? (transient (hash-map :x "fine")) :x)
IllegalArgumentException contains? not supported on type: clojure.lang.PersistentHashMap$TransientHashMap  clojure.lang.RT.contains (RT.java:724)
;; expected: true

user=> (contains? (transient [1 2 3]) 0)
IllegalArgumentException contains? not supported on type: clojure.lang.PersistentVector$TransientVector  clojure.lang.RT.contains (RT.java:724)
;; expected: true

user=> (contains? (transient #{:x}) :x)
IllegalArgumentException contains? not supported on type: clojure.lang.PersistentHashSet$TransientHashSet  clojure.lang.RT.contains (RT.java:724)
;; expected: true

user=> (:x (transient #{:x}))
nil
;; expected: :x

user=> (get (transient #{:x}) :x)
nil
;; expected: :x

Behavior with latest Clojure master as of Jun 27 2014 (same as Clojure 1.6.0) plus patch clj-700-7.diff. In all cases it matches the expected results shown in comments above:

user=> (contains? (transient {:x "fine"}) :x)
true
user=> (contains? (transient (hash-map :x "fine")) :x)
true
user=> (contains? (transient [1 2 3]) 0)
true
user=> (contains? (transient #{:x}) :x)
true
user=> (:x (transient #{:x}))
:x
user=> (get (transient #{:x}) :x) 
:x

Analysis by Alexander Redington: This is caused by expectations in clojure.lang.RT regarding the type of collections for some methods, e.g. contains() and getFrom(). Checking for contains looks to see if the instance passed in is Associative (a subinterface of PersistentCollection), or IPersistentSet.

This patch refactors several of the Clojure interfaces so that logic abstract from the issue of immutability is pulled out to a general interface (e.g. ISet, IAssociative), but preserves the contract specified (e.g. Associatives only return Associatives when calling assoc()).

With more general interfaces in place the contains() and getFrom() methods were then altered to conditionally use the general interfaces which are agnostic of persistence vs. transience. Includes tests in transients.clj to verify the changes fix this problem.

Questions on this approach from Stuart Halloway to Rich Hickey:

1. this represents working back from the defect to rethinking abstractions (good!). Does it go far enough?

2. what are good names for the interfaces introduced here?

Alex Miller: Should also keep an eye on CLJ-787 as it may have some collisions with this one.

Patch: clj-700-8.diff

One 'trailing whitespace' warning is perfectly normal when applying this patch to latest Clojure master as of Sep 1 2014, as shown below. This is simply because of carriage returns at the end of lines in file Associative.java. I know of no way to avoid such a warning without removing CRs from all Clojure source files (e.g. CLJ-1026):

% git am -s --keep-cr --ignore-whitespace < ~/clj/patches/clj-700-8.diff
Applying: Refactor of some of the clojure .java code to fix CLJ-700.
/Users/andy/clj/latest-clj/clojure/.git/rebase-apply/patch:29: trailing whitespace.
public interface Associative extends IPersistentCollection, IAssociative{
warning: 1 line adds whitespace errors.
Applying: more CLJ-700: refresh to use hasheq


 Comments   
Comment by Herwig Hochleitner [ 01/Jan/11 8:01 PM ]

the same is also true for TransientVectors

{{(contains? (transient [1 2 3]) 0)}}

false

Comment by Herwig Hochleitner [ 01/Jan/11 8:25 PM ]

As expected, TransientSets have the same issue; plus an additional, probably related one.

(:x (transient #{:x}))

nil

(get (transient #{:x}) :x)

nil

Comment by Alexander Redington [ 07/Jan/11 2:07 PM ]

This is caused by expectations in clojure.lang.RT regarding the type of collections for some methods, e.g. contains() and getFrom(). Checking for contains looks to see if the instance passed in is Associative (a subinterface of PersistentCollection), or IPersistentSet.

This patch refactors several of the Clojure interfaces so that logic abstract from the issue of immutability is pulled out to a general interface (e.g. ISet, IAssociative), but preserves the contract specified (e.g. Associatives only return Associatives when calling assoc()).

With more general interfaces in place the contains() and getFrom() methods were then altered to conditionally use the general interfaces which are agnostic of persistence vs. transience. Includes tests in transients.clj to verify the changes fix this problem.

Comment by Stuart Halloway [ 28/Jan/11 10:35 AM ]

Rich: Patch doesn't currently apply, but I would like to get your take on approach here. In particular:

  1. this represents working back from the defect to rethinking abstractions (good!). Does it go far enough?
  2. what are good names for the interfaces introduced here?
Comment by Alexander Redington [ 25/Mar/11 7:44 AM ]

Rebased the patch off the latest pull of master as of 3/25/2011, it should apply cleanly now.

Comment by Stuart Sierra [ 17/Feb/12 2:59 PM ]

Latest patch does not apply as of f5bcf647

Comment by Andy Fingerhut [ 17/Feb/12 5:59 PM ]

clj-700-patch2.txt does patch cleanly to latest Clojure head as of a few mins ago. No changes to patch except in context around changed lines.

Comment by Andy Fingerhut [ 07/Mar/12 3:23 AM ]

Sigh. Git patches applied via 'git am' are fragile beasts indeed. Look at them the wrong way and they fail to apply.

clj-700-patch3.txt applies cleanly to latest master as of Mar 7, 2012, but not if you use this command:

git am -s < clj-700-patch3.txt

I am pretty sure this is because of DOS CR/LF line endings in the file src/jvm/clojure/lang/Associative.java. The patch does apply cleanly if you use this command:

git am --keep-cr -s < clj-700-patch3.txt

Comment by Andy Fingerhut [ 23/Mar/12 6:34 PM ]

This ticket was changed to Incomplete and waiting on Rich when Stuart Halloway asked for feedback on the approach on 28/Jan/2011. Stuart Sierra changed it to not waiting on Rich on 17/Feb/2012 when he noted the patch didn't apply cleanly. Latest patch clj-700-patch3.txt does apply cleanly, but doesn't change the approach used since the time Stuart Halloway's concern was raised. Should it be marked as waiting on Rich again? Something else?

Comment by Stuart Halloway [ 08/Jun/12 12:44 PM ]

Patch 4 incorporates patch 3, and brings it up to date on hashing (i.e. uses hasheq).

Comment by Andy Fingerhut [ 08/Jun/12 12:52 PM ]

Removed clj-700-patch3.txt in favor of Stuart Halloway's improved clj-700-patch4.txt dated June 8, 2012.

Comment by Andy Fingerhut [ 18/Jun/12 3:06 PM ]

clj-700-patch5.txt dated June 18, 2012 is the same as Stuart Halloway's clj-700-patch4.txt, except for context lines that have changed in Clojure master since Stuart's patch was created. clj-700-patch4.txt no longer applies cleanly.

Comment by Andy Fingerhut [ 19/Aug/12 4:47 AM ]

Adding clj-700-patch6.txt, which is identical to Stuart Halloway's clj-700-patch4.txt, except that it applies cleanly to latest master as of Aug 19, 2012. Note that as described above, you must use the --keep-cr option to 'git am' when applying this patch for it to succeed. Removing clj-700-patch5.txt, since it no longer applies cleanly.

Comment by Stuart Sierra [ 24/Aug/12 1:08 PM ]

Patch fails as of commit 1c8eb16a14ce5daefef1df68d2f6b1f143003140

Comment by Andy Fingerhut [ 24/Aug/12 1:53 PM ]

Which patch did you try, and what command did you use? I tried applying clj-700-patch6.txt to the same commit, using the following command, and it applied, albeit with the warning messages shown:

% git am --keep-cr -s < clj-700-patch6.txt
Applying: Refactor of some of the clojure .java code to fix CLJ-700.
/Users/jafinger/clj/latest-clj/clojure/.git/rebase-apply/patch:29: trailing whitespace.
public interface Associative extends IPersistentCollection, IAssociative{
warning: 1 line adds whitespace errors.
Applying: more CLJ-700: refresh to use hasheq

Note the --keep-cr option, which is necessary for this patch to succeed. It is recommended in the "Screening Tickets" section of the JIRA workflow wiki page here: http://dev.clojure.org/display/design/JIRA+workflow

Comment by Andy Fingerhut [ 28/Aug/12 5:48 PM ]

Presumptuously changing Approval from Incomplete back to None, since the latest patch does apply cleanly if the --keep-cr option is used. It was in Screened state recently, but I'm not so presumptuous as to change it to Screened

Comment by Alex Miller [ 19/Aug/13 12:26 PM ]

I think through a series of different hands on this ticket it got knocked way back in the list. Re-marking vetted as it's previously been all the way up through screening. Should also keep an eye on CLJ-787 as it may have some collisions with this one.

Comment by Andy Fingerhut [ 08/Nov/13 10:14 AM ]

clj-700-7.diff is identical to clj-700-patch6.txt, except it applies cleanly to latest master. Only some lines of context in a test file have changed.

When I say "applies cleanly", I mean that there is one warning when using the proper "git am" command from the dev wiki page. This is because one line replaced in Associative.java has a CR/LF at the end of the line, because all lines in that file do.

Comment by Herwig Hochleitner [ 17/Feb/14 9:54 AM ]

Since clojure 1.5, contains? throws an IllegalArgumentException on transients.
In 1.6.0-beta1, transients are no longer marked as alpha.

Does this mean, that we won't be able to distinguish between a nil value and no value on a transient?

Comment by Stuart Halloway [ 27/Jun/14 10:20 AM ]

Request for someone to (1) update patch to apply cleanly, and (2) summarize approach so I don't have to read through the comment history.

Comment by Andy Fingerhut [ 27/Jun/14 11:02 AM ]

The latest patch is clj-700-7.diff dated Nov 8, 2013. I believe it is impossible to create a patch that applies any more cleanly using git for source files that have carriage returns in them, which at least one modified source file does. Here is the command I used on latest Clojure master as of today (Jun 27 2014), which is the same as that of March 25 2014:

% git am -s --keep-cr --ignore-whitespace < ~/clj/patches/clj-700-7.diff 
Applying: Refactor of some of the clojure .java code to fix CLJ-700.
/Users/admin/clj/latest-clj/clojure/.git/rebase-apply/patch:29: trailing whitespace.
public interface Associative extends IPersistentCollection, IAssociative{
warning: 1 line adds whitespace errors.
Applying: more CLJ-700: refresh to use hasheq

If you want a patch that doesn't have the 'trailing whitespace' warning in it, I think someone would have to commit a change that removed the carriage returns from file Associative.java. If you want such a patch, let me know and we can remove all of them from every source file and be done with this annoyance.

Comment by Andy Fingerhut [ 27/Jun/14 11:19 AM ]

Updated description to contain a copy of only those comments that seemed 'interesting'. Most comments have simply been "attached an updated patch that applies cleanly", or "changed the state of this ticket for reason X".

Comment by Alex Miller [ 27/Jun/14 1:19 PM ]

Looks like Andy did as requested, moving back to Screenable.

Comment by Andy Fingerhut [ 29/Aug/14 4:27 PM ]

Patch clj-700-7.diff dated Nov 8 2013 no longer applied cleanly to latest master after some commits were made to Clojure on Aug 29, 2014. It did apply cleanly before that day.

I have not checked how easy or difficult it might be to update this patch.

Comment by Andy Fingerhut [ 01/Sep/14 3:59 AM ]

Patch clj-700-8.diff dated Sep 1 2014 is identical to clj-700-7.diff, except that it applies "cleanly" to latest master, by which I mean it applies as cleanly as I think it is possible to apply for a git patch to a file with carriage return/line feed line endings, as one of the modified files still does.

Generated at Wed Oct 22 23:51:58 CDT 2014 using JIRA 4.4#649-r158309.