09-230 To: J3 From: Malcolm Cohen Subject: More editorial changes Date: 2009 May 06 1. Introduction This paper collects the editorial suggestions made in papers 09-172, 09-177, 09-180 and 09-182. Edits made in those papers that do not appear here have been rejected. Additional edits have been added to fix other glitches more recently spotted. 2. Edits [xiv:p3-2:-1] Change "a missing" to "an absent nonallocatable". [5:33] 1.3.24 codimension Before "corresponding" insert "a set of", Delete refs "(R624, 6.6)". {Clarify wording. The references were wrong: those refs are for image selectors, not for "codimension". No ref is needed because it describes codimension by reference to "coarray" and that has a reference to 2.4.7 which is where this setup is described.} [5:36] 1.3.25 coindexed object Append refs "(R624, 6.6)". {This is where those refs belonged.} [37:12] 2.4.7p4 "s" -> "cosubscripts". {Nothing else in this subclause uses the syntax terms, and neither should this.} [37:14,15] 2.4.7p5 Delete "noncoarray", After "or pointer" insert "component". {A coarray is not permitted to have a coarray subcomponent anyway; and there is no such thing as pointer selection.} [55:29] 4.4.2.4p1 Reinstate "real part" and "imaginary part" in the text (right now they are only in the index). [89:20] 5.3.1 C515 Change "function result that does not have the ALLOCATABLE or POINTER attribute" to "nonallocatable nonpointer function result". {Simplify wording.} [91:22+3] After "same image" insert "as the pointer". [92:18] 5.3.6.3 C528 Change "coarray that does not have the ALLOCATABLE attribute" to "nonallocatable coarray". {Simplify wording.} [119:12] 6.2.1 heading Delete heading. {Spurious remainder of past structure.} [127:18] "specifies" -> "determines", "coarray data" -> "a coindexed object". {It determines what the image index is, it is not a specification. And it's the image index for accessing a coindexed object or maybe for accessing a coarray.} [130:22] After "conformable" delete "(2.4.6)". {It's a defined term and needs no cross-reference.} [155:14] Hyperlink "conform" to "conformable", or better still, Change "the shapes of the variable and shall conform" to "the variable and shall be conformable". [155:23] After "corresponding kind" insert "type parameter". [189:26,27] Delete "END or RETURN ... ;" (one line), Delete "completes execution of a block (8.1.2.2) and" After "results in" insert "the", Turning those two items into "- any statement that results in the implicit deallocation of a coarray;" {Combine two cases into the obvious one case, fix grammar. Note this was already edited by 09-216.} [304:13] Change "assignmentstatement" to "assignment statement". [518:36-37] C.10.1p2, replace entire misleading paragraph with "In the following example, the COINDEX function calculates the image index from the cobounds plus a list of subscripts, and the COSUBSCRIPTS function calculates a set of cosubscripts that specify a given image index." {These are not anything like the intrinsics!} [518:38-519:23] In the example, make the following renamings: name of module "index" -> "coindex_module", "image_index" -> "coindex", "lbound" -> "lcobound" (in both functions), "ubound" -> "ucobound" (ditto), "this_image" -> "cosubscripts", "me" -> "image_number". {The interfaces of the example functions are totally unlike the intrinsics earlier mentioned, blowing the intrinsics away (and the intrinsic INDEX) with misleading function names is a bit much. In c13 we say these functions "are similar" but to me they look more like "not entirely dissimilar".} [535:Annex title] After "(Informative) Syntax rules" Append "and constraints". [535:D.1 title] Append "and constraints". ===END===