Not approved yet: 1. Do you mean to check in the change to lps.properties? That does not seem related to this bug. Also, I have a concern that saying:
compiler.catcherrors=false does not actually work because the property parser parses the properties as strings and considers "false" to be true. At least that is my experience... 2. Have you a test to ensure that non-anonymous classes that are defined in an import library and only used in the import library are actually emitted to the import library rather than the main program? 3. I don't follow the change to ImportCompiler. It seems to handle DTHML and AS3 but what about swf8? And do we need/mean to have `.dhtml.js`? --- Editorial comment: We need to reconcile snippets with binary libraries. I think snippets are too magical and binary libraries require too much work. There should be a happy medium where the designer specifies what goes in a separate library, the compiler binary-compiles that, and the optionally compiles the 'binary' library to each potential platform for runtime loading. That's my dream anyways. On 2010-01-25, at 14:23, Henry Minsky wrote: > I tried making a remodularization of how we pass the compilation environment > around, but it proved to be too much of a mess, > so this is the original change I had proposed, plus code which writes out > separate snippet object files for each runtime that you > compile, to avoid the caching problem that Max was seeing. > > > > Change 20100122-hqm-F by [email protected] on 2010-01-22 13:41:34 EST > in /Users/hqm/openlaszlo/trunk > for http://svn.openlaszlo.org/openlaszlo/trunk > > Summary: fix for bug in swf10 (and other runtimes) snippets > > New Features: > > Bugs Fixed: LPP-8727 bug in dynamic libraries > > Technical Reviewer: ptw > QA Reviewer: max > Doc Reviewer: (pending) > > Documentation: > > Release Notes: > > Overview: > > When the compiler compiles an <import> library, it makes a new > compilation environment for emitting the code. However the ClassModels > that it looks up may have the compilation environment of the main app, as they > may have been created during the updateSchema phase. > > > Details: > > ImportCompiler: Put the name of the runtime into the pathname of each > snippet object file, so that user doesn't accidentally load a swf8 > library into a swf10 runtime or vice-versa > > ClassCompiler: The change is to set the classmodel's CompilationEnvironment > to the > current Compiler's value, so that any anon classes that get made will > inherit that value. > > NodeModel: ensure that class is always emitted (this wasn't the bug in this > case, > but seemed like an oversight) > > ClassModel: make env var public, so it can be set by the ClassCompiler > > Tests: > > run in swf8,10,dhtml > > test/snippets/import-class-and-view.lzx > test/snippets/lpp-8727.lzx > test/snippets/import-view.lzx > test/snippets/import-class-and-view.lzx > test/snippets/import-class-and-view-direct.lzx > test/snippets/import-class-resource.lzx > test/snippets/import-dataload.lzx?lzr=swf10 > > > ant runlzunit > > > > Files: > A test/snippets/class-and-view-library-noproxied.lzx > M test/snippets/class-resource-library.lzx > M test/snippets/import-class-and-view-direct.lzx > M test/snippets/class-and-view-library.lzx > A test/snippets/lpp-8727-library.lzx > M test/snippets/class-and-view-library-proxied.lzx > A test/snippets/lpp-8727.lzx > M test/snippets/import-class-resource.lzx > M WEB-INF/lps/config/lps.properties > M WEB-INF/lps/server/src/org/openlaszlo/compiler/ClassCompiler.java > M WEB-INF/lps/server/src/org/openlaszlo/compiler/ImportCompiler.java > M WEB-INF/lps/server/src/org/openlaszlo/compiler/NodeModel.java > M WEB-INF/lps/server/src/org/openlaszlo/compiler/ClassModel.java > > Changeset: http://svn.openlaszlo.org/openlaszlo/patches/20100122-hqm-F.tar > > _______________________________________________ > Laszlo-reviews mailing list > [email protected] > http://www.openlaszlo.org/mailman/listinfo/laszlo-reviews _______________________________________________ Laszlo-reviews mailing list [email protected] http://www.openlaszlo.org/mailman/listinfo/laszlo-reviews
