On Wed, Jun 19, 2019 at 9:27 PM ClearScript Developers < clearscript...@gmail.com> wrote:
> "When you say "local to the evaluation context", what do you mean?" > > Suppose you have an isolate with two contexts, C1 and C2, and a module > whose evaluation produces side effects. Once you've evaluated the module in > C1, there's no way to apply its side effects to C2. You can import it into > C2, but doing so doesn't replicate the side effects. You can use its > exports in C2, but they won't work correctly if they rely on the side > effects. > > "The current API is designed to allow one evaluation per v8::Context" > > But it doesn't do that. It allows one evaluation per isolate. Evaluating a > module in one context breaks it for others. > What I'd recommend is to call CompileModule() again to get another Module object; I believe this should hit the in-memory cache and be very fast, and should let you go back through the Instantiate/Evaluate cycle in C2. > On Wednesday, June 19, 2019 at 5:21:43 PM UTC-4, Adam Klein wrote: >> >> When you say "local to the evaluation context", what do you mean? Modules >> can't have side-effects on the lexical top-level scope of the importing >> module; they can only affect the broader "realm" (in ECMAScript spec >> terms), the v8::Context (in V8 API terms) or the mutable global state (in >> more common terms). The current API is designed to allow one evaluation per >> v8::Context, since part of Instantiation >> <https://cs.chromium.org/chromium/src/v8/include/v8.h?rcl=ab5fb228bcaf387cc72dfd93cdc62c1db4867ebf&l=1302> >> is tying a v8::Module to a v8::Context. So if you instantatiate a module in >> multiple context, multiple evaluations should still work. >> >> On Thu, Jun 13, 2019 at 7:12 AM ClearScript Developers < >> clearsc...@gmail.com> wrote: >> >>> Sorry to revive this thread, but we're encountering the same issue. >>> >>> v8::Script is not a replacement for v8::Module; for one thing, scripts >>> can't import modules :) >>> >>> In any case, V8's API seems deficient here. Modules can be imported >>> specifically for their side effects. There's even unique syntax for that >>> (see >>> https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Statements/import#Import_a_module_for_its_side_effects_only). >>> And since side effects can be local to the evaluation context, it should be >>> possible to evaluate a module in multiple contexts. >>> >>> Opinions? >>> >>> On Wednesday, February 20, 2019 at 2:54:26 PM UTC-5, Adam Klein wrote: >>>> >>>> Modules are designed (in the JavaScript spec) to only have their >>>> top-level code run once. To "re-use" a cached module, the expectation is >>>> that other modules would use imports to access the exports of that module. >>>> In particular, in your code snippet, you'd return the already-evaluated >>>> module from the ResolveCallback passed to the module that wishes to import >>>> from it. >>>> >>>> If what you want to do is compile some code and run it multiple times, >>>> a v8::Script is probably want you want instead. >>>> >>>> On Tue, Feb 12, 2019 at 5:11 PM Jane Chen <jxch...@gmail.com> wrote: >>>> >>>>> Embedding v8 6.7. >>>>> >>>>> I found the following test in test-modules.cc: >>>>> >>>>> CHECK(module->Evaluate(env.local()) >>>>> .ToLocalChecked() >>>>> ->StrictEquals(v8_str("gaga"))); >>>>> CHECK_EQ(Module::kEvaluated, module->GetStatus()); >>>>> >>>>> CHECK(module->Evaluate(env.local()).ToLocalChecked()->IsUndefined()); >>>>> >>>>> and it is consistent with what I'm seeing. Basically, evaluating an >>>>> evaluated module returns undefined. I don't get the rational for this >>>>> behavior. Once a module is evaluated, it is no good for evaluation? Then >>>>> how am I supposed to re-use a cached module? Re-instantiating it doesn't >>>>> seem to help. >>>>> >>>>> -- >>>>> -- >>>>> v8-users mailing list >>>>> v8-u...@googlegroups.com >>>>> http://groups.google.com/group/v8-users >>>>> --- >>>>> You received this message because you are subscribed to the Google >>>>> Groups "v8-users" group. >>>>> To unsubscribe from this group and stop receiving emails from it, send >>>>> an email to v8-u...@googlegroups.com. >>>>> For more options, visit https://groups.google.com/d/optout. >>>>> >>>> -- >>> -- >>> v8-users mailing list >>> v8-u...@googlegroups.com >>> http://groups.google.com/group/v8-users >>> --- >>> You received this message because you are subscribed to the Google >>> Groups "v8-users" group. >>> To unsubscribe from this group and stop receiving emails from it, send >>> an email to v8-u...@googlegroups.com. >>> To view this discussion on the web visit >>> https://groups.google.com/d/msgid/v8-users/a7260cfa-cc60-4870-b11d-b2d00c79eb01%40googlegroups.com >>> <https://groups.google.com/d/msgid/v8-users/a7260cfa-cc60-4870-b11d-b2d00c79eb01%40googlegroups.com?utm_medium=email&utm_source=footer> >>> . >>> For more options, visit https://groups.google.com/d/optout. >>> >> -- > -- > v8-users mailing list > v8-users@googlegroups.com > http://groups.google.com/group/v8-users > --- > You received this message because you are subscribed to the Google Groups > "v8-users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to v8-users+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/v8-users/55686a62-e8dc-439b-9acb-f3edb6ddf60c%40googlegroups.com > <https://groups.google.com/d/msgid/v8-users/55686a62-e8dc-439b-9acb-f3edb6ddf60c%40googlegroups.com?utm_medium=email&utm_source=footer> > . > For more options, visit https://groups.google.com/d/optout. > -- -- v8-users mailing list v8-users@googlegroups.com http://groups.google.com/group/v8-users --- You received this message because you are subscribed to the Google Groups "v8-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to v8-users+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/v8-users/CAEvLGc%2BRxsCsV5PkrzHf04CKSrBAP5-XOcQ7qg6ZzqTz3UBkpA%40mail.gmail.com. For more options, visit https://groups.google.com/d/optout.
