perf: have ErrorThrower lazy-lookup the current isolate
ErrorThrower's default constructor is marked as "should rarely if ever
be used" because it's expensive to call.
Unfortunately, nearly every instance of ErrorThrower comes as an argument
in gin_helper's JS-->C++ function marshalling where a thrower is
default-constructed and then populated in gin_helper::GetNextArgument()
with an assignment operator to a temporary ErrorThrower constructed
with the gin::Arguments' isolate.
tldr: most of the time we use the slow constructor first, then throw
that work away unused by overwriting with a fast-constructed one.
This refactor avoids that cost by deferring the expensive work to
`ErrorThrower::isolate()`, where it happens only as a fallback iff
isolate_ hasn't been set.
Co-authored-by: trop[bot] <37223003+trop[bot]@users.noreply.github.com>
Co-authored-by: Charles Kerr <charles@charleskerr.com>
* chore: migrate from base::StringPiece to std::string_view in keyboard_util.cc
* chore: migrate from base::StringPiece to std::string_view in error_thrower.cc
* chore: migrate from base::StringPiece to std::string_view in electron_api_web_contents.cc
* chore: migrate from base::StringPiece to std::string_view in gin_helper/dictionary.h
* chore: migrate from base::StringPiece to std::string_view in electron_api_url_loader.cc
* chore: phase out internal use of base:::StringPiece
`base::StringPiece` is being phased out upstream. Its code has been
removed upstream and it's just a typedef for `std::string_view`.
They haven't removed the typedef yet, so this PR tries to get ahead
of future breakage by migrating "internal" use (i.e. leaving alone the
places where the `base::StringPiece` name is coming from an upstream
method that we override).
Xref: https://bugs.chromium.org/p/chromium/issues/detail?id=691162
Xref: 4294483
Xref: https://docs.google.com/document/d/1d4RnD1uAE2t4iANR0nXy82ASIPGsPuw2mpO6v6T7JKs
* chore: bump chromium in DEPS to 119.0.6036.0
* chore: bump chromium in DEPS to 119.0.6037.0
* chore: bump chromium in DEPS to 119.0.6039.0
* chore: bump chromium in DEPS to 119.0.6041.0
* chore: update chromium patches
* 4765230: Move //content/browser/renderer_host/event_with_latency_info.h to //content/common/input | 4765230
* 4890325: ScopedRunLoopTimeout: add custom timeout callback handler for testing | 4890325
* chore: update all patches
* chore: bump chromium in DEPS to 119.0.6043.0
* 4898682: [api] Add Error.cause to V8 API
4898682
* 4837192: Plumb origin through for drags.
4837192
* Prevent content analysis on web pages that don't accept drag and drop.
4814086
* Make getting displayed notifications work with notification attribution.
4738935
* 4898682: [api] Add Error.cause to V8 API
4898682
* lib,test: do not hardcode Buffer.kMaxLength
https://github.com/nodejs/node/pull/49876
* chore: remove Goma warning from mksnapshot_args
* 4776412: Remove Windows-specific wstring variants of StringPrintf() etc.
4776412
* [dPWA] Prevent WebAppInstallInfo from being included on Android
4886594
---------
Co-authored-by: electron-roller[bot] <84116207+electron-roller[bot]@users.noreply.github.com>
Co-authored-by: VerteDinde <vertedinde@electronjs.org>
Co-authored-by: John Kleinschmidt <jkleinsc@electronjs.org>
* refactor: clean up webFrame implementation to use gin wrappers
The previous implementation of webFrame in the renderer process leaked
sub-frame contexts and global objects across the context boundaries thus
making it possible for apps to either maliciously or accidentally
violate the contextIsolation boundary.
This re-implementation binds all methods in native code directly to
content::RenderFrame instances instead of relying on JS to provide a
"window" with every method request. This is much more consistent with
the rest of the Electron codebase and is substantially safer.
* chore: un-re-order for ease of review
* chore: pass isolate around instead of ErrorThrower
* chore: fix rebase typo
* chore: remove unused variables
* use gin converter in atom_api_menu
* please only put necessary includes in header
Having include in header means they have dependency relationship,
putting arbitrary includes really really really really really makes
refacoring much harder.
* remove some simple uses of callback_converter_deprecated.h
* use gin callback converter in file_dialog code
* use gin in ErrorThrower
* use gin in atom_bundle_mover
* fix mistake in node stream
* deprecate native_mate version of event_emitter_caller
* use gin in node_bindings
* remove usages of native_mate event_emitter_caller.h except for EventEmitter
* fix compilation on Windows
* gin::Arguments behaves differently on GetNext
* just use StringToV8
2019-09-06 14:52:54 +09:00
Renamed from shell/common/error_util.cc (Browse further)
