Age | Commit message (Collapse) | Author |
|
Nearly all the "loose" functions in FXJS become methods on
the CFJXS_Engine.
This is the "missing link" wrt some layering violatons that
have been around forever. We can stop passing &m_ variables
from CJS_ down into FXJS Initialization as a result.
Review-Url: https://codereview.chromium.org/2245863002
|
|
Create a new class to hold these, CFXJS_Engine (could have been
called Runtime, but there are too many "Runtimes" already). In a
subsequent patch, all the FXJS_*() functions that take an isolate
as the first argument can become methods on the engine.
CJS_ must still manage the isolates; this happens outside
the engine.
The IJS_Runtime abstraction moves up to fpdfsdk/javascript; it
remains to allow for either a real JS library or a stubb one to
be linked (for non-js builds).
Review-Url: https://codereview.chromium.org/2241483004
|
|
And call it from FPDF_DestroyLibrary(). Otherwise further attempts to
FPDF_InitLibraryWithConfig() can hit fail an assertion in
IJS_Runtime::Initialize().
BUG=604587
Review-Url: https://codereview.chromium.org/2103443004
|
|
The entire code base is single-threaded, hence the lockers
ought not be required.
Review-Url: https://codereview.chromium.org/2026933002
|
|
R=thestig@chromium.org
BUG=
Review-Url: https://codereview.chromium.org/2019383002
|
|
BUG=515137,564736
TBR=ochang@chromium.org, tsepez@chromium.org
Review URL: https://codereview.chromium.org/1533233002 .
Review URL: https://codereview.chromium.org/1544493003 .
(cherry picked from commit b426e3edde040089b70d1a223c83b90957aa571d)
(cherry picked from commit bd35d484a4027775f19ff93e9143e0b270133d42)
Review URL: https://codereview.chromium.org/1546443002 .
|