If you are looking for the decompiler itself, visit https://github.com/jindrapetrik/jpexs-decompiler
NEW : We have got a new blog where we post some interesting SWF internals info.
List of issues#2437 ConstantPool action emitted even for SWF 4
Author: 
LittleCube
LittleCubeDate created:
Type: bug
Visibility: Everybody
Assigned to:
State: upgraded 
For context, this is a *very* minor issue and an extreme nitpick.
> What steps will reproduce the problem?
Open any SWF format version 4 .swf file in JPEXS, edit any actions (even if you remove and
add back the same action), and hit Save.
> What is the expected output? What do you see instead?
The .swf file shouldn't break behavior, but it does. There is something about JPEXS saving
over a SWF 4 .swf that breaks. At first I suspected it was the extraneous addition of a
ConstantPool action (since of course ConstantPool was introduced in SWF 5), but observing
the actual actionBytes there doesn't seem to be one emitted, even though one is
displayed.
> What version of the product are you using? Is it "nightly build"? Which operating system
do you have?
I am using JPEXS 22.0.2 on Windows 10.
> Please provide any additional information below. If the problem is related to a SWF
file, attach it here, otherwise we can't help you.
If you happen to have a build of Flash Player 4, I have attached a SWF that displays
"true" upon clicking a button in both Flash Player 32 and 4. If you open the .swf in JPEXS
and Save the actions of the button (again, without actually changing any of the actions),
JPEXS breaks the button, which causes Flash Player 4 to suddenly begin displaying "false"
while Flash Player 32 continues displaying "true" as it should. I have checked the
actionBytes after editing with JPEXS and they don't seem to be any different, so I can't
imagine what is actually breaking here.
Thank you for your time, and thank you for JPEXS. No rush on this ofc. Again, obviously
this is a very minor nitpick. Let me know if I'm missing something obvious.
true32false4.swf (56 KiB)Upon further inspection, it does appear that JPEXS has indeed inserted a ConstantPool
action after all. It seems this is the issue, unless you see something else wrong as well.
Please try nightly 3036.
It seems that FP4 does not allow (ignores) more than one values in ActionPush.
FFDec was grouping Pushes together for display purposes / editation and that caused the
problem.
Also, I do not emit ConstantPool for SWF version 4 or lower.
State: new→upgraded
Ah, good catch on the Push actions.
I couldn't find 3036, but I tried 3038 and it no longer emits a ConstantPool action, nor
collapses the Push actions! Thank you very much, my SWFs work now.
Out of curiosity, do you happen to know whether FP4 ignores the entire Push action
altogether if there are multiple values, or does it only push the first value?
Yes, I meant 3036 or newer, we remove older nightly builds when we release new.
FP4 pushes only first value on ActionPush.
The second and later value in ActionPush are just additional bytes added to the action.
FP4 does not expect there to be more bytes after first value and just ignores them.
Each Action has length field and FP4 just skips the length bytes using only first value.
I guess they added the feature of grouping Pushes in FP5.
For some reasons, it's not documented in Adobes docs,
but it's good we've found out.
Perfect, I appreciate you specifying that 😋
Thanks again!