Subject:
Re: [Fwd: [Bug 13658] Changed - Rhino: null pointer exception on class with duplicate field/method]
Date:
Mon, 13 Sep 1999 20:57:32 -0400
From:
"Kurt Westerfeld" <kurt@westerfeld.com>
To:
"Norris Boyd" <norris@netscape.com>
I do have a patch for this, but it is intermixed with some other changes
that I have implemented for the get/set on Java instances (per my LC3
proposal). The bug requires changes that are a little involved actually;
basically it seems that when getting the default value for a "field and
methods" (which combines the same-named entities), the prototype of the
parent scope is deref-ed, and the parent scope is null. Hence, the scope
must be passed into the the cloned field and method values.
Also, the NativeJavaClass implementation passed "false" for isStatic on the
constructor of the FieldAndMethods Hashtable, which results in classes
having instance methods. Bad. I haven't filed a bug on that yet.
Additionally, I fixed a couple other NullPointerException nigglies thrown in
when exceptions are propagated in the same area. Finally, when getting the
default value for the field, it is helpful to convert a Scriptable to string
when that is requested (as when typing in the console).
I am attaching the changed files. The LC3++ code can be removed if you
want, which I can do for you but it will take a little longer. What is your
preference?
-----Original Message-----
From: Norris Boyd <norris@netscape.com>
To: Kurt Westerfeld <kurt@westerfeld.com>
Date: Monday, September 13, 1999 4:54 PM
Subject: [Fwd: [Bug 13658] Changed - Rhino: null pointer exception on class
with duplicate field/method]
>Kurt,
>
>Is this the bug that your patch fixes?
>
>Thanks,
>Norris
>
and propertyIsEnumerable) for JS1.5.
- Optimize obj_propertyIsEnumerable to avoid extra lookup code bloat, requiring
fix to js_GetAttributes (unset out param on successful early retunr) that it
exposed.
- Use more righteous else-if style in shaver's jsarray.c change.
- Add JS1.5 getter/setter support in all its glory:
* getter function SN() {return ++x} at top-level or as a closure binds an SN
property getter than returns the incremented value of x. Likewise for
setter function SN(y) {return y = x}.
* getters and setters may be defined in an object literal:
o = {p getter:function() {return ++this.x},
p setter:function(y){return this.x = y},
x:42};
* getter= and setter= operators (compound tokens) may be used to bind getter
and setter properties dynamically:
o = new Object;
o.p getter= function() {return ++this.x};
o.p setter= function(y){return this.x = y};
o.x = 42;
Waldemar is concerned that this form will collide semantically with JS2, so
I am not committing to keeping it in JS1.5. I'd like to check my code in
ASAP so shaver can use it, and I'd also like to see this form get used (or
not) during Mozilla betas. Caveat emptor, and if you find this "dynamic"
or "imperative" form necessary and hard to substitute, please let me know.
If this proves important to users, then I think JS1.5 should keep it.
- Cleaned up property flags (in a binary-incompatible fashion -- who cares?) by
eliminating JSPROP_ASSIGNHACK and JSPROP_TINYIDHACK.
- Added JS_DONT_PRETTY_PRINT flag to be ORed with the indent argument to the
several JS_Decompile*() API calls. This avoids any newlines or identation in
the decompiled string.
- Improved and extended (for getter/setter non-reservation) scanner lookahead
by using a circular (power-of-2 sized) token buffer.
- Fix ECMA Edition 3 deviation where function f(){function g(){}} bound f.g by
mistake (it should arrange to make a closure named g in activations of f, but
it should not bind a property of function f).
Turned off 'super' keyword - was letting through some cut'n'pasted java
code quietly and blowing big chunks out of the codegen/interpreter later.
Anybody know why 'super' had an interesting value here? - there was no
support for it on any path that I could see.
- Merged the typelib types for array and array_with_length.
- Added typelib types for string_with_size and wstring_with_size
- Fixed array param conversion and cleanup - using two passes over the params where necessary.
- Added array conversions when calling wrapped JS objects.
- Added expanded array tests.
- Avoid repeated atomization of 'value' property name.
As this changed the generated interface signatures, I had to change all of the uses to avoid bustage. Any corners of the browser that aren't built by default, or that I haven't discovered how to build, may be at risk of bustage if they use string or wstring attributes. (This could mean blackwood; sorry, guys!)
Many thanks to Alec Flett (alecf@netscape.com) for preparing diffs for the mailnews portion of the signature changes; thanks also to Ariel Backenroth (arielb@rice.edu) and Mike Shaver (shaver@mozilla.org) for help with updating the tree with NS_DECL_NSIFOO macros; everwhere where one of these macros was used was one less place I had to manually add 'const'.
Also removed extraneous space from generated method signatures, leftover from Brendan's capitalization spam, and made 'const decl must be of type short or long' an error rather than just a warning.
Subject:
Embedding the shell
Date:
Wed, 8 Sep 1999 16:01:44 -0400
From:
"Kurt Westerfeld" <kurt@westerfeld.com>
To:
"Norris Boyd" <norris@netscape.com>
Norris, please find the attached zip file containing the (minor)
modifications to the Rhino shell module that enables the shell to be
embedded in a host application.
There are two areas of change to be concerned about:
1. Any and all references to System.in/out/err have been modified to use
Main.in/out/err, which default to System.in/out/err. Methods to do a setIn,
setOut, and setErr were added. Note that in/out/err on Main were made
static public, so that the jikes compiler wouldn't complain (I had them as
static protected, but when accessed outside of the package, a warning was
issued).
2. The global and sharedGlobal static variables were made protected so
that my app can make use of them (to add extensions after an initial pass
through main()).
That's it.
I have successfully used the facility to drop a remote telnet server into
the shell interpreter, effectively giving our server a remote shell
interpreter. It's quite nice, as we have a lot of extensions to Rhino
written that blend into our server already.
PS, I am still working on the array issues, but made a lot of progress
today. I just wanted to get this stuff off my desk.
Thanks!
________________________________________________________________________
Kurt Westerfeld
Managed Object Solutions
2722 Merrilee Drive
Suite 350
Fairfax, VA 22031
Phone: 703.208.3330 x225
Fax: 703.208.3331
http://www.mosol.com
mailto:kurt@mosol.com
shell.zip
Name:
shell.zip
Type:
Zip Compressed Data (application/x-zip-compressed)
Encoding:
base64
==============================================================================
Subject:
Rhino Array Source (Fixed)
Date:
Thu, 9 Sep 1999 14:12:03 -0400
From:
"Kurt Westerfeld" <kurt@mosol.com>
To:
"Norris Boyd" <norris@netscape.com>
Attached is NativeJavaObject.java, which seems to now pass the tests supplied to me by you and Scott. Not a lot of change, but a lot of
testing and thinking was involved. <g>
PS. I also fixed a bug in reportConversionError() which was throwing an IllegalArgumentException inside of the MessageFormat class at times.
It also looks a little nicer (uses formatting from NativeJavaMethod) and closer to the C implementation.
NativeJavaObject.java
Name:
NativeJavaObject.java
Type:
Java Source File (text/java)
Encoding:
quoted-printable
beard%netscape.com
sfraser%netscape.com
shaver%netscape.com
dmose%mozilla.org
norris%netscape.com
rogerl%netscape.com
kipp%netscape.com
mscott%netscape.com
dp%netscape.com
cbegle%netscape.com
brendan%mozilla.org
brade%netscape.com
waterson%netscape.com
jband%netscape.com
mccabe%netscape.com
slamm%netscape.com
colin%theblakes.com
waldemar%netscape.com
cls%seawood.org
briano%netscape.com
toshok%hungry.com
fur%netscape.com
mcafee%netscape.com
cyeh%netscape.com