Sami Vaarala
|
31d3f4df06
|
release note update
|
11 years ago |
Sami Vaarala
|
8191338eb2
|
assert fix, missing parens
|
11 years ago |
Sami Vaarala
|
43118a541c
|
indirect eval from C
|
11 years ago |
Sami Vaarala
|
bebf480ede
|
allow eval() call from Duktape/C code (at least an issue if eval.bind() is used and result is called from C)
|
11 years ago |
Sami Vaarala
|
c9c4f809bd
|
release note update
|
11 years ago |
Sami Vaarala
|
573679398a
|
update description of Duktape/C function properties
|
11 years ago |
Sami Vaarala
|
d0cce6b77d
|
a simple testcase to check that built-in function prototype is Function.prototype (this prevents using an intermediate prototype before Function.prototype to provide custom properties for Duktape/C functions)
|
11 years ago |
Sami Vaarala
|
0a9d269cfa
|
Add a virtual 'length' property to Duktape/C functions if they don't have a concrete property of that name. Ideally the number could be overwritten with a concrete property value, but currently the value is non-writable. Making it writable is trivial, but putprop needs to be fixed so that overwriting a virtual property works. Currently putprop assumes that if an own property is writable, it cannot be a virtual property.
|
11 years ago |
Sami Vaarala
|
781b0a474b
|
release note update
|
11 years ago |
Sami Vaarala
|
d7f2f9700d
|
guide/api docs for stash API calls
|
11 years ago |
Sami Vaarala
|
89cbd53566
|
add testcases for stash api calls
|
11 years ago |
Sami Vaarala
|
aafb998851
|
add api calls to push heap, global, and thread level stash objects
|
11 years ago |
Sami Vaarala
|
b1b3f1197f
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
75f86eb13a
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
63da03eab0
|
compile warning fix for debug macros when colors are disabled
|
11 years ago |
Sami Vaarala
|
5276850b4c
|
compile warning fix
|
11 years ago |
Sami Vaarala
|
6ee91fd5f3
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
92453bc0b8
|
release note update, Array.prototype.splice() behavior change
|
11 years ago |
Sami Vaarala
|
6c78f9b8df
|
fix Array.prototype.splice() testcase to avoid the case when deleteCount is omitted (separate testcase covers that)
|
11 years ago |
Sami Vaarala
|
ebd80edcf9
|
fix Array.prototype.splice() to follow real world behavior when deleteCount is omitted entirely; real world behavior is to splice to end of array, standard behavior is to work as if deleteCount were undefined (= 0)
|
11 years ago |
Sami Vaarala
|
b15e74c7b9
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
ea32ad15ca
|
use 'nonstandard' to flag testcases which are real world deviations from spec, but not Duktape custom behavior as such
|
11 years ago |
Sami Vaarala
|
61c09a8270
|
add testcases for Array splice() without deleteCount argument: specification behavior is to work as if deleteCount were 0, but real world engines splice to end of array
|
11 years ago |
Sami Vaarala
|
1541b646bb
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
f4053b60ab
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
b3b1e8b634
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
6d837d726e
|
downgrade a few FIXMEs
|
11 years ago |
Sami Vaarala
|
8e915a8239
|
date fixme fixed
|
11 years ago |
Sami Vaarala
|
7730e0a5ab
|
downgrade a few FIXMEs
|
11 years ago |
Sami Vaarala
|
99e7897144
|
remove pretty pointless DUK_ERR_OK and DUK_ERR_FAIL defines, which were also easy to confuse with other DUK_ERR_xxx defines
|
11 years ago |
Sami Vaarala
|
b5035877f8
|
move date provider selection and date system headers to the general platform selection and system include selection part; this avoids duplicating the platform ladder
|
11 years ago |
Sami Vaarala
|
7a3ae6ae1f
|
fix amiga/atari endianness detection in rework
|
11 years ago |
Sami Vaarala
|
5e9d095e92
|
add a footprint estimate to index page
|
11 years ago |
Sami Vaarala
|
253fb15300
|
update endianness refs in guide
|
11 years ago |
Sami Vaarala
|
84ba77df23
|
use 'mixed endian' instead of 'middle endian' to refer to arm mixed endian; rename DUK_USE_LITTLE_ENDIAN to DUK_USE_INTEGER_LE (same for big, middle) for clarity; when using mixed endian, define DUK_USE_INTEGER_LE instead of DUK_USE_INTEGER_BE
|
11 years ago |
Sami Vaarala
|
325aef6d3e
|
reduce duplication in endianness detection by using DUK_F_BYTEORDER as an intermediate define to track both whether or not endianness detection has already succeeded, and also to record the byteorder in one value; at the end, define individual defines based on the intermediate one
|
11 years ago |
Sami Vaarala
|
aee6c86a30
|
add qnx to release notes and portability list
|
11 years ago |
Sami Vaarala
|
7d7a38b846
|
add QNX support to feature detection
|
11 years ago |
Sami Vaarala
|
9ff07ab039
|
add support for gcc/clang built-in endianness macros; when float word order not known, assume it matches integer endiannes (self tests will catch a false assumption at runtime); other feature detection trivia
|
11 years ago |
Sami Vaarala
|
1b824b4da7
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
413b9bed8f
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
e7197cd4b9
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
4330dcd98c
|
alternative to 'undefined' would be 'void 0'
|
11 years ago |
Sami Vaarala
|
d9f00bc01d
|
add a few tentative scan-build targets
|
11 years ago |
Sami Vaarala
|
a53774e744
|
ditz issue update
|
11 years ago |
Sami Vaarala
|
0aa5f74653
|
release checklist: more git maintenance
|
11 years ago |
Sami Vaarala
|
ab84d0fd6f
|
release notes update
|
11 years ago |
Sami Vaarala
|
fd614b79dc
|
add duk_push_object_helper_proto() which takes an actual prototype pointer (not just a built-in index)
|
11 years ago |
Sami Vaarala
|
041f911aa3
|
rename DUK_HOBJECT_SET_PROTOTYPE to DUK_HOBJECT_SET_PROTOTYPE_UPDREF to make it clearer that it updates refcounts too (decs old, incs new)
|
11 years ago |
Sami Vaarala
|
d5f202da6e
|
promote plain buffer and pointer values to Buffer and Pointer objects in Object constructor (consistent with behavior for other primitive types)
|
11 years ago |