duktape

Commit Graph

Author	SHA1	Message	Date
Sami Vaarala	c36d48a4fc	Releases: debugger reattach	9 years ago
Sami Vaarala	8b2bbb5135	Delay detached_cb to between messages Without this it's possible for the following to happen: - The debug transport gets broken by e.g. a syntax error during message handling. - The detached_cb is called when marking transport bad. The detached callback immediately reattaches. - The debug command processing continues, potentially reading from and writing to the new connection which causes the connection to go out of sync. Delaying the detached_cb call to the message loop, right after one message has been fully processed, avoids this issue and keeps the connections cleanly separated. The solution is a bit awkward: duk_debug_do_detach() is split into two. Outside callers (duk_debugger_detach(), duk_heap_free()) still have a single function helper to hide the split.	9 years ago
Sami Vaarala	aab5494a7b	Document debugger attach call from detached cb	9 years ago
Sami Vaarala	1cffa519db	Duk cmdline #if 0'd debugger reattach test	9 years ago
Sami Vaarala	20da8b7f64	Add an explicit 'finish' call to debug trans	9 years ago
Sami Vaarala	37000d8f3a	Reorder detach cleanup and detach callback This should make it easier to allow a detached callback to reattach immediately without leaving the debugger state inconsistent. Requires careful handling of the detach/reattach sequence: - Both the detached callback and its udata must be stashed, state cleared, and only then the detached callback can be called. - When the Detach command has been handled, must avoid skipping to EOM because that would happen in a potentially new connection which would cause that connection to be out of sync. Returning to the message loop is fine, and since duk_debugger_attach() marks the state dirty, a Status will be immediately sent (followed by the handshake line, sent inline by duk_debugger_attach()).	9 years ago
Sami Vaarala	eeccde0e4a	Merge pull request #412 from fatcerberus/debug-get-locals-callstack-level Implement callstack level for debugger commands	9 years ago
Bruce Pascoe	ecdad014b5	Trivia fixes for debugger callstack level	9 years ago
Bruce Pascoe	973d6cf85a	Document optional callstack level in debugger.rst	9 years ago
Bruce Pascoe	426c6bc88f	Implement callstack level for debugger commands GetLocals, GetVar, PutVar, and Eval will now accept an optional negative callstack offset specifying the function activation to operate on. This offset has the same semantics as the argument of Duktape.act(): -1 is the topmost activation, -2 is its caller, etc.	9 years ago
Sami Vaarala	c079f14ae2	Merge pull request #428 from svaarala/fix-nested-prop-assignment-gh427 Add a testcase for already fixed nested property assignment bug	9 years ago
Sami Vaarala	ed469076a2	Releases: nested property assignment	9 years ago
Sami Vaarala	5bda532d12	Add testcase for nested prop assign (GH-427)	9 years ago
Sami Vaarala	3489418401	Merge pull request #422 from svaarala/perf-string-intern-utf8-clen Improve string intern computation of UTF-8 character length	9 years ago
Sami Vaarala	bd745340a0	Releases: string utf-8 clen	9 years ago
Sami Vaarala	9931e7e21a	Makefile perf test indent fix	9 years ago
Sami Vaarala	bc7a732002	Prune and clean up string charlen algorithms	9 years ago
Sami Vaarala	1bddde8daf	Multiple alternatives for string UTF-8 charlen	10 years ago
Sami Vaarala	36b8c7b4e7	Test cases for string charlen improvement - Fix string intern test case ("intern match"), which didn't keep a reference to the string so that the string was actually not already in the string table. - Add character length computation performance test cases for both pure ASCII and non-ASCII strings. - Add character length computation correctness test case, which exercises different string lengths and random string bytes.	9 years ago
Sami Vaarala	29c9ac4a1b	Merge pull request #420 from svaarala/contributing-mergeready-note CONTRIBUTING mergeready change	9 years ago
Sami Vaarala	4591994259	CONTRIBUTING mergeready change Labels can't be applied by non-collaborators so use comments instead.	9 years ago
Sami Vaarala	e6996f9c48	Merge pull request #419 from svaarala/debugger-request-value-stack-restore More robust value stack handling for debugger message loop	9 years ago
Sami Vaarala	ac5f110251	More robust value stack handling for dbg messages - Change duk__debug_process_message() to store/restore value stack top for message handlers. This is more robust than requiring that every debug message handler code path always manages the value stack 100% correctly. - Change existing message handlers so that trivial value stack cleanups are not done in the message handler. Value stack popping is still necessary for loops where the value stack might grow without bound.	9 years ago
Sami Vaarala	b88bbabd17	Merge pull request #360 from svaarala/add-hack-const-support Add hack "const" keyword support	9 years ago
Sami Vaarala	ed424c1e95	Releases: minimal const	9 years ago
Sami Vaarala	b5a28bcbb8	Website changes for 'const'	9 years ago
Sami Vaarala	c7a47dca7a	Minimal const support, treat mostly like "var" Only difference now is that an initializer is required for const declarations.	9 years ago
Sami Vaarala	c3e0e001a7	Testcase for minimal const support	9 years ago
Sami Vaarala	bf94a57902	Merge pull request #405 from svaarala/invalid-base-value-key-name Human readable object/key summary for property access errors	9 years ago
Sami Vaarala	2de8866b68	Releases: object/key summary in property errors	9 years ago
Sami Vaarala	067e660bb2	Makefile and matrix test for prop error messages	9 years ago
Sami Vaarala	ad78aea94d	Add property error note to sandboxing doc	9 years ago
Sami Vaarala	6590b783d9	Testcases for prop error messages, expect fixes	9 years ago
Sami Vaarala	f9bf05aa4f	Config options for prop error base/key summary - Add DUK_USE_VERBOSE_PROP_ERRORS config option - Add placeholder security sensitive config example - Low memory config option update	9 years ago
Sami Vaarala	e0a06cd7bd	Remove unused string readable tests	9 years ago
Sami Vaarala	394cf6d67e	Add summary of key/value for property errors Add internal duk_push_tval_readable() which summarizes a duk_tval into a human readable, limited length sanitized string without side effects. It is useful for summarizing values safely in e.g. error messages. The internal call is a candidate for the public API later on (in a slightly different form).	9 years ago
Sami Vaarala	f254dbd259	Merge pull request #400 from svaarala/debugger-doc-gc-freeze Debugger document note on GC freezing and walking	9 years ago
Sami Vaarala	7011ebaf86	Debugger document note on GC freezing and walking	9 years ago
Sami Vaarala	4e2cc1f5a1	Merge pull request #407 from svaarala/fix-date-tzo-riscos Fix local time offset on RISC OS	9 years ago
Sami Vaarala	1b4b1933f9	Releases: mktime() RISC OS	9 years ago
Sami Vaarala	22e631a640	Force tm_isdst to 0 before mktime() Forcing tm_isdst to 0 before mktime() and removing a previous DST check after mktime() seems to work on Linux and should (potentially) fix a RISC OS issue.	9 years ago
Sami Vaarala	a3843f38a4	Merge pull request #414 from svaarala/fix-radix16-leading-zero-gh413 Fix parseInt() explicit radix 16 handling for leading zero	9 years ago
Sami Vaarala	a3a41fc3d7	Releases: parseInt radix 16	9 years ago
Sami Vaarala	1de96e94d3	parseInt() GH-413 fix and octal parsing change Don't autodetect octal in parseInt(), requires explicit radix 8 for that.	9 years ago
Sami Vaarala	f29cd9925e	Add octal parseInt() testcases - Bug testcase for GH-413 - Testcase for removing octal autodetection in parseInt() Fix expect string of numconv parseInt() testcase to match E5.1 behavior.	9 years ago
Sami Vaarala	86c448c50c	Merge pull request #403 from svaarala/perf-push-this-direct Direct valstack access for duk_push_this()	9 years ago
Sami Vaarala	bdf6829bd3	Releases: duk_push_this() valstack access	9 years ago
Sami Vaarala	5de6133ef1	Direct value stack access in duk_push_this()	9 years ago
Sami Vaarala	16ac453eaf	Merge pull request #398 from svaarala/avoid-rdflib-in-dist Avoid rdflib in non-official dist	9 years ago
Sami Vaarala	de34db7f9c	Makefile: keep compiler.jar, --create-spdx Keeping compiler.jar means that if you once "make dist-src-official" or "make compiler.jar", "make dist" will then closure-minify further builds.	9 years ago

1 2 3 4 5 ...

5696 Commits (c36d48a4fcf59b015b046cd006ee7869cb417f02) All Branches Search

5696 Commits (c36d48a4fcf59b015b046cd006ee7869cb417f02)

All Branches