Fixes #2063.
REPLServer.prototype.resetContext:
Reset the line cache
REPLServer.prototype.memory (don't know if I like that name, called from finish)
pushes what cmd's have been executed against it into this.lines
pushes the "tab depth" for bufferedCommands, in this.lines.level
REPLServer.prototype.displayPrompt:
Uses "tab depth" from this.lines.level to adjust the prompt to visually
denote this depth e.g.
> asdf = function () {
… var inner = {
….. one:1
REPLServer.prototype.complete:
Now notices if there is a bufferedCommand and attempts determine locally
scoped variables by removing any functions from this.lines and evaling these
lines in a nested REPL e.g.
> asdf = function () {
… var inner = { one: 1};
… inn\t
will complete to 'inner' and inner.o\t will complete to 'inner.one'
If the nested REPL still has a bufferedCommand it will falls back to the
default.
ArrayStream is a helper class for the nested REPL to get commands pushed to it.
new REPLServer('', new ArrayStream());
Finally added two new REPL commands .save and .load, each takes 1 parameter,
a file and attempts to save or load the file to or from the REPL
respectively.
This is a combination of 20 commits. Their commit messages are preserved
below for the benefit of future generations.
* Adding a shortcut to easily compress/decompress a string of text.
* Making the API consistent. unzip should accept a Buffer for input as well.
* Adding docs.
* Oops, typo.
* Propagate error through the callback.
* Adding zlib from string tests.
* Typo in test.
* Remove 'end' listeners, and join buffers properly instead of joining them
as a string.
* Oops, needs to be rendered to a string.
* Updated test to include multi-byte characters.
* unzip should return a raw Buffer. Updated docs to reflect.
* And finally updating test.
* EventEmitter.destroy() is a bit more customary
* Revert "EventEmitter.destroy() is a bit more customary"
* Renaming internal methods to "buffer" instead of string.
* Remove the 'error' listeners as well.
* @isaacs: spacing/style, and compress duplicate functions into one
* @isaacs: Update docs
* @isaacs: doc style fix
When trying to connect to something that is not a UNIX socket, Linux returns
ECONNREFUSED, not ENOTSOCK.
We cannot atomically determine if the other end is a) a stale socket, or b) not
a socket at all, so let's accept both error codes.
This tests passes on node v0.4, but fails on node v0.5. v0.5 seems to
generally lack error codes for various error events related to UNIX
pipes.
Fixes #2001
http2.js
protocols object to store defaults for http and https, and use as a switch for supported protocols.
options.hostname > options.host > 'localhost'
if I have an options.auth element and I do not have an Authorization header, I do basic auth.
http.request collapses to new ClientRequest since the defaults are handled by the protocol object
test-http-url.parse*
Fixes #1390
Conflicts:
lib/http2.js
added a .path property = .pathname + .search for use with http.request
And tests to verify everything.
With the tests, I changed over to deepEqual, and I would note the comment on the test
['.//g', 'f:/a', 'f://g'], which I think is a fundamental problem
This supersedes pull 1596
Fix #1484
Fix #1834
Fix #1482
Fix #771
It's been a while now, and we've seen how this separate context thing
works. It constantly confuses people, and no one actually uses '.clear'
anyway, so the benefit of that feature does not justify the constant
WTFery.
This makes repl.context actually be a getter that returns the global
object, and prints a deprecation warning. The '.clear' command is gone,
and will report that it's an invalid repl keyword. Tests updated to
allow the require, module, and exports globals, which are still
available in the repl just like they were before, by making them global.
Squashed commit:
(- re tests) Cleaning up the `Script` test suite.
For whatever reason, there were several duplicate test files related to `Script`
and the `'vm'` module. I removed these, and fixed a few other small issues.
(More fixes coming in subsequent commits.)
Squashes: 19e86045a0..1e3dcff4eb
(api fix:1801 new:1801) `'vm'` module uses sandbox as prototype
As described in GH-1801, the `'vm'` module was handling the `sandbox` object
provided by the API consumer in a particularly terrible and fragile fashion: it
was simply shallow-copying any enumerable properties from the sandbox onto the
global context before executing the code, and then eventually copying any values
on the global context back into the sandbox object *afterwards*.
This commit removes all of that implementation, and utilizes the passed sandbox
object as the *prototype of the context* instead. A bit of a hack, but a very
effective one.
This no longer allows for new variables created in the global context to be
placed into your sandbox after execution has completed, but that’s for the best
anyway, as it’s not very in line with the concept of a “box of passed-in
context.” I’m planning to further implement an interface for API consumers to
acquire the *actual global* from within the VM soon, thus allowing for
separation-of-concerns: providing data *to* the VM via the sandbox-prototype,
and exploring the internal environment of the VM itself.
// GitHub cruft: closes #1801
Squashes: 43b8e3c..209ed86
Only register once for listening when passing a callback to Server.listen(),
this prevents servers recycled using close() from invoking the callback when
Server.listen() is called later.
This commit removes one assert from test-console.js in which we check that
process.stderr.write returns true. In the case of a dump to a file we cannot
guarantee this any longer now that it goes through fs.WriteStream.
Coerce fractional, negative and non-numeric length arguments to numbers.
Fractional numbers are rounded up, negative numbers and non-numeric values
are set to zero.
Coerce fractional, negative and non-numeric length arguments to numbers.
Fractional numbers are rounded up, negative numbers and non-numeric values
are set to zero.
* remove useless clearline call at Interface start
* silence after .handleBreak()
* output '\b' if this.stdout is not a tty (debugger)
* add '\b' checks for clearline (test)
* Don't buffer command, before it's execution (repl)
* `quit` command, custom streams for .start, stubbed out test, disable
history repeation for non-tty (debugger)
If hasOwnProperty is overridden, then calling `obj.hasOwnProperty(prop)`
can fail. Any time a dictionary of user-generated items is built, we
cannot rely on hasOwnProperty being safe, so must call it from the
Object.prototype explicitly.
and lazy load modules for process.stdout
This probably breaks test-module-load-list on windows, but it hopefully is
an easy fix to replace "_posix" bindings with "_win32".
