This commit improves module loading performance by at least ~25-35%
in the module-loader benchmarks.
Some optimization strategies include:
* Try-finally/try-catch isolation
* Replacing regular expressions with manual parsing
* Avoiding unnecessary string and array creation
* Avoiding constant recompilation of anonymous functions and
function definitions within functions
PR-URL: https://github.com/nodejs/node/pull/5172
Reviewed-By: James M Snell <jasnell@gmail.com>
V8 5.0 introduced a small modification for the unexpected end of input
error.
PR-URL: https://github.com/nodejs/node/pull/5945
Reviewed-By: bnoordhuis - Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: indutny - Fedor Indutny <fedor.indutny@gmail.com>
Change the regular expression that recognizes “simple” JS expressions
to requiring that the full line needs to match it.
Previously, in terms like `a().b.`, `b.` would be a partial match.
This meant that completion would evaluate `b` and either fail with
a `ReferenceError` or, if `b` was some global, return the properties
of the global `b` object.
PR-URL: https://github.com/nodejs/node/pull/6192
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
os.userInfo() calls libuv's uv_os_get_passwd() function. It returns
an object containing the current effective user's username, uid,
gid, shell, and home directory. On Windows, the uid and gid are
-1, and the shell is null.
Refs: https://github.com/nodejs/node/issues/5582
PR-URL: https://github.com/nodejs/node/pull/6104
Reviewed-By: James M Snell <jasnell@gmail.com>
In 68990948fe (https://github.com/nodejs/node/pull/2325),
the conditions for increasing `readableState.awaitDrain` when
writing to a piping destination returns false were changed so
that they could not actually be met, effectively leaving
`readableState.awaitDrain` with a constant value of 0.
This patch changes the conditions to testing whether the
stream for which `.write()` returned false is still a piping
destination, which was likely the intention of the original
patch.
Fixes: https://github.com/nodejs/node/issues/5820
Fixes: https://github.com/nodejs/node/issues/5257
PR-URL: https://github.com/nodejs/node/pull/6023
Reviewed-By: Brian White <mscdex@mscdex.net>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Adds additional `targetStart`, `targetEnd`, `sourceStart,
and `sourceEnd` arguments to `Buffer.prototype.compare`
to allow comparison of sub-ranges of two Buffers without
requiring Buffer.prototype.slice()
Fixes: https://github.com/nodejs/node/issues/521
PR-URL: https://github.com/nodejs/node/pull/5880
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
The only test with modifications is `test-stdin-child-proc` that was
passing when it should not because the exit code of the child process
was not being checked.
PR-URL: https://github.com/nodejs/node/pull/6087
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Sakthipriyan Vairamani <thechargingvolcano@gmail.com>
Reviewed-By: Claudio Rodriguez <cjrodr@yahoo.com>
A win32-only test was verifying that path.win32._makeLong('C:')
would return the current working directory. This would only work if
current working directory was also on the C: device. Fix is to grab
the device letter for current working directory, and pass that to
_makeLong().
PR-URL: https://github.com/nodejs/node/pull/6067
Reviewed-By: Trott - Rich Trott <rtrott@gmail.com>
Reviewed-By: Joao Reis <reis@janeasystems.com>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
ParseArrayIndex() was requesting a Uint32Value(), but assigning it to an
in32_t. This caused slight differences in error message reported in edge
cases of argument parsing. Fixed by getting the IntegerValue() before
checking if the value is < 0. Added test of API that was affected.
PR-URL: https://github.com/nodejs/node/pull/6084
Reviewed-By: James M Snell <jasnell@gmail.com>
* use common.mustCall() to verify all tests have run
* eliminate unneeded removeTestFile()
* eliminate unneeded var leaking into global scope
* var -> const
* remove instance of let
PR-URL: https://github.com/nodejs/node/pull/6050
Reviewed-By: James M Snell <jasnell@gmail.com>
This allows third-party tools to check whether or not a handle that
can be unreferenced is unreferenced at a particular time.
Notably, this should be helpful for inspection via AsyncWrap.
Also, this is useful even to node's internals, particularly timers.
Refs: https://github.com/nodejs/node/pull/5828
Refs: https://github.com/nodejs/node/pull/5827
PR-URL: https://github.com/nodejs/node/pull/5834
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Trevor Norris <trev.norris@gmail.com>
The last change to this test landed before a nit about strict mode was
addressed, so this change addresses that.
PR-URL: https://github.com/nodejs/node/pull/6047
Refs: https://github.com/nodejs/node/pull/6017
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Roman Klauke <romaaan.git@gmail.com>
The tests used to rely on precise timing of when a JavaScript object
would be garbage collected to ensure that there is enough memory
available on the system. Switch the test to use a malloc/free pair
instead.
Ref: https://github.com/nodejs/node/pull/5945
PR-URL: https://github.com/nodejs/node/pull/6039
Reviewed-By: jasnell - James M Snell <jasnell@gmail.com>
Reviewed-By: evanlucas - Evan Lucas <evanlucas@me.com>
Reviewed-By: Trott - Rich Trott <rtrott@gmail.com>
Detect whether a gzip file is being passed to `unzip*` by
testing the first bytes for the gzip magic bytes, and setting
the decompression mode to `GUNZIP` or `INFLATE` according to
the result.
This enables gzip-only features like multi-member support
to be used together with the `unzip*` autodetection support
and thereby makes `gunzip*` and `unzip*` return identical
results for gzip input again.
Add a simple test for checking that features specific to
`zlib.gunzip`, notably support for multiple members, also work
when using `zlib.unzip`.
PR-URL: https://github.com/nodejs/node/pull/5884
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: James M Snell <jasnell@gmail.com>
Currently, 'keypress' events include the sequence and key info
for named keys, but only the sequence for unnamed keys. This
commit causes the key info to be included in both cases.
PR-URL: https://github.com/nodejs/node/pull/6024
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>
This commit adds documentation to the already publicly available
readline.emitKeypressEvents() method.
PR-URL: https://github.com/nodejs/node/pull/6024
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Fedor Indutny <fedor.indutny@gmail.com>
Use `global` to be explicit that a global variable is intended.
PR-URL: https://github.com/nodejs/node/pull/6014
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
There was a comment in `test-domain-crypto.js` indicating that the
pollution of the `global` object with a `domain` property was
intentional. Provide more information in the comment so someone may
easily determine why. Use `global.domain` rather than declaring `domain`
without the `var` keyword to more clearly signal that the pollution is
intentional.
PR-URL: https://github.com/nodejs/node/pull/6017
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
The test intentionally assigns a global. Use `global` namespace to make
it clear that it is intentional and not an accidental leak.
PR-URL: https://github.com/nodejs/node/pull/6026
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
This commit fixes an inconsistency in absolute path checking compared
to the absolute path detection used by the other path.win32 functions.
Fixes: https://github.com/nodejs/node/issues/6027
PR-URL: https://github.com/nodejs/node/pull/6028
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
A handful of tests in `test/pummel` were failing due to undefined
variables.
The tests in pummel are not run in CI or otherwise exercised regularly
so these failures can go unnoticed for a long time.
PR-URL: https://github.com/nodejs/node/pull/6012
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Use empty string instead of `www.google.com` for tests where we are just
doing parameter evaluation. This will avoid DNS lookups which appear to
be causing flakiness on Raspberry Pi devices in CI.
PR-URL: https://github.com/nodejs/node/pull/5996
Fixes: https://github.com/nodejs/node/issues/5554
Reviewed-By: Michael Dawson <michael_dawson@ca.ibm.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
`require-buffer` is only enabled in the `lib` directory. There is no
need to disable it in `test`.
PR-URL: https://github.com/nodejs/node/pull/6013
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Remove timer to avoid the test timing out occasionally.
PR-URL: https://github.com/nodejs/node/pull/5676
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Claudio Rodriguez <cjrodr@yahoo.com>
If `kNoZeroFill` is set here, it won't be reset in case of
pooled allocation. In case of "slow" allocation it will be
set later anyway.
Fixes: https://github.com/nodejs/node/issues/6006
PR-URL: https://github.com/nodejs/node/pull/6007
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Evan Lucas <evanlucas@me.com>
Reviewed-By: Myles Borins <myles.borins@gmail.com>
This commit adds support for passing DNS lookup hints to
createConnection().
PR-URL: https://github.com/nodejs/node/pull/6000
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: James M Snell <jasnell@gmail.com>
Informative error messages are very important for developers and could
possibly save hours of debugging and frustration. This improves the error
message thrown when writing invalid data into a socket, by communicating
what's expected compared to what the developer just tried to write.
PR-URL: https://github.com/nodejs/node/pull/5981
Reviewed-By: Brian White <mscdex@mscdex.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
IPv6 addresses in Host header (URI), must be enclosed within
square brackets, in order to properly separate the host address
from any port reference.
PR-URL: https://github.com/nodejs/node/pull/5314
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Evan Lucas <evanlucas@me.com>
`test-stdtout-close-unref.js` will fail if `process.stdin._handle` does
not exist. On UNIX-like operating systems, you can see this failure this
way:
./node test/parallel/test-stdout-close-unref.js < /dev/null
This issue has been experienced by @bengl and @drewfish in a Docker
container. I'm not sure why they are experiencing it in their
environment, but since it is possible that the `_handle` property does
not exist, let's use `child_process.spawn()` to make sure it exists.
PR-URL: https://github.com/nodejs/node/pull/5916
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Do not convert typed arrays to `Buffer` for deepEqual since
their values may not be accurately represented by 8-bit ints.
Instead perform binary comparison of underlying `ArrayBuffer`s,
but only when the array types match.
Never apply any kind of optimization for floating-point typed
arrays since bit pattern equality is not the right kind of check
for them.
PR-URL: https://github.com/nodejs/node/pull/5910
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Fixes: https://github.com/nodejs/node/issues/5907
`<`-ing a file into stdin actually results in a `fs.ReadStream`, rather
than a `tty.ReadStream`, and as such does not inherit from net.Socket,
unlike the other possible stdin options.
Refs: https://github.com/nodejs/node/pull/5916
PR-URL: https://github.com/nodejs/node/pull/5935
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Make sure that, even if an `inflate()` call only sees the first
few bytes of a following gzip member, all members are decompressed
and part of the full output.
This change also modifies behaviour for trailing garbage:
If there is trailing garbage which happens to start with the
gzip magic bytes, it is no longer discarded but rather throws
an error, since we cannot reliably tell random garbage from
a valid gzip member anyway and have to try and decompress it.
(Null byte padding is not affected, since it has been pointed
out at various occasions that such padding is normal and
discarded by `gzip(1)`, too.)
Adds tests for the special case that the first `inflate()` call
receives only the first few bytes of a second gzip member but
not the whole header (or even just the magic bytes).
PR-URL: https://github.com/nodejs/node/pull/5883
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: James M Snell <jasnell@gmail.com>
Check that piping a large chunk of data from `process.stdin`
into `process.stdout` does not lose any data by verifying that
the output has the same size as the input.
This is a regression test for #5927 and fails for the commits
in the range [ace100945..89abe8680).
PR-URL: https://github.com/nodejs/node/pull/5949
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
There were 2 tests using curl:
`test-http-304.js` is removed because it was initially included to test
that the 304 response does not contain a body, and this is already
covered by `test-http-chunked-304.js`.
`test-http-curl-chunk-problem` has been renamed and refactored so
instead of using curl, it uses 2 child node processes: one for sending
the HTTP request and the other to calculate the sha1sum. Originally,
this test was introduced to fix a bug in `nodejs@0.2.x`, and it was not
fixed until `nodejs@0.2.5`. A modified version of this test has been run
with `nodejs@0.2.0` and reproduces the problem. This same test has been
run with `nodejs@0.2.6` and runs correctly.
Fixes: https://github.com/nodejs/node/issues/5174
PR-URL: https://github.com/nodejs/node/pull/5750
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Johan Bergström <bugs@bergstroem.nu>
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Currently we are not testing that resolution of local paths is
resolved first in the repl. This addition to `test-repl-require`
adds an additional fixture an ensures we won't regress in the future
PR-URL: https://github.com/nodejs/node/pull/5689
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
This fixes a bug where a 3rd party module found in node_modules,
would be preferred over a ./local module with the same name.
Fixes: https://github.com/nodejs/node/issues/5684
PR-URL: https://github.com/nodejs/node/pull/5689
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl>
Reviewed-By: Jeremiah Senkpiel <fishrock123@rocketmail.com>
Brian White
Rich Trott
Michaël Zasso
Anna Henningsen
Matteo Collina
cjihrig
Igor Kalashnikov
silverwind
James M Snell
Santiago Gimeno
Mike Kaufman
Trevor Norris
Jeremiah Senkpiel
Ali Ijaz Sheikh
Anna Henningsen
Vladimir Kurchatkin
Phillip Johnsen
Mihai Potra
Sakthipriyan Vairamani
Myles Borins