The maximum hard limit that `util.inspect()` could theoretically handle
is the maximum string size. That is ~2 ** 28 on 32 bit systems and
~2 ** 30 on 64 bit systems.
Due to the recursive algorithm a complex object could easily exceed
that limit without throwing an error right away and therefore
crashing the application by exceeding the heap limit.
`util.inspect()` is fast enough to compute 128 MB of data below one
second on an Intel(R) Core(TM) i7-5600U CPU. This hard limit allows
to inspect arbitrary big objects from now on without crashing the
application or blocking the event loop significantly.
PR-URL: https://github.com/nodejs/node/pull/22756
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: John-David Dalton <john.david.dalton@gmail.com>
This significantly improves the inspection performance for all array
types. From now on only the visible elements cause work instead of
having to process all array keys no matter how many entries are
visible.
This also moves some code out of the main function to reduce the
overall function complexity.
PR-URL: https://github.com/nodejs/node/pull/22503
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: John-David Dalton <john.david.dalton@gmail.com>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Refael Ackermann <refack@gmail.com>
The former implementation lacked symbols on the iterator objects
without prototype. This is now fixed.
The special handling for overriding `Symbol.iterator` was removed as
it's very difficult to deal with this properly. Manipulating the
symbols is just not supported.
PR-URL: https://github.com/nodejs/node/pull/22437
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Calling `formatValue()` directly requires the indentation level to
be set manually. This was not the case so far in most cases and the
indentation was off in all these cases.
PR-URL: https://github.com/nodejs/node/pull/22291
Reviewed-By: Gus Caplan <me@gus.host>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: James M Snell <jasnell@gmail.com>
This makes sure values without prototype will still be inspected
properly and do not cause errors. It restores the original
information if possible.
Besides that it fixes an issue with boxed symbols: extra keys were
not visualized so far.
PR-URL: https://github.com/nodejs/node/pull/21869
Reviewed-By: Gus Caplan <me@gus.host>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: John-David Dalton <john.david.dalton@gmail.com>
This adds the number of not visible elements when inspecting iterators
while exceeding `maxArrayLength`.
It also fixes a edge case with `maxArrayLength` and the map.entries()
iterator. Now the whole entry will be visible instead of only the key
but not the value of the first entry.
Besides that it uses a slighly better algorithm that improves the
performance by skipping unnecessary steps.
PR-URL: https://github.com/nodejs/node/pull/20961
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Right now util.inspect will always escape single quotes. That is not
necessary though in case the string that will be escaped does not
contain double quotes. In that case the string can simply be wrapped
in double quotes instead.
If the string contains single and double quotes and it does not
contain `${` as part of the string, backticks will be used instead.
That makes sure only very few strings have to escape quotes at all.
Thus it increases the readability of these strings.
PR-URL: https://github.com/nodejs/node/pull/21624
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
This commit updates `util.inspect` to add an `n` suffix to BigInts that
appear in BigInt64Arrays. BigInts are formatted with an `n` suffix in
most cases, but this did not occur in BigInt64Arrays due to an apparent
oversight where the implementation of `inspect` for typed arrays assumed
that all typed array elements are numbers.
PR-URL: https://github.com/nodejs/node/pull/21499
Reviewed-By: Michaël Zasso <targos@protonmail.com>
Reviewed-By: Gus Caplan <me@gus.host>
Reviewed-By: Ruben Bridgewater <ruben@bridgewater.de>
Reviewed-By: Tiancheng "Timothy" Gu <timothygu99@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Yuta Hiroto <hello@hiroppy.me>
This reverts commit b994b8eff6.
This caused regressions in ecosystem code. While the change originally
was semver-major and could be postponed until after Node.js 10,
I think reverting it is a good choice at this point.
Also, I personally do not think defaulting to a shallow inspect
is a bad thing at all – quite the opposite: It makes `util.inspect()`
give an overview of an object, rather than providing a full
display of its contents. Changing the `depth` default to infinity
fundamentally changed the role that `util.inspect()` plays,
and makes output much more verbose and thus at times unusable
for `console.log()`-style debugging.
PR-URL: https://github.com/nodejs/node/pull/20017
Fixes: https://github.com/nodejs/node/issues/19405
Refs: https://github.com/nodejs/node/pull/17907
Reviewed-By: Rich Trott <rtrott@gmail.com>
Reviewed-By: Tiancheng "Timothy" Gu <timothygu99@gmail.com>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
1) So far extra keys on an (Set|Map)Iterator were ignored. Those
will now be visible.
2) Improve the performance of showing (Set|Map)Iterator by using
the cloned iterator instead of copying all entries first.
3) So far the output was strictly limited to up to 100 entries.
The limit will now depend on `maxArrayLength` instead (that
default is set to 100 as well) and the output indicates that
more entries exist than visible.
PR-URL: https://github.com/nodejs/node/pull/19259
Reviewed-By: Yosuke Furukawa <yosuke.furukawa@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Tiancheng "Timothy" Gu <timothygu99@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
This makes a effort to make sure all of these errors will actually
also show the received input.
On top of that it refactors a few tests for better maintainability.
It will also change the returned type to always be a simple typeof
instead of special handling null.
PR-URL: https://github.com/nodejs/node/pull/19445
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Joyee Cheung <joyeec9h3@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
- Moves the creation of `process.binding()`, `process._linkedBinding()`
`internalBinding()` and `NativeModule` into a separate file
`lib/internal/bootstrap_loaders.js`, and documents them there.
This file will be compiled and run before `bootstrap_node.js`, which
means we now bootstrap the internal module & binding system before
actually bootstrapping Node.js.
- Rename the special ID that can be used to require `NativeModule`
as `internal/bootstrap_loaders` since it is setup there. Also put
`internalBinding` in the object exported by `NativeModule.require`
instead of putting it inside the `NativeModule.wrapper`
- Use the original `getBinding()` to get the source code of native
modules instead of getting it from `process.binding('native')`
so that users cannot fake native modules by modifying the binding
object.
- Names the bootstrapping functions so their names show up
in the stack trace.
PR-URL: https://github.com/nodejs/node/pull/19112
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Anatoli Papirovski <apapirovski@mac.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Gus Caplan <me@gus.host>
Ruben Bridgewater