This patch splits the execution mode selection from the environment
setup in `lib/internal/bootstrap/node.js`, and split the entry point
of different execution mode into main scripts under
`lib/internal/main`:
- `check_syntax.js`: used when `-c`/`--check` which only checks the
syntax of the input instead of executing it.
- `eval_stdin.js`: used when `-e` is passed without value and stdin
is not a TTY (e.g. something is piped).
- `eval_string`: used when `-e` is passed along with a string argument
- `inspect.js`: for `node inspect`/`node debug`
- `print_bash_completion.js`: for `--completion-bash`
- `print_help.js`: for `--help`
- `prof_process.js`: for `--prof-process`
- `repl.js`: for the REPL
- `run_main_module.js`: used when a main module is passed
- `run_third_party_main.js`: for the legacy `_third_party_main.js`
support
- `worker_thread.js`: for workers
This makes the entry points easier to navigate and paves the way
for customized v8 snapshots (that do not need to deserialize
execution mode setup) and better embedder APIs.
As an example, after this patch, for the most common case where
Node.js executes a user module as an entry point, it essentially
goes through:
- `lib/internal/per_context.js` to setup the v8 Context (which is
also run when setting up contexts for the `vm` module)
- `lib/internal/bootstrap/loaders.js` to set up internal binding
and builtin module loaders (that are separate from the loaders
accessible in the user land).
- `lib/internal/bootstrap/node.js`: to set up the rest of the
environment, including various globals and the process object
- `lib/internal/main/run_main_module.js`: which is selected from
C++ to prepare execution of the user module.
This patch also removes `NativeModuleLoader::CompileAndCall` and
exposes `NativeModuleLoader::LookupAndCompile` directly so that
we can handle syntax errors and runtime errors of bootstrap
scripts differently.
PR-URL: https://github.com/nodejs/node/pull/25667
Reviewed-By: Anna Henningsen <anna@addaleax.net>
The idea is to allow the C++ layer to run arbitrary scripts
as the main script. This paves the way for
- cctest of the execution of Node.js instances
- Earlier handling of per-process CLI options that affect
execution modes (those usually do not make sense for the
embedders).
- Targets like mkcodecache or mksnapshot.
Also moves the handling of `_third_party_main.js` into C++.
PR-URL: https://github.com/nodejs/node/pull/25474
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Minwoo Jung <minwoo@nodesource.com>
This patch refactors the branches for choosing the mode to run
Node.js in `internal/bootstrap/node.js`. Instead of inlining the
decision making all in `startup`, we create a `startExecution()`
function which either detects and start the non-user-code mode,
or prepares for user code execution (worker setup, preloading modules)
and starts user code execution.
We use early returns when we decide the mode to run Node.js in for fewer
indentations and better readability.
This patch also adds a few comments about the command-line switches
and a few TODOs to remove underscore properties on `process` that
are mainly used for bootstrap mode branching. It also includes
a few other refactoring such as inlining functions/variables
that are not reused and removing the default argument of
`evalScript` for better clarity.
PR-URL: https://github.com/nodejs/node/pull/24673
Reviewed-By: Gus Caplan <me@gus.host>
Reviewed-By: Daniel Bevenius <daniel.bevenius@gmail.com>
Reviewed-By: Anna Henningsen <anna@addaleax.net>
Reviewed-By: Colin Ihrig <cjihrig@gmail.com>
- 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>
Show the stack trace for the `eventemitter.emit('error')` call
in the case of an uncaught exception.
Previously, there would be no clue in Node’s output about where
the actual `throw` comes from.
PR-URL: https://github.com/nodejs/node/pull/19003
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
