npm package for deltachat-rpc-server
This is the successor of deltachat-node,
it does not use NAPI bindings but instead uses stdio executables
to let you talk to core over jsonrpc over stdio.
This simplifies cross-compilation and even reduces binary size (no CFFI layer and no NAPI layer).
📚 Docs: https://js.jsonrpc.delta.chat/
Usage
The minimum nodejs version for this package is
16
npm i @deltachat/stdio-rpc-server @deltachat/jsonrpc-client
import { startDeltaChat } from "@deltachat/stdio-rpc-server";
import { C } from "@deltachat/jsonrpc-client";
async function main() {
const dc = await startDeltaChat("deltachat-data");
console.log(await dc.rpc.getSystemInfo());
dc.close();
}
main();
For a more complete example refer to https://github.com/deltachat-bot/echo/tree/master/nodejs_stdio_jsonrpc.
How to use on an unsupported platform
You need to have rust installed to compile deltachat core for your platform and cpu architecture. https://rustup.rs/ is the recommended way to install rust. Also your system probably needs more than 4GB RAM to compile core, alternatively your could try to build the debug build, which might take less RAM to build.
- clone the core repo, right next to your project folder:
git clone git@github.com:deltachat/deltachat-core-rust.git - go into your core checkout and run
git pullandgit checkout <version>to point it to the correct version (needs to be the same version the@deltachat/jsonrpc-clientpackage has) - run
cargo build --release --package deltachat-rpc-server --bin deltachat-rpc-server
Then you have 2 options:
point to deltachat-rpc-server via direct path:
# start your app with the DELTA_CHAT_RPC_SERVER env var
DELTA_CHAT_RPC_SERVER="../deltachat-core-rust/target/release/deltachat-rpc-server" node myapp.js
install deltachat-rpc-server in your $PATH:
# use this to install to ~/.cargo/bin
cargo install --release --package deltachat-rpc-server --bin deltachat-rpc-server
# or manually move deltachat-core-rust/target/release/deltachat-rpc-server
# to a location that is included in your $PATH Environment variable.
And make sure to enable the takeVersionFromPATH option:
startDeltaChat("data-dir", { takeVersionFromPATH: true });
How does it work when you install it
NPM automatically installs platform dependent optional dependencies when os and cpu fields are set correctly.
references:
- https://napi.rs/docs/deep-dive/release#3-the-native-addon-for-different-platforms-is-distributed-through-different-npm-packages, webarchive version
- https://docs.npmjs.com/cli/v6/configuring-npm/package-json#cpu
- https://docs.npmjs.com/cli/v6/configuring-npm/package-json#os
When you import this package it searches for the rpc server in the following locations and order:
DELTA_CHAT_RPC_SERVERenvironment variable- use the PATH when
{takeVersionFromPATH: true}is supplied in the options. - prebuilds in npm packages
so by default it uses the prebuilds.
How do you built this package in CI
- To build platform packages, run the
build_platform_package.pyscript:python3 build_platform_package.py <cargo-target> # example python3 build_platform_package.py x86_64-apple-darwin - Then pass it as an artifact to the last CI action that publishes the main package.
- upload all packages from
deltachat-rpc-server/npm-package/platform_package. - then publish
deltachat-rpc-server/npm-package,- this will run
update_optional_dependencies_and_version.js(in theprepackscript), which puts all platform packages intooptionalDependenciesand updates theversioninpackage.json
- this will run
How to build a version you can use locally on your host machine for development
You can not install the npm packet from the previous section locally, unless you have a local npm registry set up where you upload it too. This is why we have separate scripts for making it work for local installation.
- If you just need your host platform run
python scripts/make_local_dev_version.py - note: this clears the
platform_packagefolder - (advanced) If you need more than one platform for local install you can just run
node scripts/update_optional_dependencies_and_version.jsafter building multiple platforms withbuild_platform_package.py
Thanks to nlnet
The initial work on this package was funded by nlnet as part of the Delta Tauri Project.