1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236
//! Definition for encoding of custom sections within core wasm modules of
//! component-model related data.
//!
//! When creating a component from a source language the high-level process for
//! doing this is that code will be generated into the source language by
//! `wit-bindgen` or a similar tool which will be compiled down to core wasm.
//! The core wasm file is then fed into `wit-component` and a component is
//! created. This means that the componentization process is decoupled from the
//! binding generation process and intentionally affords for linking together
//! libraries into the main core wasm module that import different interfaces.
//!
//! The purpose of this module is to define an intermediate format to reside in
//! a custom section in the core wasm output. This intermediate format is
//! carried through the wasm linker through a custom section whose name starts
//! with `component-type`. This custom section is created
//! per-language-binding-generation and consumed by slurping up all the
//! sections during the component creation process.
//!
//! The custom section here contains `World`, the interpretation of a "world"
//! of a component, along with how strings are encoded for all the specified
//! interfaces. Currently the encoding is:
//!
//! * First, a version byte (`CURRENT_VERSION`). This is intended to detect
//! mismatches between different versions of the binding generator and
//! `wit-component` which may or may not become a problem over time.
//!
//! * Next a string encoding byte.
//!
//! * Afterwards a "types only" component encoding of a `World`
//! package through the `ComponentEncoder::types_only` configuration.
use crate::{decode_world, ComponentEncoder, StringEncoding};
use anyhow::{bail, Context, Result};
use indexmap::IndexMap;
use wasm_encoder::Encode;
use wasmparser::BinaryReader;
use wit_parser::World;
const CURRENT_VERSION: u8 = 0x01;
/// Result of extracting interfaces embedded within a core wasm file.
///
/// This structure is returned by the [`extract_module_interfaces`] function.
#[derive(Default)]
pub struct BindgenMetadata {
/// All interfaces found within a module, merged together into one `World`.
pub world: World,
/// Per-function options imported into the core wasm module, currently only
/// related to string encoding.
pub import_encodings: IndexMap<(String, String), StringEncoding>,
/// Per-function options exported from the core wasm module, currently only
/// related to string encoding.
pub export_encodings: IndexMap<String, StringEncoding>,
}
/// This function will parse the `wasm` binary given as input and return a
/// [`BindgenMetadata`] which extracts the custom sections describing
/// component-level types from within the binary itself.
///
/// This is used to parse the output of `wit-bindgen`-generated modules and is
/// one of the earliest phases in transitioning such a module to a component.
/// The extraction here provides the metadata necessary to continue the process
/// later on.
///
/// Note that a "stripped" binary where `component-type` sections are removed
/// is returned as well to embed within a component.
pub fn decode(wasm: &[u8]) -> Result<(Vec<u8>, BindgenMetadata)> {
let mut ret = BindgenMetadata::default();
let mut new_module = wasm_encoder::Module::new();
for payload in wasmparser::Parser::new(0).parse_all(wasm) {
let payload = payload.context("decoding item in module")?;
match payload {
wasmparser::Payload::CustomSection(cs) if cs.name().starts_with("component-type") => {
let data = BindgenMetadata::decode(cs.data())
.with_context(|| format!("decoding custom section {}", cs.name()))?;
ret.merge(data)
.with_context(|| format!("updating metadata for section {}", cs.name()))?;
}
_ => {
if let Some((id, range)) = payload.as_section() {
new_module.section(&wasm_encoder::RawSection {
id,
data: &wasm[range],
});
}
}
}
}
Ok((new_module.finish(), ret))
}
/// Creates a `component-type*` custom section to be decoded by `decode` above.
///
/// This is primarily created by wit-bindgen-based guest generators to embed
/// into the final core wasm binary. The core wasm binary is later fed
/// through `wit-component` to produce the actual component where this returned
/// section will be decoded.
pub fn encode(world: &World, encoding: StringEncoding) -> Vec<u8> {
let component = ComponentEncoder::default()
.types_only(true)
.world(world.clone(), encoding)
.unwrap()
.encode()
.unwrap();
let mut ret = Vec::new();
ret.push(CURRENT_VERSION);
ret.push(match encoding {
StringEncoding::UTF8 => 0x00,
StringEncoding::UTF16 => 0x01,
StringEncoding::CompactUTF16 => 0x02,
});
world.name.encode(&mut ret);
ret.extend(component);
ret
}
impl BindgenMetadata {
fn decode(data: &[u8]) -> Result<BindgenMetadata> {
let mut reader = BinaryReader::new(data);
let version = reader.read_u8()?;
if version != CURRENT_VERSION {
bail!("component-type version {version} does not match supported version {CURRENT_VERSION}");
}
let encoding = match reader.read_u8()? {
0x00 => StringEncoding::UTF8,
0x01 => StringEncoding::UTF16,
0x02 => StringEncoding::CompactUTF16,
byte => bail!("invalid string encoding {byte:#x}"),
};
let name = reader.read_string()?;
Ok(BindgenMetadata::new(
decode_world(name, &data[reader.original_position()..])?,
encoding,
))
}
/// Creates a new `BindgenMetadata` instance holding the given set of
/// interfaces which are expected to all use the `encoding` specified.
pub fn new(world: World, encoding: StringEncoding) -> BindgenMetadata {
let mut ret = BindgenMetadata {
world,
import_encodings: Default::default(),
export_encodings: Default::default(),
};
if let Some(iface) = &ret.world.default {
for func in iface.functions.iter() {
let name = func.core_export_name(None);
let prev = ret.export_encodings.insert(name.to_string(), encoding);
assert!(prev.is_none());
}
}
for (name, import) in ret.world.imports.iter() {
for func in import.functions.iter() {
let key = (name.clone(), func.name.clone());
let prev = ret.import_encodings.insert(key, encoding);
assert!(prev.is_none());
}
}
for (name, export) in ret.world.exports.iter() {
for func in export.functions.iter() {
let name = func.core_export_name(Some(name));
let prev = ret.export_encodings.insert(name.to_string(), encoding);
assert!(prev.is_none());
}
}
ret
}
/// Merges another `BindgenMetadata` into this one.
///
/// This operation is intended to be akin to "merging worlds" when the
/// abstraction level for that is what we're working at here. For now the
/// merge operation only succeeds if the two metadata descriptions are
/// entirely disjoint.
///
/// Note that at this time there's no support for changing string encodings
/// between metadata.
pub fn merge(&mut self, other: BindgenMetadata) -> Result<()> {
let BindgenMetadata {
world,
import_encodings,
export_encodings,
} = other;
// TODO: instead of returning an error here on duplicate interfaces
// this should merge the two interfaces. This probably requires world
// files to exist first though.
for (name, import) in world.imports {
let prev = self.world.imports.insert(name.clone(), import);
if prev.is_some() {
bail!("import interface `{name}` specified twice");
}
}
for (name, export) in world.exports {
let prev = self.world.exports.insert(name.clone(), export);
if prev.is_some() {
bail!("export interface `{name}` specified twice");
}
}
if let Some(default) = world.default {
if self.world.default.is_some() {
bail!("default export interface specified twice");
}
self.world.default = Some(default);
}
for (name, encoding) in export_encodings {
let prev = self.export_encodings.insert(name.clone(), encoding);
if let Some(prev) = prev {
if prev != encoding {
bail!("conflicting string encodings specified for export `{name}`");
}
}
}
for ((module, name), encoding) in import_encodings {
let prev = self
.import_encodings
.insert((module.clone(), name.clone()), encoding);
if let Some(prev) = prev {
if prev != encoding {
bail!("conflicting string encodings specified for import `{module}::{name}`");
}
}
}
Ok(())
}
}