| commit | 8974ba2518603710c9e602d6d8518b7494eadc09 | [log] [tgz] |
|---|---|---|
| author | adonovan <adonovan@google.com> | Sun Nov 22 19:45:14 2020 -0800 |
| committer | Copybara-Service <copybara-worker@google.com> | Sun Nov 22 19:47:20 2020 -0800 |
| tree | 32cca8cf4ae4c4f0102a90d4e1160451481eecba | |
| parent | 0881c80d29acecdfbb58c49156f805e8c50db117 [diff] |
starlark resolver: implement "flat globals" optimization This change reworks the Resolver.Module interface (which abstracts eval.Module) so that each global is statically mapped to a small integer, so that they can be represented during execution as a flat array, not a hash table, similar to the treatment of locals in CL 343371120. It also resolves PREDECLARED and UNIVERSE references distinctly, so that we needn't try both maps during execution. And it sets the stage for those maps to be treated as flat arrays too. Furthermore, it materializes for the first time a "file-local block", which will be used to hold bindings created by load statements, when we make loads bind locally and eliminate the getExportedGlobals hack. The Resolver.Module interface no longer requires the implementation to enumerate all the defined names; names are looked up on demand as references are encountered in the program during resolution, and Bindings are created for them. This reduces unnecessary allocation and copying when a small program is executed in a big environment. The resolver still provides spelling suggestions over all available names; it does this by having the Module implementation provide the set of top-level candidates lazily: only when resolve() fails. PiperOrigin-RevId: 343776166
{Fast, Correct} - Choose two
Build and test software of any size, quickly and reliably.
Speed up your builds and tests: Bazel rebuilds only what is necessary. With advanced local and distributed caching, optimized dependency analysis and parallel execution, you get fast and incremental builds.
One tool, multiple languages: Build and test Java, C++, Android, iOS, Go, and a wide variety of other language platforms. Bazel runs on Windows, macOS, and Linux.
Scalable: Bazel helps you scale your organization, codebase, and continuous integration solution. It handles codebases of any size, in multiple repositories or a huge monorepo.
Extensible to your needs: Easily add support for new languages and platforms with Bazel's familiar extension language. Share and re-use language rules written by the growing Bazel community.
Follow our tutorials:
See CONTRIBUTING.md
