Description
With Biome 2, we introduced the scanner, which is used for finding nested biome.json files as well as .gitignore files if configured (both these are handled by the "light scanner"). If any rules from the project domain are enabled, we enable the "full scanner" which indexes the entire project, including node_modules, in order to populate the module graph and extract type information.
Whenever only the "light scanner" is enabled and we are running on the CLI and we are given an exact list of files to process, we should be able to perform an even more targeted version of the light scan. Instead of scanning "wide" from the project root, we can start from the files to process and only scan upwards to the project root. That should give us all the information we need for the invocation, and speed up use cases such as Git hooks.
Description
With Biome 2, we introduced the scanner, which is used for finding nested
biome.jsonfiles as well as.gitignorefiles if configured (both these are handled by the "light scanner"). If any rules from the project domain are enabled, we enable the "full scanner" which indexes the entire project, includingnode_modules, in order to populate the module graph and extract type information.Whenever only the "light scanner" is enabled and we are running on the CLI and we are given an exact list of files to process, we should be able to perform an even more targeted version of the light scan. Instead of scanning "wide" from the project root, we can start from the files to process and only scan upwards to the project root. That should give us all the information we need for the invocation, and speed up use cases such as Git hooks.