Provides an easy-to-use API for parsing users' js/ts/tsx files using their project's installed Babel version and config.
This way it won't fail if they're using funky features like the smart pipeline operator. This is a big problem for codemod
tools; for example, jscodeshift would choke on the smart pipeline operator unless you pass a custom parser. I want my
codemod tools to just work.
Note: only Babel 7 is currently supported.
babel-parse-wild-code also improves performance. Until recently, doing require('@babel/core').loadOptionsAsync and passing that config
to a bunch of require('@babel/core').parseAsync calls didn't improve performance as expected; parseAsync was accidentally re-doing
a lot of the work that had already been done by loadOptionsAsync. That bug has been fixed since I reported it, but babel-parse-wild-code
works around this for older versions of Babel 7 by extracting the options for @babel/parser from the user's Babel config.
If babel-parse-wild-code fails to load @babel/core, @babel/parser, or the Babel config from the user's
project, it falls back to parsing with reasonable default options.
import { parseSync } from 'babel-parse-wild-code'Parses the given file synchronously, returning the File node.
encoding defaults to utf8. The remaining options are passed to @babel/parser's parse function.
options.plugins will be merged with the resolved options for the file.
import { parseAsync } from 'babel-parse-wild-code'Parses the given file asynchronously, returning a Promise that will resolve to the File node.
encoding defaults to utf8. The remaining options are passed to @babel/parser's parse function.
options.plugins will be merged with the resolved options for the file.
Instances of @babel/core, @babel/parser and parser options are cached on a per-directory basis.
Calling clearCache() clears this cache, and deletes instances of @babel/core and @babel/parser
from require.cache.
You should probably do this before any bulk parsing operation. It would be nice to bust the cache automatically when the user's Babel version or config changes, but setting up the watchers would be complicated. Clearing the cache before you parse a bunch of files is simpler and won't have a huge impact on performance.
import { getParserSync } from 'babel-parse-wild-code'Gets a fully-configured parser for the given file synchronously.
options is additional options for @babel/parser's parse function. For example when working
with jscodeshift or recast, you should pass { tokens: true }.
options.plugins will be merged with the resolved options for the file.
import { getParserSync } from 'babel-parse-wild-code'Gets a fully-configured parser for the given file asynchronously.
options is additional options for @babel/parser's parse function. For example when working
with jscodeshift or recast, you should pass { tokens: true }.
options.plugins will be merged with the resolved options for the file.
import { Parser } from 'babel-parse-wild-code'Type defs:
import * as defaultBabelParser from '@babel/parser'
type BabelParser = Pick<typeof defaultBabelParser, 'parse' | 'parseExpression'>
export class Parser {
readonly babelParser: BabelParser
readonly parserOpts: ParserOptions
constructor(babelParser: BabelParser, parserOpts: ParserOptions)
parse(code: string, parserOpts?: ParserOptions): t.File
parseExpression(code: string, parserOpts?: ParserOptions): t.Expression
bindParserOpts(parserOpts: ParserOptions): Parser
}import { tsParser } from 'babel-parse-wild-code'The fallback parser used for .ts files.
import { tsxParser } from 'babel-parse-wild-code'The fallback parser used for .tsx files.
import { jsParser } from 'babel-parse-wild-code'The fallback parser used for parsing .js files when babel-parse-wild-code failed to load Babel modules or config for the file's directory.
Note: this is the same as jsxParser, but jsxParser is exported for completeness.
import { jsxParser } from 'babel-parse-wild-code'The fallback parser used for parsing .jsx files when babel-parse-wild-code failed to load Babel modules or config for the file's directory.