ignore/index.js

// A simple implementation of make-array
function makeArray (subject) {
  return Array.isArray(subject)
    ? subject
    : [subject]
}

const REGEX_TEST_BLANK_LINE = /^\s+$/
const REGEX_REPLACE_LEADING_EXCAPED_EXCLAMATION = /^\\!/
const REGEX_REPLACE_LEADING_EXCAPED_HASH = /^\\#/
const REGEX_SPLITALL_CRLF = /\r?\n/g
// /foo,
// ./foo,
// ../foo,
// .
// ..
const REGEX_TEST_INVALID_PATH = /^\.*\/|^\.+$/

const SLASH = '/'
const KEY_IGNORE = typeof Symbol !== 'undefined'
  ? Symbol.for('node-ignore')
  /* istanbul ignore next */
  : 'node-ignore'

const define = (object, key, value) =>
  Object.defineProperty(object, key, {value})

const REGEX_REGEXP_RANGE = /([0-z])-([0-z])/g

// Sanitize the range of a regular expression
// The cases are complicated, see test cases for details
const sanitizeRange = range => range.replace(
  REGEX_REGEXP_RANGE,
  (match, from, to) => from.charCodeAt(0) <= to.charCodeAt(0)
    ? match
    // Invalid range (out of order) which is ok for gitignore rules but
    //   fatal for JavaScript regular expression, so eliminate it.
    : ''
)

// > If the pattern ends with a slash,
// > it is removed for the purpose of the following description,
// > but it would only find a match with a directory.
// > In other words, foo/ will match a directory foo and paths underneath it,
// > but will not match a regular file or a symbolic link foo
// >  (this is consistent with the way how pathspec works in general in Git).
// '`foo/`' will not match regular file '`foo`' or symbolic link '`foo`'
// -> ignore-rules will not deal with it, because it costs extra `fs.stat` call
//      you could use option `mark: true` with `glob`

// '`foo/`' should not continue with the '`..`'
const DEFAULT_REPLACER_PREFIX = [

  // > Trailing spaces are ignored unless they are quoted with backslash ("\")
  [
    // (a\ ) -> (a )
    // (a  ) -> (a)
    // (a \ ) -> (a  )
    /\\?\s+$/,
    match => match.indexOf('\\') === 0
      ? ' '
      : ''
  ],

  // replace (\ ) with ' '
  [
    /\\\s/g,
    () => ' '
  ],

  // Escape metacharacters
  // which is written down by users but means special for regular expressions.

  // > There are 12 characters with special meanings:
  // > - the backslash \,
  // > - the caret ^,
  // > - the dollar sign $,
  // > - the period or dot .,
  // > - the vertical bar or pipe symbol |,
  // > - the question mark ?,
  // > - the asterisk or star *,
  // > - the plus sign +,
  // > - the opening parenthesis (,
  // > - the closing parenthesis ),
  // > - and the opening square bracket [,
  // > - the opening curly brace {,
  // > These special characters are often called "metacharacters".
  [
    /[\\^$.|*+(){]/g,
    match => `\\${match}`
  ],

  [
    // > [abc] matches any character inside the brackets
    // >    (in this case a, b, or c);
    /\[([^\]/]*)($|\])/g,
    (match, p1, p2) => p2 === ']'
      ? `[${sanitizeRange(p1)}]`
      : `\\${match}`
  ],

  [
    // > a question mark (?) matches a single character
    /(?!\\)\?/g,
    () => '[^/]'
  ],

  // leading slash
  [

    // > A leading slash matches the beginning of the pathname.
    // > For example, "/*.c" matches "cat-file.c" but not "mozilla-sha1/sha1.c".
    // A leading slash matches the beginning of the pathname
    /^\//,
    () => '^'
  ],

  // replace special metacharacter slash after the leading slash
  [
    /\//g,
    () => '\\/'
  ],

  [
    // > A leading "**" followed by a slash means match in all directories.
    // > For example, "**/foo" matches file or directory "foo" anywhere,
    // > the same as pattern "foo".
    // > "**/foo/bar" matches file or directory "bar" anywhere that is directly
    // >   under directory "foo".
    // Notice that the '*'s have been replaced as '\\*'
    /^\^*\\\*\\\*\\\//,

    // '**/foo' <-> 'foo'
    () => '^(?:.*\\/)?'
  ]
]

const DEFAULT_REPLACER_SUFFIX = [
  // starting
  [
    // there will be no leading '/'
    //   (which has been replaced by section "leading slash")
    // If starts with '**', adding a '^' to the regular expression also works
    /^(?=[^^])/,
    function startingReplacer () {
      return !/\/(?!$)/.test(this)
        // > If the pattern does not contain a slash /,
        // >   Git treats it as a shell glob pattern
        // Actually, if there is only a trailing slash,
        //   git also treats it as a shell glob pattern
        ? '(?:^|\\/)'

        // > Otherwise, Git treats the pattern as a shell glob suitable for
        // >   consumption by fnmatch(3)
        : '^'
    }
  ],

  // two globstars
  [
    // Use lookahead assertions so that we could match more than one `'/**'`
    /\\\/\\\*\\\*(?=\\\/|$)/g,

    // Zero, one or several directories
    // should not use '*', or it will be replaced by the next replacer

    // Check if it is not the last `'/**'`
    (_, index, str) => index + 6 < str.length

      // case: /**/
      // > A slash followed by two consecutive asterisks then a slash matches
      // >   zero or more directories.
      // > For example, "a/**/b" matches "a/b", "a/x/b", "a/x/y/b" and so on.
      // '/**/'
      ? '(?:\\/[^\\/]+)*'

      // case: /**
      // > A trailing `"/**"` matches everything inside.

      // #21: everything inside but it should not include the current folder
      : '\\/.+'
  ],

  // intermediate wildcards
  [
    // Never replace escaped '*'
    // ignore rule '\*' will match the path '*'

    // 'abc.*/' -> go
    // 'abc.*'  -> skip this rule
    /(^|[^\\]+)\\\*(?=.+)/g,

    // '*.js' matches '.js'
    // '*.js' doesn't match 'abc'
    (_, p1) => `${p1}[^\\/]*`
  ],

  // trailing wildcard
  [
    /(\^|\\\/)?\\\*$/,
    (_, p1) => {
      const prefix = p1
        // '\^':
        // '/*' does not match ''
        // '/*' does not match everything

        // '\\\/':
        // 'abc/*' does not match 'abc/'
        ? `${p1}[^/]+`

        // 'a*' matches 'a'
        // 'a*' matches 'aa'
        : '[^/]*'

      return `${prefix}(?=$|\\/$)`
    }
  ],

  [
    // unescape
    /\\\\\\/g,
    () => '\\'
  ]
]

const POSITIVE_REPLACERS = [
  ...DEFAULT_REPLACER_PREFIX,

  // 'f'
  // matches
  // - /f(end)
  // - /f/
  // - (start)f(end)
  // - (start)f/
  // doesn't match
  // - oof
  // - foo
  // pseudo:
  // -> (^|/)f(/|$)

  // ending
  [
    // 'js' will not match 'js.'
    // 'ab' will not match 'abc'
    /(?:[^*/])$/,

    // 'js*' will not match 'a.js'
    // 'js/' will not match 'a.js'
    // 'js' will match 'a.js' and 'a.js/'
    match => `${match}(?=$|\\/)`
  ],

  ...DEFAULT_REPLACER_SUFFIX
]

const NEGATIVE_REPLACERS = [
  ...DEFAULT_REPLACER_PREFIX,

  // #24, #38
  // The MISSING rule of [gitignore docs](https://git-scm.com/docs/gitignore)
  // A negative pattern without a trailing wildcard should not
  // re-include the things inside that directory.

  // eg:
  // ['node_modules/*', '!node_modules']
  // should ignore `node_modules/a.js`
  [
    /(?:[^*])$/,
    match => `${match}(?=$|\\/$)`
  ],

  ...DEFAULT_REPLACER_SUFFIX
]

// A simple cache, because an ignore rule only has only one certain meaning
const regexCache = Object.create(null)

// @param {pattern}
const makeRegex = (pattern, negative, ignorecase) => {
  const r = regexCache[pattern]
  if (r) {
    return r
  }

  const replacers = negative
    ? NEGATIVE_REPLACERS
    : POSITIVE_REPLACERS

  const source = replacers.reduce(
    (prev, current) => prev.replace(current[0], current[1].bind(pattern)),
    pattern
  )

  return regexCache[pattern] = ignorecase
    ? new RegExp(source, 'i')
    : new RegExp(source)
}

const isString = subject => typeof subject === 'string'

// > A blank line matches no files, so it can serve as a separator for readability.
const checkPattern = pattern => pattern
  && isString(pattern)
  && !REGEX_TEST_BLANK_LINE.test(pattern)

  // > A line starting with # serves as a comment.
  && pattern.indexOf('#') !== 0

const splitPattern = pattern => pattern.split(REGEX_SPLITALL_CRLF)

class IgnoreRule {
  constructor (
    origin,
    pattern,
    negative,
    regex
  ) {
    this.origin = origin
    this.pattern = pattern
    this.negative = negative
    this.regex = regex
  }
}

const createRule = (pattern, ignorecase) => {
  const origin = pattern
  let negative = false

  // > An optional prefix "!" which negates the pattern;
  if (pattern.indexOf('!') === 0) {
    negative = true
    pattern = pattern.substr(1)
  }

  pattern = pattern
  // > Put a backslash ("\") in front of the first "!" for patterns that
  // >   begin with a literal "!", for example, `"\!important!.txt"`.
  .replace(REGEX_REPLACE_LEADING_EXCAPED_EXCLAMATION, '!')
  // > Put a backslash ("\") in front of the first hash for patterns that
  // >   begin with a hash.
  .replace(REGEX_REPLACE_LEADING_EXCAPED_HASH, '#')

  const regex = makeRegex(pattern, negative, ignorecase)

  return new IgnoreRule(
    origin,
    pattern,
    negative,
    regex
  )
}

const throwError = (message, Ctor) => {
  throw new Ctor(message)
}

const returnFalse = () => false

const checkPath = (path, doThrow) => {
  if (!isString(path)) {
    return doThrow(
      `path must be a string, but got \`${path}\``,
      TypeError
    )
  }

  // We don't know if we should ignore '', so throw
  if (!path) {
    return doThrow(`path must not be empty`, TypeError)
  }

  //
  if (REGEX_TEST_INVALID_PATH.test(path)) {
    const r = '`path.relative()`d'
    return doThrow(
      `path should be a ${r} string, but got "${path}"`,
      RangeError
    )
  }

  return true
}

class Ignore {
  constructor ({
    ignorecase = true
  } = {}) {
    this._rules = []
    this._ignorecase = ignorecase
    define(this, KEY_IGNORE, true)
    this._initCache()
  }

  _initCache () {
    this._ignoreCache = Object.create(null)
    this._testCache = Object.create(null)
  }

  _addPattern (pattern) {
    // #32
    if (pattern && pattern[KEY_IGNORE]) {
      this._rules = this._rules.concat(pattern._rules)
      this._added = true
      return
    }

    if (checkPattern(pattern)) {
      const rule = createRule(pattern, this._ignorecase)
      this._added = true
      this._rules.push(rule)
    }
  }

  // @param {Array<string> | string | Ignore} pattern
  add (pattern) {
    this._added = false

    makeArray(
      isString(pattern)
        ? splitPattern(pattern)
        : pattern
    ).forEach(this._addPattern, this)

    // Some rules have just added to the ignore,
    // making the behavior changed.
    if (this._added) {
      this._initCache()
    }

    return this
  }

  // legacy
  addPattern (pattern) {
    return this.add(pattern)
  }

  //          |           ignored : unignored
  // negative |   0:0   |   0:1   |   1:0   |   1:1
  // -------- | ------- | ------- | ------- | --------
  //     0    |  TEST   |  TEST   |  SKIP   |    X
  //     1    |  TESTIF |  SKIP   |  TEST   |    X

  // - SKIP: always skip
  // - TEST: always test
  // - TESTIF: only test if checkUnignored
  // - X: that never happen

  // @param {boolean} whether should check if the path is unignored,
  //   setting `checkUnignored` to `false` could reduce additional
  //   path matching.

  // @returns {TestResult} true if a file is ignored
  _testOne (path, checkUnignored) {
    let ignored = false
    let unignored = false

    this._rules.forEach(rule => {
      const {negative} = rule
      if (
        unignored === negative && ignored !== unignored
        || negative && !ignored && !unignored && !checkUnignored
      ) {
        return
      }

      const matched = rule.regex.test(path)

      if (matched) {
        ignored = !negative
        unignored = negative
      }
    })

    return {
      ignored,
      unignored
    }
  }

  // @returns {TestResult}
  _test (path, cache, checkUnignored, slices) {
    checkPath(path, throwError)

    if (path in cache) {
      return cache[path]
    }

    if (!slices) {
      // path/to/a.js
      // ['path', 'to', 'a.js']
      slices = path.split(SLASH)
    }

    slices.pop()

    // If the path has no parent directory, just test it
    if (!slices.length) {
      return cache[path] = this._testOne(path, checkUnignored)
    }

    const parent = this._test(
      slices.join(SLASH) + SLASH,
      cache,
      checkUnignored,
      slices
    )

    // If the path contains a parent directory, check the parent first
    return cache[path] = parent.ignored
      // > It is not possible to re-include a file if a parent directory of
      // >   that file is excluded.
      ? parent
      : this._testOne(path, checkUnignored)
  }

  ignores (path) {
    return this._test(path, this._ignoreCache, false).ignored
  }

  createFilter () {
    return path => !this.ignores(path)
  }

  filter (paths) {
    return makeArray(paths).filter(this.createFilter())
  }

  // @returns {TestResult}
  test (path) {
    return this._test(path, this._testCache, true)
  }
}

// Windows
// --------------------------------------------------------------
/* istanbul ignore if  */
if (
  // Detect `process` so that it can run in browsers.
  typeof process !== 'undefined'
  && (
    process.env && process.env.IGNORE_TEST_WIN32
    || process.platform === 'win32'
  )
) {
  const test = Ignore.prototype._test

  /* eslint no-control-regex: "off" */
  const make_posix = str => /^\\\\\?\\/.test(str)
  || /[^\x00-\x80]+/.test(str)
    ? str
    : str.replace(/\\/g, '/')

  Ignore.prototype._test = function testWin32 (path, ...args) {
    path = make_posix(path)
    return test.call(this, path, ...args)
  }
}

const factory = options => new Ignore(options)

factory.isPathValid = path => checkPath(path, returnFalse)

module.exports = factory