cspell

A Spell Checker for Code!

cspell is a command line tool and library for spell checking code.

• Spell Checks Code -- Able to spell check code by parsing it into words before checking against the dictionaries.
  • Supports CamelCase, snake_case, and compoundwords naming styles.
• Self contained -- does not depend upon OS libraries like Hunspell or aspell. Nor does it depend upon online services.
• Fast -- checks 1000's of lines of code in seconds.
• Programming Language Specific Dictionaries -- Has dedicated support for:
  • JavaScript, TypeScript, Python, PHP, C#, C++, LaTex, Go, HTML, CSS, etc.
• Customizable -- supports custom dictionaries and word lists.
• Continuous Integration Support -- Can easily be added as a linter to Travis-CI.

cspell was initially built as the spell checking service for the spell checker extension for Visual Studio Code.

Available as part of the Tidelift Subscription.

The maintainers of cspell and thousands of other packages are working with Tidelift to deliver commercial support and maintenance for the open source packages you use to build your applications. Save time, reduce risk, and improve code health, while paying the maintainers of the exact packages you use. Learn more.

npm install -g cspell

There is a docker image at Packages for cspell.

cspell commands below can be run by replacing

cspell

with:

docker run -it -v $PWD:/workdir ghcr.io/streetsidesoftware/cspell:latest

Example:

docker run -it -v $PWD:/workdir ghcr.io/streetsidesoftware/cspell:latest lint --help

Example: recursively spell check all JavaScript files in src

JavaScript files

cspell "src/**/*.js"
# or
cspell lint "src/**/*.js"

Check everything

cspell "**"

Git: Check Only Changed Files

 git diff --name-only | npx cspell --file-list stdin

The lint command is used for spell checking files.

cspell lint --help

Usage: cspell lint [options] [globs...] [file:https://<path> ...] [stdin[:https://<path>]]

Patterns:
 - [globs...]            Glob Patterns
 - [stdin]               Read from "stdin" assume text file.
 - [stdin:https://<path>]      Read from "stdin", use <path> for file type and config.
 - [file:https://<path>]       Check the file at <path>

Examples:
    cspell .                        Recursively check all files.
    cspell lint .                   The same as "cspell ."
    cspell "*.js"                   Check all .js files in the current directory
    cspell "**/*.js"                Check all .js files recursively
    cspell "src/**/*.js"            Only check .js under src
    cspell "**/*.txt" "**/*.js"     Check both .js and .txt files.
    cspell "**/*.{txt,js,md}"       Check .txt, .js, and .md files.
    cat LICENSE | cspell stdin      Check stdin
    cspell stdin:https://docs/doc.md      Check stdin as if it was "./docs/doc.md"

Check spelling

Options:
  -c, --config <cspell.json>   Configuration file to use.  By default cspell
                               looks for cspell.json in the current directory.
  -v, --verbose                Display more information about the files being
                               checked and the configuration.
  --locale <locale>            Set language locales. i.e. "en,fr" for English
                               and French, or "en-GB" for British English.
  --language-id <file-type>    Force programming language for unknown
                               extensions. i.e. "php" or "scala"
  --words-only                 Only output the words not found in the
                               dictionaries.
  -u, --unique                 Only output the first instance of a word not
                               found in the dictionaries.
  -e, --exclude <glob>         Exclude files matching the glob pattern. This
                               option can be used multiple times to add
                               multiple globs.
  --file-list <path or stdin>  Specify a list of files to be spell checked. The
                               list is filtered against the glob file patterns.
                               Note: the format is 1 file path per line.
  --file [file...]             Specify files to spell check. They are filtered
                               by the [globs...].
  --no-issues                  Do not show the spelling errors.
  --no-progress                Turn off progress messages
  --no-summary                 Turn off summary message in console.
  -s, --silent                 Silent mode, suppress error messages.
  --no-exit-code               Do not return an exit code if issues are found.
  --quiet                      Only show spelling issues or errors.
  --fail-fast                  Exit after first file with an issue or error.
  -r, --root <root folder>     Root directory, defaults to current directory.
  --no-relative                Issues are displayed with absolute path instead
                               of relative to the root.
  --show-context               Show the surrounding text around an issue.
  --show-suggestions           Show spelling suggestions.
  --no-show-suggestions        Do not show spelling suggestions or fixes.
  --no-must-find-files         Do not error if no files are found.
  --cache                      Use cache to only check changed files.
  --no-cache                   Do not use cache.
  --cache-reset                Reset the cache file.
  --cache-strategy <strategy>  Strategy to use for detecting changed files.
                               (choices: "metadata", "content")
  --cache-location <path>      Path to the cache file or directory. (default:
                               ".cspellcache")
  --dot                        Include files and directories starting with `.`
                               (period) when matching globs.
  --gitignore                  Ignore files matching glob patterns found in
                               .gitignore files.
  --no-gitignore               Do NOT use .gitignore files.
  --gitignore-root <path>      Prevent searching for .gitignore files past
                               root.
  --validate-directives        Validate in-document CSpell directives.
  --no-color                   Turn off color.
  --color                      Force color.
  --no-default-configuration   Do not load the default configuration and
                               dictionaries.
  --debug                      Output information useful for debugging
                               cspell.json files.
  --reporter <module|path>     Specify one or more reporters to use.
  --issue-template [template]  Use a custom issue template. See --help
                               --issue-template for details.
  -h, --help                   display help for command

More Examples:

    cspell "**/*.js" --reporter @cspell/cspell-json-reporter
        This will spell check all ".js" files recursively and use
        "@cspell/cspell-json-reporter".

    cspell . --reporter default
        This will force the default reporter to be used overriding
        any reporters defined in the configuration.

    cspell . --reporter ./<path>/reporter.cjs
        Use a custom reporter. See API for details.

    cspell "*.md" --exclude CHANGELOG.md --files README.md CHANGELOG.md
        Spell check only check "README.md" but NOT "CHANGELOG.md".

    cspell "/*.md" --no-must-find-files --files $FILES
        Only spell check the "/*.md" files in $FILES,
        where $FILES is a shell variable that contains the list of files.

References:
    https://cspell.org
    https://github.com/streetsidesoftware/cspell

Do a quick visual check of a file. This is a great way to see which text is included in the check.

cspell check <filename>

It will produce something like this:

To get color in less, use --color and less -r

cspell check <filename> --color | less -r

Trace shows a the list of known dictionaries and a * next to the ones that contain the word.

A ! will appear next to the ones where the word is forbidden.

Usage: cspell trace [options] [words...]

Trace words -- Search for words in the configuration and dictionaries.

Options:
  -c, --config <cspell.json>  Configuration file to use.  By default cspell
                              looks for cspell.json in the current directory.
  --locale <locale>           Set language locales. i.e. "en,fr" for English
                              and French, or "en-GB" for British English.
  --language-id <language>    Use programming language. i.e. "php" or "scala".
  --allow-compound-words      Turn on allowCompoundWords
  --no-allow-compound-words   Turn off allowCompoundWords
  --ignore-case               Ignore case and accents when searching for words.
  --no-ignore-case            Do not ignore case and accents when searching for
                              words.
  --dictionary-path <format>  Configure how to display the dictionary path.
                              (choices: "hide", "short", "long", "full",
                              default: Display most of the path.)
  --stdin                     Read words from stdin.
  --all                       Show all dictionaries.
  --only-found                Show only dictionaries that have the words.
  --no-color                  Turn off color.
  --color                     Force color
  --no-default-configuration  Do not load the default configuration and
                              dictionaries.
  -h, --help                  display help for command

Mega-Linter aggregates 70 linters ready to use out of the box, including cspell

• Can run as a GitHub Action, on other CI tools and locally
• Provides an updated .cspell.json file with new unknown words

Setup

Quick setup following installation guide in Mega-Linter documentation

Setup

npm install -SD cspell

.git/hooks/pre-commit

#!/bin/sh

exec git diff --cached --name-only | npx cspell --no-summary --no-progress --no-must-find-files --file-list stdin

cspell needs Node 14 and above.

The concept is simple, split camelCase and snake_case words before checking them against a list of known words.

• camelCase -> camel case
• HTMLInput -> html input
• srcCode -> src code
• snake_case_words -> snake case words
• camel2snake -> camel snake -- (the 2 is ignored)
• function parseJson(text: string) -> function parse json text string

• Escape characters like \n, \t are removed if the word does not match:
  • \narrow -> narrow - because narrow is a word
  • \ncode -> code - because ncode is not a word.
  • \network -> network - but it might be hiding a spelling error, if \n was an escape character.

• This spellchecker is case insensitive. It will not catch errors like english which should be English.
• The spellchecker uses dictionaries stored locally. It does not send anything outside your machine.
• The words in the dictionaries can and do contain errors.
• There are missing words.
• Only words longer than 3 characters are checked. "jsj" is ok, while "jsja" is not.
• All symbols and punctuation are ignored.

It is possible to add spell check settings into your source code. This is to help with file specific issues that may not be applicable to the entire project.

All settings are prefixed with cspell: or spell-checker:.

• disable -- turn off the spell checker for a section of code.
• enable -- turn the spell checker back on after it has been turned off.
• ignore -- specify a list of words to be ignored.
• words -- specify a list of words to be considered correct and will appear in the suggestions list.
• ignoreRegExp -- Any text matching the regular expression will NOT be checked for spelling.
• includeRegExp -- Only text matching the collection of includeRegExp will be checked.
• enableCompoundWords / disableCompoundWords -- Allow / disallow words like: "stringlength".
• dictionaries -- specify a list of the names of the dictionaries to use.

It is possible to disable / enable the spell checker by adding comments to your code.

• /* cspell:disable */
• /* spell-checker: disable */
• /* spellchecker: disable */
• // cspell:disable-line -- disables checking for the current line.
• /* cspell:disable-next-line */ -- disables checking till the end of the next line.

• /* cspell:enable */
• /* spell-checker: enable */
• /* spellchecker: enable */

// cspell:disable
const wackyWord = ['zaallano', 'wooorrdd', 'zzooommmmmmmm'];
/* cspell:enable */

const words = ['zaallano', 'wooorrdd', 'zzooommmmmmmm']; // cspell:disable-line disables this entire line

// To disable the next line, use cspell:disable-next-line
const moreWords = ['ieeees', 'beees', 'treeees'];

// Nesting disable / enable is not Supported

// spell-checker:disable
// It is now disabled.

var liep = 1;

/* cspell:disable */
// It is still disabled

// cspell:enable
// It is now enabled

const str = 'goededag'; // <- will be flagged as an error.

// spell-checker:enable <- doesn't do anything

// cSPELL:DISABLE <-- also works.

// if there isn't an enable, spelling is disabled till the end of the file.
const str = 'goedemorgen'; // <- will NOT be flagged as an error.

Ignore allows you the specify a list of words you want to ignore within the document.

// cspell:ignore zaallano, wooorrdd
// cspell:ignore zzooommmmmmmm
const wackyWord = ['zaallano', 'wooorrdd', 'zzooommmmmmmm'];

Note: words defined with ignore will be ignored for the entire file.

The words list allows you to add words that will be considered correct and will be used as suggestions.

// cspell:words woorxs sweeetbeat
const companyName = 'woorxs sweeetbeat';

Note: words defined with words will be used for the entire file.

In some programming language it is common to glue words together.

// cspell:enableCompoundWords
char * errormessage;  // Is ok with cspell:enableCompoundWords
int    errornumber;   // Is also ok.

Note: Compound word checking cannot be turned on / off in the same file. The last setting in the file determines the value for the entire file.

By default, the entire document is checked for spelling. cspell:disable/cspell:enable above allows you to block off sections of the document. ignoreRegExp and includeRegExp give you the ability to ignore or include patterns of text. By default the flags gim are added if no flags are given.

The spell checker works in the following way:

1. Find all text matching includeRegExp
2. Remove any text matching ignoreRegExp
3. Check the remaining text.

// cspell:ignoreRegExp 0x[0-9a-f]+     -- will ignore c style hex numbers
// cspell:ignoreRegExp /0x[0-9A-F]+/g  -- will ignore upper case c style hex numbers.
// cspell:ignoreRegExp g{5} h{5}       -- will only match ggggg, but not hhhhh or 'ggggg hhhhh'
// cspell:ignoreRegExp g{5}|h{5}       -- will match both ggggg and hhhhh
// cspell:ignoreRegExp /g{5} h{5}/     -- will match 'ggggg hhhhh'
/* cspell:ignoreRegExp /n{5}/          -- will NOT work as expected because of the ending comment -> */
/*
   cspell:ignoreRegExp /q{5}/          -- will match qqqqq just fine but NOT QQQQQ
*/
// cspell:ignoreRegExp /[^\s]{40,}/    -- will ignore long strings with no spaces.
// cspell:ignoreRegExp Email           -- this will ignore email like patterns -- see Predefined RegExp expressions
var encodedImage = 'HR+cPzr7XGAOJNurPL0G8I2kU0UhKcqFssoKvFTR7z0T3VJfK37vS025uKroHfJ9nA6WWbHZ/ASn...';
var email1 = '[email protected]';
var email2 = '<[email protected]>';

Note: ignoreRegExp and includeRegExp are applied to the entire file. They do not start and stop.

In general you should not need to use includeRegExp. But if you are mixing languages then it could come in helpful.

# cspell:includeRegExp #.*
# cspell:includeRegExp ("""|''')[^\1]*\1
# only comments and block strings will be checked for spelling.
def sum_it(self, seq):
    """This is checked for spelling"""
    variabele = 0
    alinea = 'this is not checked'
    for num in seq:
        # The local state of 'value' will be retained between iterations
        variabele += num
        yield variabele

The dictionaries list allows you to specify dictionaries to use for the file.

// cspell:dictionaries lorem-ipsum
const companyName = 'Lorem ipsum dolor sit amet';

Note: dictionaries specified with dictionaries will be used for the entire file.

• Urls¹ -- Matches urls
• HexValues -- Matches common hex format like #aaa, 0xfeef, \u0134
• Base64¹ -- matches base64 blocks of text longer than 40 characters.
• Email -- matches most email addresses.

• Everything¹ -- By default we match an entire document and remove the excludes.
• string -- This matches common string formats like '...', "...", and `...`
• CStyleComment -- These are C Style comments /* */ and //
• PhpHereDoc -- This matches PHPHereDoc strings.

^1. These patterns are part of the default include/exclude list for every file.

cspell's behavior can be controlled through a config file. By default it looks for any of the following files:

• .cspell.json
• cspell.json
• .cSpell.json
• cSpell.json
• cspell.config.js
• cspell.config.cjs
• cspell.config.json
• cspell.config.yaml
• cspell.config.yml
• cspell.yaml
• cspell.yml

Or you can specify a path to a config file with the --config <path> argument on the command line.

// cSpell Settings
{
    // Version of the setting file.  Always 0.2
    "version": "0.2",
    // language - current active spelling language
    "language": "en",
    // words - list of words to be always considered correct
    "words": [
        "mkdirp",
        "tsmerge",
        "githubusercontent",
        "streetsidesoftware",
        "vsmarketplacebadge",
        "visualstudio"
    ],
    // flagWords - list of words to be always considered incorrect
    // This is useful for offensive words and common spelling errors.
    // For example "hte" should be "the"
    "flagWords": [
        "hte"
    ]
}

• version - currently always 0.2 - controls how the settings in the configuration file behave.

• language - this specifies the language locale to use in choosing the general dictionary. For example: "language": "en-GB" tells cspell to use British English instead of US English.

• words - a list of words to be considered correct.

• flagWords - a list of words to be always considered incorrect

• ignoreWords - a list of words to be ignored (even if they are in the flagWords).

• ignorePaths - a list of globs to specify which files are to be ignored.

  Example

  "ignorePaths": ["node_modules/**"]

  will cause cspell to ignore anything in the node_modules directory.

• maxNumberOfProblems - defaults to 100 per file.

• minWordLength - defaults to 4 - the minimum length of a word before it is checked.

• allowCompoundWords - defaults to false; set to true to allow compound words by default.

• dictionaries - list of the names of the dictionaries to use. See Dictionaries below.

• dictionaryDefinitions - this list defines any custom dictionaries to use. This is how you can include other languages like Spanish.

  Example

  "language": "en",
  // Dictionaries "spanish", "ruby", and "corp-term" will always be checked.
  // Including "spanish" in the list of dictionaries means both Spanish and English
  // words will be considered correct.
  "dictionaries": ["spanish", "ruby", "corp-terms", "fonts"],
  // Define each dictionary.  Relative paths are relative to the config file.
  "dictionaryDefinitions": [
      { "name": "spanish", "path": "./spanish-words.txt"},
      { "name": "ruby", "path": "./ruby.txt"},
      { "name": "company-terms", "path": "./corp-terms.txt"}
  ],

• ignoreRegExpList - list of patterns to be ignored

• includeRegExpList - (Advanced) limits the text checked to be only that matching the expressions in the list.

• patterns - this allows you to define named patterns to be used with ignoreRegExpList and includeRegExpList.

• languageSettings - this allow for per programming language configuration settings. See LanguageSettings

The spell checker includes a set of default dictionaries.

• en_US - Derived from Hunspell US English words.
• en-gb - Derived from Hunspell GB English words.
• companies - List of well known companies
• softwareTerms - Software Terms and concepts like "coroutine", "debounce", "tree", etc.
• misc - Terms that do not belong in the other dictionaries.

• typescript - keywords for TypeScript and JavaScript
• node - terms related to using nodejs.
• php - php keywords and library methods
• go - go keywords and library methods
• python - python keywords
• powershell - powershell keywords
• html - html related keywords
• css - css, less, and scss related keywords
• cpp - C++ related keywords
• csharp - C# related keywords
• latex - LaTex related words
• bash - Bash/shell script keywords

• fonts - long list of fonts - to assist with css
• filetypes - list of file types
• npm - list of top 500+ package names on npm.

• name - The reference name of the dictionary, used with program language settings
• description - Optional description
• path - Path to the file, can be relative or absolute. Relative path is relative to the current cspell.json file.
• repMap - Optional replacement map use to replace character prior to searching the dictionary. Example:

      // Replace various tick marks with a single '
      "repMap": [["'|`|’", "'"]]

  // Use Compounds
• useCompounds - allow compound words

// Define each dictionary.  Relative paths are relative to the config file.
"dictionaryDefinitions": [
    { "name": "spanish", "path": "./spanish-words.txt"},
    { "name": "ruby", "path": "./ruby.txt"},
    { "name": "company-terms", "path": "./corp-terms.txt"}
],

It is possible to prevent a dictionary from being loaded. This is useful if you want to use your own dictionary or just turn off an existing dictionary.

"dictionaries": ["!cpp"],
"overrides": [
  {
      "filename": "legacy/**/*.cpp",
      "dictionaries": ["!!cpp"], // add it back for *.cpp files under the legacy folder
  },
]

The number of !s is important.

• !cpp remove cpp dictionary
• !!cpp add it back
• !!!cpp remove it again.

The Language Settings allow configuration to be based upon the programming language and/or the locale. There are two selector fields locale and languageId.

• languageId defines which programming languages to match against. A value of "python,javascript" will match against python and javascript files. To match against ALL programming languages, use "*".
• locale defines which spoken languages to match against. A value of "en-GB,nl" will match against British English or Dutch. A value of "*" will match all spoken languages.
• Most configuration values allowed in a cspell.json file can be define or redefine within the languageSettings.

    "languageSettings": [
        {
            // VSCode languageId. i.e. typescript, java, go, cpp, javascript, markdown, latex
            // * will match against any file type.
            "languageId": "c,cpp",
            // Language locale. i.e. en-US, de-AT, or ru. * will match all locales.
            // Multiple locales can be specified like: "en, en-US" to match both English and English US.
            "locale": "*",
            // To exclude patterns, add them to "ignoreRegExpList"
            "ignoreRegExpList": [
                "/#include.*/"
            ],
            // List of dictionaries to enable by name in `dictionaryDefinitions`
            "dictionaries": ["cpp"],
            // Dictionary definitions can also be supplied here. They are only used iff "languageId" and "locale" match.
            "dictionaryDefinitions": []
        }
    ]

Overrides are useful for forcing configuration on a per file basis.

Example:

    "overrides": [
        // Force `*.hrr` and `*.crr` files to be treated as `cpp` files:
        {
            "filename": "**/{*.hrr,*.crr}",
            "languageId": "cpp"
        },
        // Force `*.txt` to use the Dutch dictionary (Dutch dictionary needs to be installed separately):
        {
            "language": "nl",
            "filename": "**/dutch/**/*.txt"
        }
    ]

Brought to you by Street Side Software