Our npm packages

We have been publishing to npm for 10.8 years now: 156 packages in that time, roughly 14.3 a year, of which 138 are current and 18 have been retired. The oldest two are still here and still maintained — detergent and object-boolean-combinations, both from 2015. At the other end sit html-entity-codec and string-typo-match, both from 2026.

Our speciality is non-parsing code processing — algorithms which read the source as a string and never build a syntax tree, so broken and templated input stays workable. It is not an HTML-only habit: html-crush applies it to HTML, email-comb to CSS, json-comb to JSON, and csv-sort to the awkward CSV that banks export.

The shapes vary more than the subject matter suggests. That same scannerless approach turns up over Sass variables, package.json, TypeScript declarations, CSV, IANA language codes, Commitizen changelogs and plain English prose. The APIs run from classes ( ranges-push) through callback-only traversal ( ast-monkey-traverse-with-lookahead) to a dozen command-line apps ( csv-sort-cli), and there is a tokeniser and a parser producing our own AST design, a monkey to walk it, a linter built on Node’s EventEmitter ( emlint), a Gulp plugin and an ESLint one.

Then the oddities: a string uglifier built on cryptographic principles ( string-uglify), a program which reverse-engineers Bitbucket’s slug algorithm ( bitbucket-slug), one which generates all 2ⁿ boolean combinations of an options object so the unit tests can cover every one of them ( test-mixer), and a mock program whose only job is to normalise everybody else’s benchmark scores ( perf-ref).

Everything below is grouped by what it operates on: HTML and CSS, ranges, strings, plain objects and arrays, ASTs, monorepo tooling, the command line and everything that fits nowhere else — with the four flagship libraries up front.

Of the 110 packages in the monorepo, 22 (20%) declare no dependencies at all, 90 (82%) never reach outside Codsen code however deep you follow them, and another 9 bottom out at exactly one outside library. The lot pulls in 18 distinct third-party packages, while our own codsen-utils is used by 77 of them.

Every one of those 110 is gated at 100% line coverage, and 32 of them at 100% branches, functions and statements as well — a package which slips below its threshold fails the build instead of shipping. Benchmarks are checked in for 99 of them, 1,003 recorded runs so far, and a release more than 10% slower than its own baseline fails the build too. There are 852 runnable examples across 101 of the packages, all automatically tested, and 10,093 commits behind the lot.

Downloads and dependencies

See how downloads are shared across our packages and how the libraries fit together. The download window is printed on each chart.

Counts include dependency and automated installs. They measure package downloads, not unique people. Open the download chart at full size.

Follow the connections between Codsen packages

Arrows lead from a package to its dependency. Each package name links to its documentation. Dashed boxes mark packages whose outgoing dependencies are unknown. This map shows only Codsen dependencies.

Open the dependency map at full size to zoom into the labels on a small screen.

Explore one package’s dependencies in 3D

Choose a package, then rotate or zoom its dependency graph. This snapshot includes resolved production and optional dependencies, including third-party packages; a fresh install can resolve different versions.

Open the dependency explorer on its own page. The explorer also includes a linked text list of the dependencies.

View all charts, rankings and download history or inspect the chart data.

Flagship Libraries

string-strip-html ^13.7.3

Strip HTML tags from strings. No parser, accepts mixed sources.

open playground

email-comb ^7.4.3

Remove unused CSS from email templates

open playground

html-crush ^6.3.3

Minify email templates

open playground

detergent ^9.4.7

Extract, clean, encode text and fix English style

open playground

↑ back to top

HTML Processing Libraries ^Σ=16

They all process HTML and CSS. Our speciality is non-parsing algorithms — we don’t parse and then work on AST and then render. We aim to work on the source code directly while traversing it as a string.

The idea is, if you don’t parse the HTML, you can support broken or mixed code. Unless you write your parser, it becomes a bottleneck — parser throws here and there, and you can do nothing about it.

It is vital to support broken code because this allows us to make broken code-fixing programs.

Also, we often aim to support the processing of HTML mixed with other, known or unknown, source code (typically, programming or templating languages).

all-named-html-entities ^3.2.1

List of all named HTML entities

charcode-is-valid-xml-name-character ^3.2.2

Does a given character belong to XML spec’s “Production 4 OR 4a” type (is acceptable for XML element’s name)

color-shorthand-hex-to-six-digit ^5.2.4

Convert shorthand hex color codes into full

detect-is-it-html-or-xhtml ^6.2.4

Does the string resemble an HTML or XHTML (or neither)?

generate-atomic-css ^3.2.4

Generate Atomic CSS

html-all-known-attributes ^6.2.1

All HTML attributes known to the Humanity

html-entities-not-email-friendly ^0.10.1

All HTML entities which are not email template friendly

html-entity-codec ^1.1.0

Decode, encode, and escape HTML character references

html-img-alt ^4.2.5

Adds missing alt attributes to img tags. Non-parsing.

html-table-patcher ^6.2.5

Visual helper to place templating code around table tags into correct places

is-char-suitable-for-html-attr-name ^4.2.3

Is given character suitable to be in an HTML attribute’s name?

is-html-attribute-closing ^4.2.5

Is a character on a given index a closing of an HTML attribute?

is-html-tag-opening ^4.2.5

Does an HTML tag start at given position?

is-language-code ^5.2.2

Is given string a language code (as per IANA)

open playground

is-media-descriptor ^5.2.5

Is given string a valid media descriptor (including media query)?

is-relative-uri ^5.2.1

Is given string a relative URI?

↑ back to top

Range Libraries ^Σ=12

We needed a way to compose multiple string amendment operations. We started to mark each operation as an array, for example, [1, 5] means “delete from string index 1 to 5”.

If you wish to put something after deletion as well, it goes as the third element in the array.

All these packages process arrays of range arrays in various ways.

ranges-apply ^7.2.5

Take an array of string index ranges, delete/replace the string according to them

ranges-crop ^6.2.5

Crop array of ranges when they go beyond the reference string’s length

ranges-ent-decode ^6.2.5

Recursive HTML entity decoding for Ranges workflow

ranges-invert ^6.2.5

Invert string index ranges

ranges-is-index-within ^4.2.1

Checks if index is within any of the given string index ranges

ranges-iterate ^4.2.4

Iterate a string and any changes within given string index ranges

ranges-merge ^9.2.5

Merge and sort string index ranges

ranges-process-outside ^6.2.5

Iterate string considering ranges, as if they were already applied

ranges-push ^7.3.2

Gather string index ranges

ranges-regex ^6.2.5

Integrate regex operations into Ranges workflow

ranges-sort ^6.2.4

Sort string index ranges

string-range-expander ^4.2.5

Expands string index ranges within whitespace boundaries until letters are met

↑ back to top

String Processing Libraries ^Σ=27

They process string inputs, which might be text, code or something else as long as it is of a string-type.

arrayiffy-if-string ^5.2.3

Put non-empty strings into arrays, turn empty-ones into empty arrays. Bypass everything else.

edit-package-json ^0.10.6

Edit package.json without parsing, as string, to keep the formatting intact

email-all-chars-within-ascii ^5.2.5

Scans all characters within a string and checks are they within ASCII range

js-row-num ^7.2.5

Update all row numbers in all console.logs in JS code

str-indexes-of-plus ^5.2.2

Like indexOf but returns array and counts per-grapheme

string-apostrophes ^4.2.6

Comprehensive, HTML-entities-aware tool to typographically-correct the apostrophes and single/double quotes

string-character-is-astral-surrogate ^3.2.2

Tells, is given character a part of astral character, specifically, a high and low surrogate

string-collapse-leading-whitespace ^7.3.1

Collapse the leading and trailing whitespace of a string

string-collapse-white-space ^11.2.6

Replace chunks of whitespace with a single spaces

string-convert-indexes ^6.2.5

Convert between native JS string character indexes and grapheme-count-based indexes

string-dashes ^1.4.6

Comprehensive, HTML-entities-aware tool to typographically-correct the dashes and hyphens

string-extract-class-names ^8.4.0

Extracts class and ID names from isolated CSS selector fragments

string-extract-sass-vars ^4.2.4

Parse SASS variables file into a plain object of CSS key-value pairs

string-find-heads-tails ^6.2.5

Finds where are arbitrary templating marker heads and tails located

string-find-malformed ^4.2.5

Search for a malformed string. Think of Levenshtein distance but in search.

string-fix-broken-named-entities ^7.2.5

Finds and fixes common and not so common broken named HTML entities, returns ranges array of fixes

string-left-right ^6.2.4

Looks up the first non-whitespace character to the left/right of a given index

string-match-left-right ^9.2.5

Match substrings on the left or right of a given index, ignoring whitespace

string-process-comma-separated ^4.3.4

Extracts chunks from possibly comma or whatever-separated string

string-remove-duplicate-heads-tails ^7.2.5

Detect and (recursively) remove head and tail wrappings around the input string

string-remove-thousand-separators ^7.2.5

Detects and removes thousand separators (dot/comma/quote/space) from string-type digits

string-remove-widows ^4.2.6

Helps to prevent widow words in a text

string-split-by-whitespace ^4.2.6

Split string into array by chunks of whitespace

string-trim-spaces-only ^5.3.1

Like String.trim() but you can choose granularly what to trim

string-typo-match ^1.1.0

Match typing errors against candidate strings with weighted omissions, swaps, and explainable ambiguity

string-uglify ^3.2.2

Shorten sets of strings deterministically

string-unfancy ^6.3.0

Replace all n/m dashes, curly quotes with their simpler equivalents

↑ back to top

Plain Object or Array Libraries ^Σ=15

Usually, plain objects come from JSON files, and often they are deeply nested. The following libraries help us to traverse them, set and delete keys and compare objects.

Array-processing libraries deal with tasks like sorting, deletion or conversion to other structures.

array-group-str-omit-num-char ^6.2.5

Groups array of strings by omitting number characters

array-includes-with-glob ^5.2.4

Check strings against whole-string wildcard patterns

array-pull-all-with-glob ^7.2.4

Remove array entries using whole-string wildcard patterns

json-comb-core ^8.2.6

The inner core of json-comb

json-variables ^12.2.5

Resolves custom-marked, cross-referenced paths in parsed JSON

object-all-values-equal-to ^4.2.4

Does the AST/nested-plain-object/array/whatever contain only one kind of value?

object-boolean-combinations ^6.3.4

Consumes a defaults object with booleans, generates all possible variations of it

object-delete-key ^4.2.5

Delete keys from all arrays or plain objects, nested within anything, by key or by value or by both, and clean up afterwards. Accepts wildcards.

object-fill-missing-keys ^11.2.5

Add missing keys into plain objects, according to a reference object

object-flatten-all-arrays ^7.2.4

Merge and flatten any arrays found in all values within plain objects

object-flatten-referencing ^7.2.5

Flatten complex nested objects according to a reference objects

object-merge-advanced ^14.2.4

Deeply merge JSON-like data structures

object-no-new-keys ^5.3.4

Check, does a plain object (AST/JSON) has any unique keys, not present in a reference object (another AST/JSON)

object-set-all-values-to ^6.2.5

Recursively walk the input and set all found values in plain objects to something

test-mixer ^4.4.2

Test helper to generate function opts object variations

↑ back to top

AST Libraries ^Σ=12

AST stands for Abstract Syntax Tree, it’s a fancy name for nested output of a parser. Typically, it’s an array or a plain object with many layers of nested elements.

ast-compare ^4.2.5

Compare anything: AST, objects, arrays, strings and nested thereof

ast-contains-only-empty-space ^4.2.2

Does AST contain only empty space?

ast-deep-contains ^5.2.4

Like t.same assert on array of objects, where element order doesn’t matter.

ast-delete-object ^4.2.5

Delete all plain objects in AST if they contain a certain key/value pair

ast-get-object ^4.2.5

Getter/setter for nested parsed HTML AST’s, querying objects by key/value pairs

ast-get-values-by-key ^5.2.5

Extract values and paths from AST by keys OR set them by keys

ast-is-empty ^4.2.4

Find out, is nested array/object/string/AST tree is empty

ast-loose-compare ^4.2.4

Compare anything: AST, objects, arrays and strings

ast-monkey ^9.2.5

Traverse and edit AST

ast-monkey-traverse ^4.3.1

Utility library to traverse AST

ast-monkey-traverse-with-lookahead ^4.2.5

Utility library to traverse AST, reports upcoming values

ast-monkey-util ^3.3.1

Utility library of AST helper functions

↑ back to top

Lerna Libraries ^Σ=3

While maintaining our monorepo we found that some essential tools were missing, so we created them!

lerna-clean-changelogs ^5.2.5

Removes frivolous entries from commitizen generated changelogs

lerna-clean-changelogs-cli ^4.2.6

CLI application to cleanse the lerna/commitizen-generated changelogs

update-versions ^7.2.6

Like npm-check-updates but supports Lerna monorepos and enforces strict semver values

↑ back to top

CLI Apps ^Σ=6

All the following libraries are command-line applications. You install them using -g flag via npm, for example, npm i -g json-sort-cli. Often a package/library/program would have its CLI counterpart: you can use a package programmatically, inside your programs, or you can use its CLI in the terminal, as a standalone program.

For example, csv-sort package is string-in, string-out function. It’s meant to be used by websites, CLIs and Node programs.csv-sort-cli “taps it” and adds file I/O layer and lets you read/write/sort files directly, via a command line.

csv-sort-cli ^4.2.6

Command line app to sort double-entry CSVs coming from internet banking statements

email-all-chars-within-ascii-cli ^4.2.6

Command line app to scan email templates, are all their characters within ASCII range

generate-atomic-css-cli ^4.2.6

Generates and updates all HTML templates’ atomic CSS

js-row-num-cli ^4.2.6

Update all row numbers in all console.logs in given files

json-comb ^0.11.6

Command line app to manage sets of JSON files

json-sort-cli ^4.3.1

Deep-sort JSON files or standard input; package.json retains its special key order

↑ back to top

Miscellaneous Libraries ^Σ=14

That’s all programs which don’t belong to any of the categories above. Here we have programs doing everything, from CSV sorting to Tap output parsing; from regexes to converting colour hex codes.

check-types-mini ^8.2.4

Validate options object

codsen-format-diagnostic-value ^1.1.1

Safely format untrusted JavaScript values for diagnostics

codsen-glob ^1.1.2

Fast file globbing

codsen-utils ^1.10.2

Various utility functions

csv-sort ^7.2.5

Sort double-entry bookkeeping CSV from internet banking

csv-split-easy ^7.2.5

Splits the CSV string into array of arrays, each representing a row of columns