Text Case Converter

What this tool does

Type once and every case style appears at the same time, each with its own copy button. There is no dropdown, because the usual problem is not converting to a case you have in mind — it is that you know the shape you want and not its name. Seeing screaming_snake beside Train-Case beside dot.case settles that in a glance.

The interesting work happens before any conversion: deciding where the words are. That step is what separates a converter that turns XMLHttpRequest into xml_http_request from one that produces x_m_l_http_request.

Common uses

• Renaming an API field across a stack — userAccountId in TypeScript, user_account_id in Postgres, USER_ACCOUNT_ID in the environment file, user-account-id in the URL.
• Turning a Jira ticket title into a branch name, or a heading into a CSS class.
• Converting a column of spreadsheet headers into identifiers — switch on "convert each line separately" and paste the whole column.
• Writing an HTTP header name correctly: Content-Type is Train-Case, not content-type or Content_Type.

A short example

Given getHTTPResponseCode, the splitter finds four words and produces:

camelCase             getHttpResponseCode
PascalCase            GetHttpResponseCode
snake_case            get_http_response_code
SCREAMING_SNAKE_CASE  GET_HTTP_RESPONSE_CODE
kebab-case            get-http-response-code
Train-Case            Get-Http-Response-Code
dot.case              get.http.response.code
path/case             get/http/response/code

How the word splitter handles acronyms

Splitting is done with an ordered set of rules, and the order is the whole trick. The first rule peels a capital run off the front of a following capitalised word, which is why XMLHttpRequest becomes XML + Http + Request rather than XMLHttp + Request. The same rule recovers HTTP from the middle of getHTTPResponseCode and URL from the end of parseURL.

Digits attach to the word they follow, so utf8Encoder splits as utf8 + Encoder and useHTTP2Pool as use + HTTP2 + Pool — not as a stray number in the middle. Existing delimiters (space, _, -, ., / and any punctuation) are boundaries, so already-converted input round-trips cleanly. Apostrophes are removed rather than split on, so user's profile gives usersProfile instead of userSProfile.

Worth knowing

Title Case and Sentence case work differently from the rest. They operate on your original text rather than on the split words, because punctuation and spacing carry meaning in a sentence and would be destroyed by rejoining. Title Case applies the usual convention of leaving short function words — a, of, the, to — lowercase unless they start or end the title, and leaves any word with an interior capital alone, so iPhone and macOS survive.

The delimiter conversions are lossy in one specific way: case boundaries are the only record of where words begin, so converting ID to camelCase and back yields id, not ID. Acronym capitalisation cannot be recovered once it has been flattened into snake_case.