Convert text case

They're conventions rather than rules, but each one has become standard in a particular place, and following the local convention is worth more than preferring one personally.

camelCase for variables in JavaScript, Java and C#. snake_case for Python and for database columns. kebab-case for URLs, CSS class names and HTML attributes. PascalCase for class and component names in most languages. SCREAMING_SNAKE_CASE for constants and environment variables.

Why kebab-case for URLs

Search engines treat hyphens as word separators and underscores as joiners, so blue-suede-shoes reads as three words and blue_suede_shoes reads as one. That's a real difference for anything you want found.

URLs are also often case-insensitive on the server and case-sensitive elsewhere, which is a reliable source of broken links. Lowercase with hyphens avoids both problems, which is why slugs look the way they do.

Why snake_case for databases

Most SQL databases fold unquoted identifiers to one case. PostgreSQL lowercases them, so a column created as firstName becomes firstname, and every query referring to firstName works by accident until someone quotes it and it doesn't.

Using first_name sidesteps the whole issue, which is why it's near-universal in database schemas even in projects whose application code is camelCase.

Title case is not just capitalising everything

The convention most style guides follow: capitalise the first word, the last word, and everything in between except articles (a, an, the), short prepositions (in, on, of, to, for) and coordinating conjunctions (and, but, or, nor).

So it's "The Man in the High Castle", not "The Man In The High Castle".

Guides disagree on the details, particularly on how long a preposition can be before it gets capitalised, with three, four and five letters all having adherents. Automatic title case gets the common cases right and can't know whether "Turn On" is a phrasal verb needing the capital.

Check anything going on a cover or a title page by eye.

Sentence case is usually better for headings

Most modern style guides prefer sentence case for headings on the web: capital on the first word and on proper nouns only.

It's easier to read at a glance, it avoids the title case arguments entirely, and it looks less shouty. If you're writing documentation or a product interface, sentence case is the safer default.

What breaks when converting

Acronyms. parseHTMLDocument in snake_case should probably be parse_html_document, and a naive conversion produces parse_h_t_m_l_document. Handling this properly means knowing that HTML is an acronym, which no converter can do reliably.

Numbers. Whether address2 becomes address_2 or address2 depends on what the number means, and both conventions exist.

Names and trademarks. iPhone, eBay, McDonald and van der Berg all have capitalisation that isn't algorithmic. Any bulk conversion will damage them.

Non-English text. German nouns are always capitalised. Turkish has a dotless ı whose uppercase is I while i uppercases to İ, which has caused real bugs in software that lowercases identifiers.

Check the output when the input contained any of these.

Converting a whole codebase

Don't do it with a text tool. Use your editor's rename feature or a language- aware refactoring tool, which knows which occurrences are the same symbol.

Find and replace across files will change strings, comments and unrelated identifiers that happen to share a name, and the result compiles while doing something slightly different.

Where this tool is genuinely useful

One-off conversions. Turning a heading into a slug, a label into a variable name, a spreadsheet column into a database column, or a list of names into a consistent format.

Everything happens as you type, in your browser. Nothing is sent anywhere.