Convert camel, snake, kebab, Pascal, title and sentence case, sort, dedupe, wrap and shuffle lines, slugify, count words, and escape for HTML, JS, C, shell and SQL.
How it works
The output follows as you type, and the arrow between the boxes makes it the next input, so trimming, sorting and wrapping one text is three clicks rather than three pastes. Both boxes are counted: characters, graphemes where they differ, characters without spaces, words, lines and UTF-8 bytes.
The five identifier cases rebuild each line from its words. A run of capitals is an acronym up to the capital that starts the next word, so XMLHttpRequest is three words, and a run of digits is a word of its own. Whatever sat between the words is dropped. Title and sentence case only change letters, and the headline variant keeps short words such as of and the lowercase unless they come first or last. A colon, question mark or exclamation mark starts a phrase with a first and last word of its own, so a tale of two cities: the story so far becomes A Tale of Two Cities: The Story so Far.
The line operations read \r\n, \r or \n, write \n, and leave a final newline at the end rather than sorting it as an empty line. The Lines count agrees, so a, b and a final newline count as 2 lines. Sorting uses English collation whatever the browser's language, so a shared link sorts the same way anywhere. Reversing characters goes by grapheme.
JavaScript and C escaping give the body alone, with both quotes escaped (and the backtick, for JavaScript), so it fits between whichever you write. The shell and SQL give the quoted word, because there the quotes are the escaping. HTML decoding reads numeric references from 128 to 159 as Windows-1252, as browsers do.
Examples
A column of names in snake_case
XMLHttpRequest
user id
utf8Decoder
parse-JSON-responsexml_http_request
user_id
utf_8_decoder
parse_json_responseEach line is converted on its own; camelCase gives xmlHttpRequest, userId, utf8Decoder and parseJsonResponse.
One string, three kinds of quoting
it's $HOMEShell word with single quotes, shell word with double quotes, and SQL string, standard:
'it'\''s $HOME'
"it's \$HOME"
'it''s $HOME'Inside single quotes nothing is special, so the quote itself closes the string, is escaped, and reopens it. Inside double quotes it is the $ that needs escaping.
Counting and reversing a decomposed letter and a flag
noël 🇬🇧Here the diaeresis is a separate combining mark, U+0308. The counts read 8 characters, 6 graphemes and 15 bytes, and reversing the characters gives:
🇬🇧 lëonThe mark stays on its e and the flag's two regional indicators stay together. The Unicode inspector lists all eight code points.
Common problems
The count disagrees with length
JavaScript's length counts UTF-16 code units. The page counts code points as characters, and graphemes too when the two differ. 👍🏽 has a length of 4, is 2 characters, 1 grapheme and 8 bytes. A column limited in bytes wants the UTF-8 figure, and a limit someone reads on screen wants graphemes.
An acronym does not survive the round trip
Converting rebuilds a name from its words and forgets how they were capitalised, so XMLHttpRequest to snake_case and back to PascalCase is XmlHttpRequest. Likewise utf8 becomes utf_8, because digits are a word of their own.
file10 sorts before file9
A to Z compares character by character and 1 comes before 9. Natural (10 after 9) reads runs of digits as numbers: file1.txt, File2.txt, file9.txt, file10.txt. Case and accents only break ties, so apple sorts next to Apple and Éclair among the other words starting with e.
A MySQL-escaped string breaks elsewhere
The MySQL variant writes a quote as \' and a backslash as \\. PostgreSQL, with its default standard_conforming_strings, and SQLite treat a backslash as an ordinary character, so there \' ends the string. The standard variant doubles the quote, 'it''s', and leaves backslashes alone.
Frequently asked questions
Why does the C escape use octal rather than hex?
Because a C hex escape takes as many hex digits as follow it, so \xE9 followed by the letter a is read as one escape that overflows a byte. Octal escapes stop after three digits. With the non-ASCII variant, é becomes its two UTF-8 bytes, \303\251, and unescaping reads them back as one letter.
Why does sentence case lowercase names?
It lowercases the whole line and capitalises the first letter of each sentence, and it has no way to know that London or I should keep their capitals. Proper nouns and acronyms need putting back by hand afterwards.
What does the URL slug do with letters that have no accent to strip?
It spells them out. Accents are removed by decomposing the text and dropping the marks, but letters like ß, æ, ø and þ do not decompose, so they are written as ss, ae, o and th. Straße & Café: Größe 2.0! becomes strasse-cafe-grosse-2-0.