Text Tools

Seventeen tools, and nearly all of them exist because a piece of text arrived in the wrong shape. Copy pulled out of a PDF carries invisible characters. A list of headings needs sorting. A variable name needs to become a URL. A paragraph has to fit inside a field with a hard character limit.

None of it is difficult, which is precisely why people reach for a tool instead of writing a script: the job takes eight seconds, and doing it properly in a text editor takes eight minutes of hunting through menus.

Counting is not as simple as it looks

Word Counter reports characters, characters without spaces, words, sentences, paragraphs, lines, the longest word, and reading and speaking time. Two details change the numbers enough to be worth knowing.

Words are matched with a Unicode-aware pattern covering letters, numbers, apostrophes and hyphens, so “café” counts as one word rather than two and a hyphenated compound stays in one piece. Characters are counted as code points rather than UTF-16 units, so an emoji or an accented letter counts once instead of twice. If the limit you are writing against is enforced by a system that counts bytes, your figure and theirs can differ on non-English text.

Reading time assumes 238 words per minute and speaking time 150 wpm — the two rates most style guides converge on. Both are estimates for an average adult; dense technical material reads slower, and so does a nervous presenter.

When a limit is hard rather than advisory, Text Trimmer cuts to an exact character or word count, optionally appending an ellipsis and breaking at a word boundary. That last part is the whole point: the difference between a description that ends cleanly and one that ends “...straight in your bro”.

Case conversion, and the modes people actually need

Case Converter offers twelve transformations that split into two families worth keeping apart.

The editorial ones are UPPERCASE, lowercase, Title Case, Sentence case and Capitalise Each Word. Title Case is the only one with genuine rules: around thirty short words — a, an, the, of, in, for, with, and the rest — stay lowercase unless they are the first or last word of the title. That is why “The Sound and the Fury” is correct and “The Sound And The Fury” is not. Sentence case capitalises the first letter of the string, of every line, and of everything following a full stop, question mark or exclamation mark.

The programming ones are camelCase, PascalCase, snake_case, kebab-case and CONSTANT_CASE. These have to find word boundaries in text that contains no spaces, so the converter splits camelCase and PascalCase runs first: getUserName becomes three words before it can become get_user_name. The conventions themselves come from elsewhere — kebab-case for URLs and CSS, snake_case for Python and database columns, CONSTANT_CASE for environment variables.

The invisible characters that break imports

The most common text bug has nothing visible in it. Copying from Word, Google Docs, a PDF or a web page quietly brings along:

• Zero-width spaces and joiners (U+200B to U+200F), inserted by layout engines to permit line breaks
• Bidirectional control marks (U+202A to U+202E), which can reorder what is displayed
• Word joiners and invisible operators (U+2060 to U+2064)
• A byte-order mark (U+FEFF), often sitting at the very start of a file
• Soft hyphens (U+00AD), which appear only where a line wraps
• Non-breaking spaces (U+00A0), which look exactly like a space and are not one

Remove Extra Spaces collapses runs of spaces and tabs, trims line ends and can join lines that were broken by layout — the standard repair for text pasted out of a PDF, where every visual line became a real one. It strips the invisible characters above as well, because a comparison against a non-breaking space fails silently and is miserable to debug.

Remove Accents is the neighbouring job: it decomposes each character and drops the combining marks, turning “Zürich” into “Zurich”. Use it for slugs, filenames and search indexes that only accept ASCII. Never use it on display copy, where the accent is part of the spelling.

Markup going in and out

Strip HTML Tags removes every tag and leaves the words, which is how you get readable prose out of a rich-text field or an email body. HTML Entity Encoder goes the other way, turning <, >, & and quotes into entity references so a snippet can be displayed as text instead of being parsed as markup.

The two are not symmetric and should not be treated as a round trip. Stripping tags destroys structure permanently — headings, links and emphasis all become plain words. Encoding is lossless, because entities decode back to exactly the characters they stood for.

Text to Slug solves the publishing version of the same problem: lowercase everything, drop the accents, turn any run of non-alphanumeric characters into a single hyphen, trim both ends.

Sorting, deduplicating and replacing

Sort Lines handles seven orders — A to Z, Z to A, original, reversed, random, by length and numeric — with optional case sensitivity, blank-line removal, trimming and deduplication in the same pass. Numeric ordering matters more than it sounds: sorting a column of prices alphabetically puts 100 before 20, because it compares character by character.

Remove Duplicate Lines is the same operation specialised. It ignores case and whitespace by default, which is the right default — trailing spaces and stray tabs are exactly why a list looks deduplicated but is not.

Find and Replace works in literal or regular-expression mode. Literal mode can be restricted to whole words and is case-insensitive unless you say otherwise, so “The” and “the” both match. Regex mode passes the pattern straight to JavaScript's engine and keeps $1 and $& replacement semantics, reporting a readable message when the pattern fails to compile.

Reversing and palindromes

Reverse Text can flip character order, word order or line order — three genuinely different operations that all get called “reverse”. Character reversal walks code points rather than UTF-16 units, so an emoji or an accented letter survives intact instead of arriving as two broken halves.

Palindrome Checker normalises before comparing: case is folded and punctuation and spaces are dropped, because “A man, a plan, a canal, Panama” is only a palindrome under those rules. It shows the raw comparison alongside the normalised one, so you can see which question you were really asking.

Filler and random words

Lorem Ipsum Generator produces placeholder copy in paragraphs, sentences or words. The value of lorem ipsum is not that it is Latin but that it is not your content: a reviewer who can read the words reads them, instead of judging the layout, which defeats the purpose of a mockup.

Random Word Generator draws from a built-in list filtered by length, for naming exercises, icebreakers, passphrase building and testing autocomplete or search with realistic-looking input.

Two tools that analyse rather than transform

Readability Analyzer scores a passage with all six classic formulas — Flesch Reading Ease on its 0–100 scale, Flesch-Kincaid Grade Level, Gunning Fog, SMOG, Coleman-Liau and the Automated Readability Index — and reports a grade band alongside them. They disagree with each other routinely, because they estimate from different signals: some count syllables, Coleman-Liau counts letters per word, ARI counts characters per word. Read the spread rather than trusting a single number, and remember that all six measure sentence and word length, not clarity of thought.

AI Text Humanizer rewrites machine-generated prose to vary its rhythm, cut formulaic connectives and remove the tics that give it away. Both tools are honest about the limits of a heuristic: they measure surface patterns, so a genuinely clear passage with unusual sentence lengths can score badly while a vacuous one scores well.

Everything stays in the tab

Every tool here is a pure string transformation running in your own JavaScript runtime. Nothing is posted anywhere, so the only length limit is your browser's memory and there is no reason to hesitate before pasting a contract clause, a customer email or an unpublished draft.