Word & Character Counter

Eight figures update on every keystroke, and two of them catch people out. Characters counts the raw string including every space and newline. No Spaces strips all whitespace first, which is the number a form limit or a database column actually enforces. The Word & Character Counter shows both at once rather than making you guess which one a 280-character cap means.

Below the grid sits a keyword strip. It lowercases the text, splits hyphenated compounds so search-engine-optimizationcontributes three tokens rather than one, discards anything under three letters, filters 71 common stop words, and lists the eight most frequent survivors with a count beside each. Enough to notice that a draft leans on pricing nine times.

The whole analysis is a handful of regular expressions run against the string sitting in the textarea. Nothing is uploaded, so an unpublished manuscript or a client contract can be measured without it leaving the machine you opened it on.

What Each Figure Counts

• Words splits the trimmed text on runs of whitespace, so don't and end-to-end are one word each.
• Sentences splits on ., ! and ? and keeps the non-empty pieces, which means a closing sentence with no full stop is still counted.
• Paragraphs needs a genuinely blank line between blocks. A single newline is a line break, and Lines is the figure that tracks those.
• Avg Word Length divides characters-without-spaces by words. Ordinary English prose lands near 4.7; technical writing climbs past 6.
• Reading Time shows a dash until the first word arrives, then switches from a plain seconds count to 1m 30s form once you cross a minute.

The Reading-Time Formula

Reading time is words ÷ 200 × 60, rounded to whole seconds. 200 words per minute is the usual silent-reading rate quoted for adult non-fiction, so a 1,000-word article reports 5m 0s. Read the same text aloud and you will be nearer 150 wpm, which means a script takes roughly a third longer than the figure shown here.

What Happens When Punctuation Goes Missing

• Unpunctuated text counts as one sentence, not zero. An earlier build required a trailing terminator and reported 0 for a whole paragraph that simply had not been punctuated yet.
• Runs of terminators collapse into a single split, so typing Well... adds one sentence rather than three.
• Abbreviations do split. Dr. Reyes arrived at 9 a.m. registers as four sentences, which is the price of splitting on punctuation instead of parsing language.
• Digits are stripped before keywords are gathered, so 2026 never reaches the keyword strip however often you write it.
• An empty box reports zero everywhere except Lines, which reads 1, because an empty string is still one line.