Keyword Position Checker

How it actually works

This is a local text locator, not a rank checker: the whole document stays in the textarea and no request leaves the page. The text is split on newlines and each line is searched in turn with a regular expression built from your keyword - every regex metacharacter in the term is escaped first, so a phrase containing a dot, a bracket or a plus sign matches those characters literally instead of acting as a pattern.

Two switches change what counts as a hit. Case Sensitive toggles the case-insensitive flag; with it off, Fox, fox and FOX all match. Whole Word Only, which is on by default, wraps the term in word boundaries so a match is found only as a standalone word and not inside longer words such as foxes. For a term containing punctuation the boundaries still apply, so a hyphenated phrase behaves like a word search rather than a substring search.

For every match the tool records three numbers: the line it sits on, the word index on that line counting from 1, and the character offset where the match starts. The word index is derived by splitting the text before the match on whitespace, which is why an answer is a position in reading order rather than a raw byte count.

What that means for your results

The preview is built by escaping the pasted text before wrapping each match in highlight markup, so a keyword search over a block that contains angle brackets or ampersands renders as text rather than being interpreted as HTML - a real safety property when you point the tool at markup or code.

Every answer is reproducible and inspectable: the same text and keyword always give the same line, word and character triple, and the three stat counts - matches, lines with matches and total characters - let you cross-check the list at a glance.

Where it falls short — honestly

It cannot tell you where a page ranks. Positions here are line, word and character offsets inside the text you paste, and nothing about Google, Bing or any search index is measured, because rank data would need a crawl index and a network query, neither of which this page performs.

Whole-word matching assumes an alphanumeric term; searching a string made of non-word characters with Whole Word Only on can match nothing, since a word boundary needs a word character next to it. Matching is exact-substring, so there is no stemming, no synonyms and no fuzzy tolerance - run will not find running unless you search the stem yourself. And counting is per line, so a phrase that wraps across a newline is not recognised as one match.

A real end-to-end workflow

Use it to audit repetition in a draft, to find every place a term must be changed in a manuscript, or to locate a token in a long log or subtitle file - then jump by the line number and read the character offset to see exactly where each hit starts.

To study how often a term appears rather than where, Keyword Density counts frequency and percentage; to normalise casing before searching, run the text through Case Converter first, and for the literal reversal of order see Reverse Text.