4629 · Analysis · v1.0.0

Text Inspector

Every invisible detail, made visible.

Measure code points, graphemes, words, lines, byte sizes, normalization differences, invisible characters and empirical character frequencies.

A glowing sheet of text-like lines examined under a magnifying glass with small floating charts
AI-generated conceptual artwork; not an application screenshot.
Processed locally in your browser

Text-only workspace. Never executes input. General limit: 200,000 UTF-16 units per editor; some operations have tighter limits.

Private session · history disabled

Opt in to store up to five recent text transformations on this device. Never stored on an account or synchronized. Files are never included. Disable history stops new saves; Clear history deletes saved items.

How to use Text Inspector

Paste input, select a command and its options, then run the transformation. Review diagnostics before copying or downloading the result. Input is not executed.

Words and sentences follow the browser’s locale-aware segmenter; counts are linguistic estimates, not universal definitions. Empty input has zero lines. A trailing newline contributes an empty final line. Empirical Shannon entropy summarizes observed code-point frequencies; it does not measure password strength, unpredictability or encryption security.

Privacy and boundaries

Editor contents stay in your browser. Transformations run in a disposable worker and are never sent to a formatting or AI service. History starts disabled; enabling it writes text to local browser storage only. Clearing that history removes this tool’s saved entries, not data saved by other applications. The surrounding website may load its normal navigation and analytics, but the utility workspace is excluded from content-interaction capture.

Use the labeled operation selector for the available capabilities. Keyboard users can reach every control with Tab; Ctrl or Command plus K opens the command palette. Copy requires clipboard permission. Download creates a local text file. Clear removes the current workspace; it does not automatically delete an opted-in history. Inputs and outputs are always rendered as text, never as a runnable preview.

Why do character counts disagree?

A visible character may use multiple code points, UTF-16 code units and bytes. Counts must say which unit they measure.

Unicode Explorer ↗

See the structure inside a glyph.

ASCII Explorer ↗

128 signals. A complete foundation.

Diff Lab ↗

Two versions. One clear explanation.

FIELD GUIDE · PURPOSE AND BOUNDARIES

Text Inspector: practical use and context

Text Inspector is a diagnostic workbench for editors, translators and developers comparing the different units that make up a string. It exposes representation details rather than grading writing quality.

↑ Back to the application
A glowing sheet of text-like lines examined under a magnifying glass with small floating charts
Application artwork reused from its publication image family. The interactive interface is above.

Inspect before changing text

Paste the exact text, keeping tabs and line endings intact, and run the inspection. Compare the named measures rather than looking for one universal character count. Keep the original while investigating. A text field specified in UTF-8 bytes, a programming string indexed in UTF-16 units and a user-visible character limit can legitimately produce different answers for the same sample.

For a localization exercise, inspect the included accented greeting and emoji. Add a trailing newline and inspect again. Note which measures changed and why an additional empty final line is not another visible word. Replace a plain space with a tab to investigate a formatting difference without changing the surrounding letters.

Use the measure that matches the contract

Word and sentence counts use the browser’s language-aware segmentation. They are useful estimates, not a universal editorial definition. A publication can define hyphenated terms, headings or footnotes differently. Document the counting rule if a length requirement affects acceptance; do not compare two counters without first checking whether they measure the same units.

Normalization differences help explain why visually similar strings compare differently. Inspection does not authorize rewriting a person’s name or changing an identifier. Examine the code points in Unicode Explorer when you need to identify a specific mark. Keep descriptive diagnosis separate from the policy decision about what your application should accept.

Frequency is not a security score

Empirical Shannon entropy summarizes the observed distribution of code points in this sample. It does not estimate password strength, encryption quality, human creativity or the probability of the next message. A short string with diverse characters can produce an interesting frequency distribution without being secret or unpredictable. Use the figure as a description of the supplied sample only.