This Bar Chart Maker processes data locally in the current browser tab, so labels, values, titles, previews, and SVG exports are not uploaded. Parsing, layout, preview, Blob creation, and export all stay on the local device.
That supports a precise privacy claim, but “safe” also concerns how user text enters the SVG. The maker escapes text for XML, validates chart geometry, and keeps the downloaded file identical to the reviewed preview. It does not decide whether the underlying chart is accurate or responsible to publish.

What Makes This Bar Chart Maker Safe
The relevant safeguards are specific implementation facts rather than a general security guarantee. The Bar Chart Maker turns strict label,value rows into a deterministic vertical bar chart in the browser and exports a standalone SVG from the same markup shown on screen.
| Safety question | Verified behavior |
|---|---|
| Does the source data leave the browser? | No. Labels, values, the optional title, the SVG, and interactions remain in the current browser tab. |
| Can the export differ from the preview? | No. One generated SVG string powers both the visible preview and the Blob download. |
| How is user-supplied text handled? | It is XML-escaped before entering text, tooltip, title, or accessibility attributes. |
| How is numerical geometry protected? | Coordinates come from validated finite values, are bounds-checked, and are rounded to three decimals. |
These controls protect the data-handling and output paths. They do not turn the tool into a data-quality checker, an accessibility auditor, or an editorial review system.
Privacy From Input to Download
The privacy boundary covers the full workflow. No label, value, title, generated SVG, or interaction is uploaded. The tool does not need to send the dataset elsewhere because parsing, layout, previewing, Blob creation, and export all happen locally.
A successful result creates one Blob URL for the download. Editing the title or data immediately clears the previous preview, status, and download, then revokes the old URL. A replacement result, failed generation, or component unmount also revokes it. If Blob creation or URL creation fails, the tool reports an error and leaves no stale download available. This lifecycle avoids immediate-revocation download races and accumulated ObjectURL leaks.
Local processing protects data transport, but it does not decide what belongs in a published chart. The SVG still presents the categories and values you entered, so inspect the result for publication suitability before distributing it.
Safe SVG Handling for Labels and Titles
Every user string is XML-escaped before it enters the SVG. Ampersands, angle brackets, quotes, and apostrophes become entities, while XML-forbidden controls and unpaired surrogates become U+FFFD. Numeric coordinates come only from validated finite values. As a result, a script-like label remains visible text rather than becoming an SVG element, and malformed numeric geometry cannot be inserted directly into the markup.
The generated file uses the W3C namespace and standard rect, line, text, g, and title elements, consistent with the document structure described in the W3C SVG 2 specification. It is a standalone SVG rather than a raster image or an HTML page containing an embedded chart.
One generated SVG string powers both the visible preview and the Blob download, so the exported file matches the markup you reviewed. The chart also includes tooltips and an accessible description. If the visible title is blank, the accessible label remains “Bar chart.” These details make the output easier to inspect without treating the maker as a substitute for responsible chart review.
Create and Download a Reviewed Bar Chart
Use the supported workflow to turn a small categorical dataset into an SVG:
- Open the maker and add an optional title. The title appears on the chart, but leaving it blank does not prevent generation.
- Paste 2 to 30 non-empty rows. Every row must have the form Label,number, with exactly one comma. The label comes before the comma, and the numeric token comes after it.
- Generate the chart. A successful request creates one vertical bar for every accepted row, together with a baseline, five numeric reference ticks, rotated category labels, tooltips, and an accessible description.
- Check the preview carefully. Verify the category labels, five tick values, baseline, zero-valued bars, and any extreme-range visibility issue. Hover over bars when you need to confirm their values, and look for the explicit note when all values are zero.
- Download the standalone SVG. The file is created from the exact string shown in the preview, so the downloaded markup and reviewed preview are the same.
If you change the title or any data after generating a chart, the previous preview, status, and download are cleared. Generate the updated chart again before downloading it.
Input Rules for Complete, Valid Results
The input format is deliberately CSV-lite rather than full CSV. It is designed for a small set of simple category and value pairs, not spreadsheet exports with quoting or extra columns.
| Input area | Accepted | Not accepted |
|---|---|---|
| Dataset size | 2 to 30 non-empty rows | Fewer or more than 30; the entire request fails |
| Row structure | Exactly one comma, with a label before it and a numeric token after it | Quoted fields, quoted commas, additional columns, or an empty label |
| Numeric values | Nonnegative finite base-ten numbers; ordinary decimals and e/E notation; nonzero values from 1e-300 through 1e300 | Typed units, expressions, hexadecimal, Infinity, NaN, nonzero input that underflows to zero, or input above the maximum |
| Text length | A label of at most 24 UTF-16 code units and a numeric token of at most 40 | One code unit beyond either limit |
| Whitespace | Surrounding whitespace is trimmed, and blank or whitespace-only lines are ignored | A non-empty line that still lacks a valid comma, label, or number |
Validation applies to the whole request. If the count is wrong or any row is malformed, generation fails instead of skipping the bad row, substituting zero, retaining only the first 30 rows, or showing a partial chart. Exact row-count and text-length limits are accepted. Negative zero is normalized to zero.
Duplicate labels are allowed and remain separate bars in input order. The tool does not aggregate equal labels, sort categories, or infer units, so the order and meaning of the source data remain the author’s responsibility.
Scales, Zeros, and Extreme Visibility
The chart uses a nonnegative linear scale. Zero maps to the baseline, and the displayed axis maximum maps to the top of the plot. The upper bound rounds upward to a finite 1, 2, 5, or 10 multiple of a power of ten, with five evenly spaced reference ticks exposing the range. This is the product’s deterministic nice-domain approach; it is not a promise to reproduce every D3 tick choice. The general idea of proportional linear mapping is covered in D3’s linear-scale documentation.
- All values are zero: The chart uses an explicit zero-to-one reference range, keeps every bar at zero height on the baseline, and displays an all-zero note. This avoids division by zero without changing the supplied values.
- A tiny positive value sits beside a huge value: The smaller bar can become visually indistinguishable from zero because of SVG coordinate precision. Its tooltip still reports the value. The maker does not invent a minimum-height bar because that would break the stated linear scale.
- Visibility is essential across an extreme range: Use a logarithmic or separately grouped chart when tiny categories must remain visible beside values hundreds of orders larger.
- There are many categories: The SVG remains 520 pixels high and expands horizontally as rows are added. The preview scrolls instead of squeezing all labels into a mobile width.
Axis labels use up to six practical significant digits or scientific notation. Keep the source numbers when exact analysis is required, because a displayed axis label is not a replacement for the original values.
Checks to Complete Before Publishing
Run a final visual and data review after generating the chart:
- Confirm every label: Check spelling, order, duplicates, and whether any category needs a unit or clearer name.
- Read the scale: Verify the baseline, all five tick values, the displayed maximum, and the all-zero note when applicable.
- Check small bars: Use the tooltips to confirm values that may be difficult to distinguish visually.
- Review the full width: With numerous categories, inspect the horizontally scrolling preview rather than judging only the first visible labels.
- Confirm accessibility details: Check the visible title and the generated accessible description or label.
- Compare the export: Download the SVG and confirm that it presents the same reviewed chart rather than a separate rendering.
Limits on the Safety Claim
A safely encoded SVG is not automatically a scientifically meaningful chart. The maker does not evaluate formulas, fetch data, parse full CSV, create raster images, add legends or series, or choose colors by meaning. It also does not use logarithmic scaling, normalize values to percentages, or imply statistical significance.
The local-processing and escaping controls address data transfer and output construction, but they do not verify whether the chart communicates responsibly. Keep the original dataset, review the axis and labels, and use a different charting method when logarithmic scaling, multiple series, richer units, or another missing capability is required.
Related reading: Is a Compliment Generator Safe? A Clear, Practical Guide.
Related reading: How Do I Get Started With a Line Graph Maker?.