SVG Inspect
What SVG Inspect actually does
Before changing a vector file it helps to know what kind of file it is, and the answer is not visible from the drawing. A graphic that looks like line art may be a photograph in a wrapper; a file with no text elements will return nothing from a text extractor no matter how many words are visible in the picture; duplicate ids explain rendering that seems random. This reports those facts directly, beside the rendered drawing, so the file's structure and its appearance can be compared in one place.
How to use it
- Choose one or more SVG files.
- Run it and read the counts beside the rendered drawing.
- Check the warnings, which name the conditions that break downstream tools.
- Download the report as JSON if it is feeding something else.
Useful for
- Finding out why an SVG is far larger than its drawing suggests.
- Checking whether visible labels are real text or just shapes before trying to extract them.
- Diagnosing an icon that renders inconsistently when duplicate ids are the cause.
Limits worth knowing
- This reports structure, not appearance. It cannot tell you whether the drawing is correct.
- Visible words drawn as paths are not text and are counted as paths; recognising them needs OCR.
- Element counts are of the document as written, so content hidden by clipping or opacity is still counted.
Questions people ask
Why does it say there are no text nodes when I can read words?
Because those words were converted to outlines when the file was exported. They are shapes now, and only an OCR pass can read them back.
What makes duplicate ids a problem?
References resolve to one of them, so gradients, masks and clips can apply to the wrong element. The id prefix tool is the fix.
Can I use the report in a script?
Yes, the JSON report is downloadable and is the same structure the command line tool emits.