About

MathML is the W3C standard for encoding mathematical expressions in XML. It is verbose by design: a simple expression like x² + 1 requires over 40 characters of markup. ASCIIMathML compresses the same semantics into a human-readable shorthand (x^2 + 1) that tools like MathJax can render. Manual conversion is error-prone. A misread <mfrac> nesting level turns ab into ab. This tool parses the MathML DOM tree recursively, maps each presentation element to its ASCIIMathML token, and emits a flat string you can paste directly into any ASCIIMathML-aware renderer. It handles <mfrac>, <msqrt>, <mroot>, <msup>, <msub>, <msubsup>, <mover>, <munder>, <mfenced>, <mtable>, and standard <mo> operator entities. This tool approximates Presentation MathML only. Content MathML (<apply>, <ci>) is outside scope.

Formulas

The converter applies a recursive tree-walking algorithm. Each MathML presentation element maps to an ASCIIMathML token pattern:

convert(node) →

{

text if node is text nodemapOp(content) if tag = momapIdent(content) if tag = mi(convert(child₀)) / (convert(child₁)) if tag = mfracconvert(child₀)^(convert(child₁)) if tag = msupjoin(convert(children)) otherwise

Where mapOp translates Unicode operators to ASCIIMathML tokens (e.g., × → xx, ÷ → -:, ≤ → <=). mapIdent converts Unicode Greek letters to their ASCII names (e.g., α → alpha, Σ → Sigma). The recursion depth equals the MathML DOM nesting depth, typically under 20 levels.

Reference Data

MathML Element	Description	ASCIIMathML Output	Example Input
`<mn>`	Number	Literal number	`<mn>42</mn>` → 42
`<mi>`	Identifier / variable	Literal or Greek name	`<mi>α</mi>` → `alpha`
`<mo>`	Operator	ASCII operator symbol	`<mo>×</mo>` → `xx`
`<mfrac>`	Fraction	`(a)/(b)`	`<mfrac><mi>a</mi><mi>b</mi></mfrac>`
`<msqrt>`	Square root	`sqrt(x)`	`<msqrt><mi>x</mi></msqrt>`
`<mroot>`	N-th root	`root(n)(x)`	`<mroot><mi>x</mi><mn>3</mn></mroot>`
`<msup>`	Superscript / power	`x^2`	`<msup><mi>x</mi><mn>2</mn></msup>`
`<msub>`	Subscript	`x_i`	`<msub><mi>x</mi><mi>i</mi></msub>`
`<msubsup>`	Sub + superscript	`x_i^2`	`<msubsup><mi>x</mi><mi>i</mi><mn>2</mn></msubsup>`
`<mover>`	Accent above	`hat(x)`, `bar(x)`, `vec(x)`, `dot(x)`	`<mover><mi>x</mi><mo>^</mo></mover>`
`<munder>`	Script below	`underset(b)(a)`	`<munder><mo>∑</mo><mi>i</mi></munder>`
`<munderover>`	Scripts above & below	`sum_(i=0)^n`	`<munderover><mo>∑</mo><mrow>...</mrow><mi>n</mi></munderover>`
`<mfenced>`	Fenced group (parens, brackets)	`(a, b)` or `[a, b]`	`<mfenced open="[" close="]">...</mfenced>`
`<mrow>`	Group / row	Concatenation	Transparent wrapper
`<mtext>`	Text annotation	`text(hello)`	`<mtext>hello</mtext>`
`<mspace>`	Whitespace	Single space	`<mspace/>`
`<mtable>`	Matrix / table	`[[a,b],[c,d]]`	Nested `<mtr>` / `<mtd>`
`<mstyle>`	Style wrapper	Pass-through	Children processed recursively
`<mpadded>`	Spacing wrapper	Pass-through	Children processed recursively
`<menclose>`	Enclosure (box, strike)	`cancel(x)` for `updiagonalstrike`	`<menclose notation="updiagonalstrike">...</menclose>`
`<merror>`	Error marker	Pass-through content	Children processed recursively

Frequently Asked Questions

No. This converter processes Presentation MathML only - elements like , , , and . Content MathML uses a different semantic model (, , ) that maps to function application rather than visual layout. If your source contains Content MathML, convert it to Presentation MathML first using an XSLT stylesheet such as the one provided by the W3C (ctop.xsl).

The converter maintains a dictionary of approximately 60 Unicode code points mapped to ASCIIMathML operator tokens. For example, × (U+00D7) maps to "xx", ± (U+00B1) maps to "+-", ∑ (U+2211) maps to "sum", ∫ (U+222B) maps to "int", ≤ (U+2264) maps to "<=", and ∞ (U+221E) maps to "oo". Operators not in the dictionary pass through as their literal Unicode character.

The converter uses the browser's DOMParser in XML mode. If the input is not well-formed XML, DOMParser returns a document containing a element. The tool detects this and displays a toast notification with the parser error message. Common issues include unclosed tags, unescaped ampersands (&), and missing the root

element.

is a transparent grouping element in MathML. The converter treats it as a pass-through: it recursively converts all children and concatenates the results with spaces. Redundant nesting (e.g., 1) collapses cleanly to just "1". No extra parentheses are added unless the is a child of a structural element like where grouping is implicit.

Yes. If the input lacks a

root element, the converter automatically wraps the input in ... before parsing. This allows you to paste raw fragments like \frac{1}{2} directly. The xmlns declaration is required for DOMParser to correctly identify the MathML namespace.

The converter inspects the second child of to determine the accent type. A combining circumflex (ˆ, U+005E or U+0302) maps to "hat(x)", a combining macron (¯, U+00AF or U+0304) maps to "bar(x)", a right arrow (→, U+2192) maps to "vec(x)", a dot (˙, U+02D9 or U+0307) maps to "dot(x)", and a double dot (¨, U+0308) maps to "ddot(x)". Unrecognized accents fall back to "overset(accent)(base)".