physics/named-fn-to-operatorname

Level

AuthoringFaithfulCorpusEquivFaithful, Corpus, and Equiv apply this rule.

Fidelity

readingPreserves notation, reading order, and structural roles.

Package

physics

Source

physics/faithful/named_fn_expand/named_fn_to_operatorname.rs (opens in new tab)

Triggers

\asin, \acos, \atan, \csch, \arccsc, \acsc, \sech, \arcsec, \asec, \arccot, \acot, \tr, \Tr, \erf, \trace, \Trace, \sine, \cosine, \naturallogarithm, \Probability commands

Eliminates

\asin, \acos, \atan, \csch, \arccsc, \acsc, \sech, \arcsec, \asec, \arccot, \acot, \tr, \Tr, \erf, \trace, \Trace commands

Produces

\operatorname, \left, \right commands

Touches

\sine, \cosine, \naturallogarithm, \Probability commands

Rewrite patterns#

1. bare\asin\operatorname{asin}
2. trig-paren\asin(#1)\operatorname{asin}\left(#1\right)
3. trig-power-paren\asin[#1](#2)\operatorname{asin}^{#1}\left(#2\right)
4. trig-literal-power\asin[#1]\operatorname{asin}[#1]
5. operator-brace\tr{#1}\operatorname{tr}\left\{#1\right\}
6. operator-bracket\tr[#1]\operatorname{tr}\left[#1\right]
7. operator-paren\tr(#1)\operatorname{tr}\left(#1\right)
8. plain-operator\erf\operatorname{erf}
9. trace-alias\trace\operatorname{tr}
10. long-name-argument\sine(#1)\operatorname{sin}(#1)

Examples#

Each example shows this rule applied in isolation. A complete normalization profile may make additional changes.

\asin[2](x(y+z)\frac{a}{b})

\operatorname{asin}^{2}\left(x(y+z)\frac{a}{b}\right)

\sech(\frac{x}{y})

\operatorname{sech}\left(\frac{x}{y}\right)

\csch(x)y

\operatorname{csch}\left(x\right)y

\asin[2]x

\operatorname{asin}[2]x

\csch x

\operatorname{csch} x

\tr{\frac{A}{2}}

\operatorname{tr}\left\{\frac{A}{2}\right\}

Related rules#

Upstream rules#

• physics/trace-capital-to-Tr Collapse a bare Trace with no following argument group to the local Tr anchor.
• physics/trace-to-tr Collapse a bare trace with no following argument group to the local tr anchor.

Downstream rules#

• ams/operatorname-to-named-function Rewrite plain operatorname arguments to equivalent base named-function or named-operator commands.

Same family#

• ams/operatorname-to-named-function Rewrite plain operatorname arguments to equivalent base named-function or named-operator commands.
• ams/idotsint-expand Expand idotsint to explicit repeated integral surfaces.
• base/limits-drop Drop explicit limits from an audited operator while preserving attached scripts.
• base/nolimits-drop Drop explicit nolimits from an audited operator while preserving attached scripts.
• physics/cosine-to-cos Collapse a bare cosine helper to the standard cos operator.
• physics/naturallogarithm-to-ln Collapse a bare natural logarithm helper to the standard ln operator.
• physics/probability-to-pr Collapse a bare Probability helper to the standard Pr operator.
• physics/sine-to-sin Collapse a bare sine helper to the standard sin operator.
• physics/trace-capital-to-Tr Collapse a bare Trace with no following argument group to the local Tr anchor.
• physics/trace-to-tr Collapse a bare trace with no following argument group to the local tr anchor.
• physics/named-fn-paren-expand Expand physics power and auto-sized parenthesis arguments of standard named functions.
• physics/physics-operator-fence-expand Expand the physics auto-sized argument of exp, det, and Pr into an explicit fence.
• physics/rank-to-operatorname-rank Rewrite rank to an explicit operatorname form instead of a package-local alias.

Notes#

Semantics follow LaTeX physics.sty 1.3. None of the short names here has a base command, so each one becomes \operatorname with its printed name (\trace and \Trace print tr and Tr). Trig-family names use \trigbraces (o d()): \asin[n](...), \acos[n](...), \atan[n](...), \csch[n](...), \arccsc[n](...), \acsc[n](...), \sech[n](...), \arcsec[n](...), \asec[n](...), \arccot[n](...), \acot[n](...) get a power and an auto-sized parenthesis fence, and \asin[n] without following parentheses keeps [n] as literal text. \tr{...}, \tr[...], \tr(...), \Tr{...}, \Tr[...], and \Tr(...) use \opbraces (g o d()): only the first group becomes an auto-sized fence. The physics records hold the grabbed group as a d() or paired d<(,)><[,]><{,}> slot, so the parser attaches it. \erf, \trace, and \Trace are plain \DeclareMathOperator commands, so following delimiters stay literal. A command that carries scripts (\tr_A(\rho)) is renamed without grabbing an argument. The long names \sine, \cosine, \naturallogarithm, and \Probability are the original operators saved with \let, so they grab nothing either; the rule rewrites them to \operatorname{sin}, \operatorname{cos}, \operatorname{ln}, and \operatorname{Pr} only when (, [, or a brace group follows, where the short command would grab it. Bare long names are left to the authoring aliases (sine-to-sin, cosine-to-cos, naturallogarithm-to-ln, probability-to-pr), so the long names are touches: every Applied on them removes one long name followed by an argument opener, and bare long names return Skipped. Fidelity is reading: physics closes the fence inside \mathopen{}\mathclose\bgroup...\egroup, which acts as a closing atom, while an explicit \left...\right fence is an inner atom. The explicit fence therefore adds about 1/6 em before the opening delimiter and another 1/6 em after the fence when an ordinary atom or an opening delimiter follows, about 1/3 em in the worst case; script styles show no difference. No compensating space is inserted. MathJax differs: MathJax physics treats \asin[2]x as a power, auto-sizes \erf(...) and \trace(...), hides the braces of \tr{...}, and prints \tr[...] literally.

Disable this rule#

TransformEngine(profile="corpus", disable_rules=["physics/named-fn-to-operatorname"])

new TransformEngine({ profile: 'corpus', disableRules: ['physics/named-fn-to-operatorname'] })

Open in Playground