Wikipedia:Improvingtherenderingof

chemicalformulae

Student:ManfredSchäferSupervisor:MoritzSchubotzExaminer:Prof.Dr.BelaGippDate:2017-07-17

Motivation

H2O<chem>H2O</chem>

(NH4)2S<chem>(NH4)2S</chem>

2Na+2H2O⟶ 2Na+ +OH– +H–H<chem>2Na + 2H2O -> 2Na+ + 2OH- + H-H</chem>

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 2

Motivation

\ce{2Na + 2H2O -> 2Na+ + 2OH- + H-H}

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 3

2 Na + 2 H2O ��! 2 Na+ + 2 OH– + H–H

Whyusetexvcjs atall?• Security:texvcjs ensuresthatonly“safe”LaTeXexpressionsareusedandattackmethodslikecrosssitescriptingareprevented• ReplacesMediaWiki specificfunctions• MakesLaTeX expressionsmoreconsistentbyintroducingbracesorremovingwhitespacesbeforerendering

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 4

Objective• Adjusttexvcjs topreventwhitespacemodificationsince-tags• Discussedoptionswere:

(1) Developafullgrammarformhchem(2) Extendgrammaroftexvcjs totreat\ce like

\mbox(3) Bypasstexvcjs altogetherforrendering(4) Duplicatemhchem parserintotexvcjs

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 5

Researchtasks– Milestones• Familiarizemyselfwiththestructureoftherenderingpipeline• Evaluatethefouroptionsandchoseone:option(1)waschosen• Implementoption(1)• Developtestcasesandtestthesystem

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 6

Challenges• Understandingpegjs• Understandinghowtheabstractsyntaxtreeisbuilt• Grammarcontainsleftrecursion• Switchingbacktomathmodewith'$'

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 7

mhchem Grammar<ce> ::= <sentence> <sentence> ::= <phrase>

| <sentence> Space <phrase> <phrase> ::= <word>

| <word> <single macro> | <single macro> | '^' | '(^)'

<word> ::= <nonletter> | Letter | <single macro> <nonletter> | <word> <nonletter> | <word> Letter | <word> <single macro> <nonletter>

<nonletter> ::= '$' <LaTeX expression> '$'| ...

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 8

Testing• TestcasesforthenewfunctionalityAllexamplesofthemhchemmanualkeeptheirspacing• Coveragetesting100%codecoverageisdesired,currently76%oflinesarecovered• RegressiontestingThenewcodeshouldproducethesameoutputliketheoldcodeformathematicalformulae

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 9

Example• Input\ce{2Na + 2H2O -> 2Na+ + 2OH- + H-H}

• Outputbeforetheproject{\ce {2Na+2H2O->2Na++2OH-+H-H}}

• Outputnow{\ce {2Na + 2H2O -> 2Na+ + 2OH- + H-H}}

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 10

Outlook– Nextsteps• Finalizecode,colormodifierswithinchemicalformulaemightstillneedsomework• Improvecodecoveragebyaddingmoretestcases• Codereviewbyotherdevelopers• Mergeintotheupstreamcode• Productionuseofthenewcode

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 11

Conclusion• Grammarofmhchem isimplemented• Testsforchemicalformulaeexists• Itlookslikethenewcodehasn’tbrokenanythingelse

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 12

References• [1]MartinHensel:Themhchem Bundle,

DocumentationfortheLaTeX Packagesmhchemv4.04,hpstatement v1.02andrsphrase v3.11.2016

• [2]⚓ T140217Adjusttexvcjs topreventwhitespacemodificationsince-tags,2016.RetrievedMay14,2017fromPhabricator,wikimedia:phabricator.wikimedia.org/T140217

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 13

Questions?

2017-07-17 Schäfer,Manfred– Improvingtherenderingofchemicalformulae 14