should section tests touching an api directly contrast coupling api?


i have combined array section tests test coupling around ftp server api.



both section tests ftp server same machine.



the coupling api gets deployed the height used both remoting web use scenarios. coupling api radically takes xml messages perform tasks such adding/deleting/updating users, changing passwords, modifying permissions...that kinda thing.



in section test, contend supplement user practical domain, i emanate xml summary send api. api does it's work advantage response station information presumably operation successful unsuccessful (error codes, validation failures etc).



to establish presumably api coupling formula unequivocally right thing (if response indicated success), i plead ftp server's com api query the store directly if, instance formulating user account, user comment unequivocally created.



does smell bad?



update 1: @jeremy/nick: coupling concentration testing, ftp server the com api 3rd jubilee products, presumably good tested stable. coupling api parse xml summary following plead ftp server's api. i verify, competence stupid case, sole ability user comment set justly wrapper. instance sourroundings wrong ability charge an ftp comment due typo coupling code. good instance being sourroundings upload download speed limits, competence reversed coupling code.



update 2: interjection answers. folks suggested controlling mocks, crossed mind, nonetheless light hasn't switched there nonetheless i'm still struggling conduct spin i coupling work ridicule ftp server. where mocks reside i pass an instance conspicuous mocks coupling api instead job com api? i'm wakeful scathing nonetheless struggling conduct spin it, mostly since i many examples tutorials summary (i'm ashamed say) verging incomprehensible.



Comments

Popular posts from this blog

list macos calm editors formula editors

how hibernate @any-related annotations?

why does floated <input> control floated component slip over too distant right ie7, nonetheless firefox?