# I'm looking for a more elegant way to solve the problem **URL:** \ **Category:** General\ **Tags:** discussion, how-to, software\ **Created:** [December 21, 2023, 11:01pm UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098 "2023-12-21T23:01:49Z")\ **Posts on this page:** 11\ **Page:** 1
**Author:** ![heiheshang](https://yyz2.discourse-cdn.com/free1/user_avatar/swi-prolog.discourse.group/heiheshang/32/4419_2.png) [@heiheshang](https://swi-prolog.discourse.group/u/heiheshang)\ **Post date:** [December 21, 2023, 11:01pm UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/1 "2023-12-21T23:01:49Z")
I’m making wrappers over rest api methods. my code represents similar pieces of code. I’m looking for a generalized method to wrap a rest api ```prolog 'crm.lead.update'(conf(AccessToken, ClientEndpoint), Fields, Reply) :- bitrix24_auth:check_token, format(atom(Url), '~w~w', [ClientEndpoint, 'crm.lead.update']), bearer(AccessToken, Bearer), bitrix24_request:post(Url, json( Fields ), Reply, [Bearer, status_code(StatusCode)]), (200 == StatusCode -> memberchk(result= @(true), Reply) ; debug(http(error), 'error ~q', [Reply]) ). ``` This code is repeated for all API methods. I’m looking for a way to get rid of this repetition. What if there is a way to write a generalized predicate? ---
**Author:** ![brebs](https://avatars.discourse-cdn.com/v4/letter/b/e9c0ed/32.png) [@brebs](https://swi-prolog.discourse.group/u/brebs)\ **Post date:** [December 22, 2023, 12:03am UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/2 "2023-12-22T00:03:53Z")
> [@heiheshang](#): > > What if there is a way to write a generalized predicate? Why wouldn’t there be? Could add another variable to indicate success/failure of the generalized predicate, as an argument. There’s specifics of how you want to handle failure, debugging info, errors raised, etc. ---
**Author:** ![heiheshang](https://yyz2.discourse-cdn.com/free1/user_avatar/swi-prolog.discourse.group/heiheshang/32/4419_2.png) [@heiheshang](https://swi-prolog.discourse.group/u/heiheshang)\ **Post date:** [December 22, 2023, 1:52am UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/3 "2023-12-22T01:52:47Z")
I don’t want to constantly write in code ‘crm.method1’() ‘crm.method2’() ‘crm.methodn’ I will have to wrap all the APIs in my code, how can I transform the call ‘crm.method’ into call (MethodName , args …). if there is no method call will simply fail ---
**Author:** ![peter.ludemann](https://yyz2.discourse-cdn.com/free1/user_avatar/swi-prolog.discourse.group/peter.ludemann/32/48_2.png) [@peter.ludemann](https://swi-prolog.discourse.group/u/peter.ludemann)\ **Post date:** [December 22, 2023, 2:15am UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/4 "2023-12-22T02:15:44Z")
Is this what you want? call/2 But this will throw an error if the predicate doesn’t exist; you can either set the `unknown` flag appropriately (see current\_prolog\_flag/2) or you can wrap it: ```prolog catch(call(C,...), error(existence_error(procedure,_),_), fail) ``` ---
**Author:** ![brebs](https://avatars.discourse-cdn.com/v4/letter/b/e9c0ed/32.png) [@brebs](https://swi-prolog.discourse.group/u/brebs)\ **Post date:** [December 22, 2023, 2:30am UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/5 "2023-12-22T02:30:31Z")
> [@heiheshang](#): > > ‘crm.method1’() ‘crm.method2’() Can use term\_expansion/2 to make it elegant. Kinda-relevant example: [prolog - What would be an elegant/idiomatic way to introduce a predicte on a list of atoms? - Stack Overflow](https://stackoverflow.com/a/77013370/) ---
**Author:** ![heiheshang](https://yyz2.discourse-cdn.com/free1/user_avatar/swi-prolog.discourse.group/heiheshang/32/4419_2.png) [@heiheshang](https://swi-prolog.discourse.group/u/heiheshang)\ **Post date:** [December 22, 2023, 2:38am UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/6 "2023-12-22T02:38:24Z")
no, that’s not what I want. Making a wrapper for a third-party rest api is a tedious task. I need a way to write to call ‘rest method’ (args …) in my program code:- predicat('rest method ', args …) I don’t want to implement every third-party rest api method. I need some syntactic sugar for this. I could do function\_post(Method Rest Api , args ) :- http\_client:post() and it works, but I would like to explicitly write ‘Method Rest Api’( in the code. If I explicitly implement each method, it’s a lot of boilerplate code. Explained how could I ---
**Author:** ![emiruz](https://yyz2.discourse-cdn.com/free1/user_avatar/swi-prolog.discourse.group/emiruz/32/4495_2.png) [@emiruz](https://swi-prolog.discourse.group/u/emiruz)\ **Post date:** [December 22, 2023, 8:12am UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/7 "2023-12-22T08:12:19Z")
Perhaps you can write down for us in pseudo-code what an ideal result would look like for you? That’d help to make the suggestions more relevant. ---
**Author:** ![heiheshang](https://yyz2.discourse-cdn.com/free1/user_avatar/swi-prolog.discourse.group/heiheshang/32/4419_2.png) [@heiheshang](https://swi-prolog.discourse.group/u/heiheshang)\ **Post date:** [December 22, 2023, 8:50am UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/8 "2023-12-22T08:50:13Z")
How do you solve the boilerplate problem? ---
**Author:** ![heiheshang](https://yyz2.discourse-cdn.com/free1/user_avatar/swi-prolog.discourse.group/heiheshang/32/4419_2.png) [@heiheshang](https://swi-prolog.discourse.group/u/heiheshang)\ **Post date:** [December 22, 2023, 8:55am UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/9 "2023-12-22T08:55:09Z")
I want to get rid of this code. ```prolog 'crm.lead.update'(conf(AccessToken, ClientEndpoint), Id, Fields) :- bitrix24_auth:check_token, format(atom(Url), '~w~w', [ClientEndpoint, 'crm.lead.update']), bearer(AccessToken, Bearer), bitrix24_request:post(Url, json( json([ id = Id, fields = Fields]) ), Reply, [Bearer, status_code(StatusCode)]), (200 == StatusCode -> memberchk(result= @(true), Reply) ; debug(http(error), 'error ~q', [Reply]) ). 'crm.lead.get'(conf(AccessToken, ClientEndpoint), Id, Reply) :- bitrix24_auth:check_token, format(atom(Url), '~w~w', [ClientEndpoint, 'crm.lead.get.json']), bearer(AccessToken, Bearer), bitrix24_request:post(Url, json( json([ id = Id ]) ), Reply, [Bearer]). 'crm.lead.list'(conf(AccessToken, ClientEndpoint), Body, Reply) :- bitrix24_auth:check_token, format(atom(Url), '~w~w', [ClientEndpoint, 'crm.lead.list.json']), bearer(AccessToken, Bearer), bitrix24_request:post(Url, Body, Reply, [Bearer]). 'crm.lead.productrows.get'(conf(AccessToken, ClientEndpoint), Body, Reply) :- bitrix24_auth:check_token, format(atom(Url), '~w~w', [ClientEndpoint, 'crm.lead.productrows.get.json']), bearer(AccessToken, Bearer), bitrix24_request:post(Url, Body, Reply, [Bearer]). ``` this is how I imagine it ```prolog Method(conf(AccessToken, ClientEndpoint), Id, Fields) :- call_method(Method, conf(AccessToken, ClientEndpoint), Id, Fields). ``` ---
**Author:** ![brebs](https://avatars.discourse-cdn.com/v4/letter/b/e9c0ed/32.png) [@brebs](https://swi-prolog.discourse.group/u/brebs)\ **Post date:** [December 22, 2023, 10:55am UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/10 "2023-12-22T10:55:47Z")
> [@heiheshang](#): > > How do you solve the boilerplate problem? By writing a program to create your predicates. Simple example: ```prolog ?- assert(mytest(X,Y) :- b(Y,X)). true. ?- listing(mytest). :- dynamic mytest/2. mytest(A, B) :- b(B, A). ``` ---
**Author:** ![ixmrm01](https://avatars.discourse-cdn.com/v4/letter/i/858c86/32.png) [@ixmrm01](https://swi-prolog.discourse.group/u/ixmrm01)\ **Post date:** [December 26, 2023, 7:29pm UTC](https://swi-prolog.discourse.group/t/im-looking-for-a-more-elegant-way-to-solve-the-problem/7098/11 "2023-12-26T19:29:47Z")
> > > Could this help you?