Im working on a project that has scala backend and scalatra as a servlet. For now, the backend is running inside a Vagrant box, and sbt is used to build the program.
My problem is that I need to handle incoming Json-formatted data, but I cant extract it properly - though I'm able to send Json data. First I played around with Scalas own Json parser, but realized that it's deprecated, so I had to find external library. After that I came up with json4s.
In my app I have the following endpoint listening for incoming post requests
post("/my/endpoint") {
// run the code in here
}
And that works just fine. I can put trivial println inside the brackets and it displays in the sbt shell after the endpoint has been called.
After that I defined a case class to be used to extract the data
case class Person(name : String) {
override def toString() = s"My name is $name"
}
And lastly I did the following code inside my endpoint function
post("/my/endpoint") {
val json = parse(request.body)
val person = json.extract[Person]
println(person)
}
So right now when calling the println it should print whatever the toString method of Person class returns, but it prints nothing.
Lets take a step back and modify the code so that we have the following
post("/my/endpoint") {
val json = parse(request.body)
println(json)
}
This piece of code prints the following into sbt shell
JObject(List((name,JString(joe))))
So we can see that the parse-function is actually doing something and data is being received.
But, if we add
person = json.extract[Person]
after println(json), the endpoint stops functioning. Why I think it's weird, is because it's not a compiling error, so everything before that line should work properly? Also it doesn't cause the program to crash on runtime, nor does it give any errors, warnings, etc. Every other endpoint and functions works still properly.
Also, I did try the example in json4s.org under the section 'Extracting values'.
I copied the code from word to word, but that did not work either.
Also FYI, my versions are SBT, 1.1.6, Scala 2.12.6, Scalatra 2.6.3, json4s 3.5.2 and Vagrant is using Xenial 16.04 as its basebox.
pls help mi
Related
You will find instructions to reproduce on your own device at the bottom.
I have a basic Angular project I created using Angular CLI, running on TypeScript 3.1.3, with nothing much added aside a class and a json file.
I created a class ResourcesService with the following command with Angular CLI:
ng generate service Resources
I'm basically using it to load json files, as a mean of localising (instead of using Angular unfinished builtin internationalisation features).
The following is my class, as well as the json file:
ResourcesBundle.json
{
"label.changeLanguage": "Change language",
"label.education": "Education",
"label.experience": "Experiences",
"label.skills": "Skills",
"label.summary": "Summary",
"label.language.english": "English",
"label.language.french": "French"
}
resources.service.ts
import * as resources from '../assets/resources/ResourcesBundle.json';
#Injectable({
providedIn: 'root'
})
export class ResourcesService {
constructor() {}
public getString(label: string): string {
let resource: string = resources[label];
return resource;
}
}
Of course, in order to be able to import the json file that way, I've set "resolveJsonModule": true in tsconfig.json.
The service by itself is working properly. I can inject it and call the getString method, and it's running without any error.
However, no matter what value I pass to the getString method, the returned value is always undefined. I've even tried to hard code the value for label = 'label.summary', but it's still returning undefined. The only time it's working properly is when I write the string directly between the brackets:
let resource: string;
label = 'label.summary';
resource = resources[label]; // resource == undefined
resource = resources['label.summary']; // resource == 'Summary'
Within the TS on VSCode, the content of resources is as following:
label.changeLanguage
label.education
label.experience
label.language.english
label.language.french
label.skills
label.summary
When using console.log(resources), the console was displaying something like this on Firefox:
Object {
label.changeLanguage: "Change language"
label.education: "Education"
label.experience: "Experience"
label.language.english: "English"
label.language.french: "French"
label.skills: "Skills"
label.summary: "Summary"
}
So the json is properly loaded, but it can only be used with hard coded string.
The only other solution I found was to give up the json file and initialise an indexed type directly in the code:
private resources: { [key: string]: string } = {
'label.changeLanguage': 'Change language',
'label.education': 'Education',
'label.experience': 'Experiences',
'label.skills': 'Skills',
'label.summary': 'Summary',
'label.language.english': 'English',
'label.language.french': 'French'
};
However, I don't think that's a good approach, at all...
In the case of the imported json file, why does it always return undefined when I use a variable? Or otherwise, why does it work only with a hard coded string between the brackets?
Edit:
You will find below a stackblitz link to a demo project:
https://stackblitz.com/edit/angular-h2aspf?file=tsconfig.json
If you run it on the browser, it will work properly (the console will properly display Change language).
However, if you download it and run it locally, you will notice that the console will display undefined instead.
To run it locally:
You must have npm and Angular CLI
Download and unzip the stackblitz demo in a folder
Run npm i in the project folder
Run ng serve --open
Open the console on your browser, it should be displaying undefined, instead of the expected value (Change language on stackblitz)
Edit:
According to a comment on the Angular CLI issue, a workaround is to set "esModuleInterop": true in tsconfig.json, and to change the import statement from:
import * as resources from '../assets/resources/ResourcesBundle.json';
To this:
import resources from '../assets/resources/ResourcesBundle.json';
Original answer:
After checking multiple times on different devices, I think this is a bug directly related to Angular (current version: 7.0.2).
To take the example I gave in the question again:
https://stackblitz.com/edit/angular-h2aspf?file=tsconfig.json
On the browser, this demo is outputting Change language on the console.
On locale device:
Download and unzip the stackblitz demo in a folder
Run npm i in the project folder
If you run with ng serve, you will notice undefined in the web browser console
Stop Angular, then run again with ng serve --prod. The web browser console is now properly outputting Change language
I've opened the following issues for Angular and Angular CLI projects on GitHub for this problem:
Angular: Issue #26785: Imported json file as indexed type always giving undefined when Angular is not running in production mode
Angular CLI: Issue #12781: Imported json file as indexed type always giving undefined, but not when running ng serve --prod
const TelegramBot= require('./telegram-bot') // It's currently only local.
var bot = new TelegramBot()
console.log(bot)
// This does not print a complete JSON of the object. It misses stuff like
constructor, method, prototype and super_.
Is there some way or npm module that prints a JSON compatible output of the object?
My only work around so far is console logging it out like this and repeatedly checking the log and printing out another but I think it'll be a lot of easier by having a JSON export and using a JSON online viewer that views like a directory tree.
console.log(`
TelegramBot:
> ${Object.getOwnPropertyNames(TelegramBot)}
TelegramBot.prototype:
> ${Object.getOwnPropertyNames(TelegramBot.prototype)}
TelegramBot.prototype.constructor:
> ${Object.getOwnPropertyNames(TelegramBot.prototype.constructor)}
TelegramBot.prototype.constructor.super_:
> ${Object.getOwnPropertyNames(TelegramBot.prototype.constructor.super_)}
`)
I'm aware functions can't be seen with JSON.parse(). I don't mind if they appear as a string like "Anonymous Function()" or "FunctionWithAName()". Or something like this.
I'm doing this since I'm having another go trying to learn prototypes and I've used util.inherits(TelegramBot, EventEmitter) in the TelegramBot object.
To avoid name clashes between TelegramBot methods I've made and the super class of EventEmitter names. I'd like to keep a clear view of the whole object structure. Or do I not have to worry since they use this variable shadowing thing? If I'm correct it checks the object's instance first, then it's prototype. Not sure if EventEmitter prototype checked first or TelegramBot's.
I'm experiencing a discrepancy between the first compilation of a Grails app and the compilation that happens when a file changes while the app is running.
Background:
My app creates some spring beans from Spring LDAP (docs) using conf/spring/resources.groovy.
I have an LdapUser.groovy class in src/groovy (I'm using it similarly to a domain class, except it isn't in grails-app/domain as it doesn't map to a database table).
In BootStrap.groovy I register a JSON marshaller for LdapUser (using JSON.registerObjectMarshaller).
I have a controller with an index method that responds a list of LdapUser objects. This renders correctly in JSON (according to the marshaller).
With that background, here are the pieces of the problem:
When the show method, which responds a single LdapUser, gets called, I get an exception that LdapUser cannot be converted to grails.converters.JSON. (fair enough)
But, if I save the LdapUser.groovy file, thus invoking a recompile on the file while the app is running, the JSON marshaller suddenly works fine.
Before saving the LdapUser.groovy, my controller has a to an LdapUserRepo (a class instantiated via an #EnableLdapRepositories annotation on the controller), but this reference becomes null after I save LdapUser.groovy. I'm not sure how this relates to the problem, as I was also able to reproduce the problem in a controller lacking an injected LdapUserRepo (but with the annotated controllers still in the app).
I also at one point was setting an asType method on the LdapUser class, which was called as expected before the save-invoked recompile. After the recompile, however, my asType method was no longer called and the JSON marshaller was taking over. ( I was doing exception-worthy things in the asType that were throwing before recompile and not throwing after... )
My understanding of the problem is therefore:
Somehow the asType method of the LdapUser.groovy class is not being automatically generated on first compile when running the app, but is being generated on subsequent compiles.
The LdapUser class is tied to the LdapUserRepo in more ways than merely being a type the Repo uses, and the recompile is not reflecting that connection correctly.
Methods rendering lists of objects are somehow unaffected by the asType method. This leads me to believe that the JSON marshaller gets called directly on list elements (instead of via asType) when the list asType has been called (whether or not the "as" operation is implicit...).
My question then is:
what is the Grails compiler doing differently on run-app vs on compile while app is running that could be causing this behavior?
how can I restructure things to ensure it works properly out of the box?
If I need to RTFM, what would be the FM section? (My google-fu is sadly quite weak).
Note: this question is vaguely similar, but doesn't have any meaningfulness to the answer:
Grails: Defining a JSON custom marshaller as static method in domain
I have a Grails service that sends out e-mails using a 3rd-party service by doing a HTTP call:
class EmailService {
def sendEmail(values) {
def valueJson = values as JSON
... // does HTTP call to 3rd party service
}
}
I've written a unit test to test this service (because an integration test spins up Hibernate and the entire domain framework, which I don't need):
#TestFor(EmailService)
class EmailServiceTests {
void testEmailServiceWorks() {
def values = [test: 'test', test2: 'test2']
service.sendEmail(values)
}
}
However, when I execute this unit test, it fails with this exception when it tries to do the as JSON conversion:
org.apache.commons.lang.UnhandledException: org.codehaus.groovy.grails.web.converters.exceptions.ConverterException: Unconvertable Object of class: java.util.LinkedHashMap
I then re-wrote my unit test to just do the following:
void testEmailServiceWorks() {
def value = [test: 'test', test2: 'test2']
def valueJson = value as JSON
}
And I get the same exception when it tries to do the as JSON conversion.
Does anyone know why I'm getting this exception, and how I can fix it?
Even though you are testing a service, you can apply the #TestMixin(ControllerUnitTestMixin) annotation to your test class to get Grails to set up the JSON converter.
The as JSON magic is created when the domain framework spins up.
You have to either change your test to an integration one or mock the asType.
def setUp(){
java.util.LinkedHashMap.metaClass.asType = { Class c ->
new grails.converters."$c"(delegate)
}
}
Rember to clean up after yourself in the tearDown, you wouldn't want metaprogramming leaks in your test suite.
def tearDown(){
java.util.LinkedHashMap.metaClass.asType = null
}
Edit:
If you come from the future, consider this answer: https://stackoverflow.com/a/15485593/194932
As Grails 3.3.x grails-test-mixins plugin is deprecated. #see migration guide.
For this problem you should implement GrailsWebUnitTest which is coming from Grails Testing Support Framework.
you can initialise the JSON in the setUp() . There are various marshallers which implement ObjectMarshaller , which need to be added to the ConverterConfiguration for JSON conversion to work.
http://grails.github.io/grails-doc/2.4.4/api/index.html?org/codehaus/groovy/grails/web/converters/marshaller/json/package-summary.html
example :
DefaultConverterConfiguration<JSON> defaultConverterConfig = new DefaultConverterConfiguration<JSON>()
defaultConverterConfig.registerObjectMarshaller(new CollectionMarshaller())
defaultConverterConfig.registerObjectMarshaller(new MapMarshaller())
defaultConverterConfig.registerObjectMarshaller(new GenericJavaBeanMarshaller())
ConvertersConfigurationHolder.setTheadLocalConverterConfiguration(JSON.class, defaultConverterConfig);
I just ran into this, and I really didn't want to implement GrailsWebUnitTest as recommended in another answer here. I want to keep my service test as "pure" and lean as possible. I ended up doing this:
void setupSpec() {
defineBeans(new ConvertersGrailsPlugin())
}
void cleanupSpec() {
ConvertersConfigurationHolder.clear()
}
This is how it happens under the hood when you implement GrailsWebUnitTest (via WebSetupSpecInterceptor and WebCleanupSpecInterceptor).
That said, the converters seem to be meant for use in the web tier, primarily for making it easy to transparently return data in different formats from a controller. It's worth considering why the service you're testing needs the converters in the first place.
For example, in my case, someone used the JSON converter to serialize some data to a string so it could be stored in a single field in the database. That doesn't seem like an appropriate user of the converters, so I plan on changing how it's done. Making the converters available in my service test is a temporary solution to allow me to improve our test coverage before I refactor things.
I was getting the same error when trying to unit test a controller that calls "render myMap as JSON". We use Grails 1.3.7 and none of the other solutions worked for me without introducing other problems. Upgrading Grails was not an alternative for us at the moment.
My solution was to use JSONBuilder instead of "as JSON", like this:
render(contentType: "application/json", {myMap})
See http://docs.grails.org/latest/guide/theWebLayer.html#moreOnJSONBuilder
(I realize this is old, but came here in search for a solution and so might others)
I'm getting this error:
Compilation error [package views.json.Runs does not exist]
when it clearly does exist. I can't figure out what I could be doing wrong.
the action in the Runs controller:
#BodyParser.Of(play.mvc.BodyParser.Json.class)
public static Result view(Long task_id, Long run_id) {
Run run = Run.find.byId(run_id);
return ok(views.json.Runs.view.render(run));
}
app/views/Runs/view.scala.json:
#(run: Run)
{
"started": "#run.started",
"completed": "#run.completed"
}
I've gotten a few html templates working, but this is the first JSON template I've tried with 2.0. I'm not sure what else to try since it's already about as basic as possible. Anyone have any ideas?
Update: I've gotten several suggestions for workarounds, but I'm still interested in knowing how to get the templates working, if only to better understand the changes to 2.0.
Only html, xml and txt appear to be supported by default. For other file extensions and formats, you'd have to register additional "templateTypes" in $PLAY_HOME/framework/src/sbt-plugin/src/main/scala/PlaySettings.scala (see also: SBT Settings, near the bottom).
It may be helpful to look at the standard template types definitions which are in $PLAY_HOME/framework/src/play/src/main/scala/play/api/templates/Templates.scala.
You could also cheat and serve your json from a txt file, but do response().setContentType("application/json") before calling the render method.
For Json, why don't you directly produce a Json string using the Json helper:
public static Result view(Long task_id, Long run_id) {
Run run = Run.find.byId(run_id);
return ok(play.libs.Json.toJson(run));
}
I recommend following the documentation and let the Json library serialize your data instead of writing the Json text on your own: See Serving Json Response.
I still don't understand, why does people want to render theirs JSON with views.
Note: this is the same way as #nico_ekito showed before and I agree with him totally , below code just demonstrates how to select part of the object for JSON
public static Result view(Long task_id, Long run_id){
Run run = Run.find.byId(run_id);
ObjectNode result = Json.newObject();
result.put("started", run.started);
result.put("completed", run.completed);
return ok(Json.toJson(result));
}