Use flatMap
The example below will use a Observable to generate a String, where the String output is used as input parameter for another Observable to split the string within flatMap.
Since there is only one result, we could use Single instead.
Note: flatMap return Observable, while map return actual object.
public Observable<String[]> splitWords(final String phrase) { return Observable.fromCallable(new Callable<String[]>() { @Override public String[] call() throws Exception { return phrase.split("\\s+"); } });}public void testPassParameter() { Observable.fromCallable(new Callable<String>() { @Override public String call() throws Exception { return "I am here"; } }).flatMap(new Function<String, ObservableSource<String[]>>() { @Override public ObservableSource<String[]> apply(@NonNull String s) throws Exception { return splitWords(s); } }).subscribe(new Consumer<String[]>() { @Override public void accept(String[] strings) throws Exception { // 3 outputs: I, am, here for (String text : strings) { Log.d(TAG, text); } } });}Use blocking call
You can execute Observable sequentially using blocking call.
public Observable<String[]> splitWords(final String phrase) { return Observable.fromCallable(new Callable<String[]>() { @Override public String[] call() throws Exception { return phrase.split("\\s+"); } });}public Observable<String> generateString() { return Observable.fromCallable(new Callable<String>() { @Override public String call() throws Exception { return "I am here"; } });}public void testPassParameter() { Observable.fromCallable(new Callable<String[]>() { @Override public String[] call() throws Exception { // String phrase = "I am here"; String phrase = generateString().blockingSingle(); return splitWords(phrase).blockingSingle(); } }).subscribe(new Consumer<String[]>() { @Override public void accept(String[] strings) throws Exception { // 3 outputs: I, am, here for (String text : strings) { Log.d(TAG, text); } } });}