我有一个函数，它将ID作为参数并进行HTTP调用，具体取决于ID . 它返回一个Observable . 我希望不同参数的调用完全独立，但它们相互影响 .
我用
RxJS 5.0.0-beta.2
Angular 2.0.0-beta.7用于http调用
打字稿1.8.2
Code:
getCharacterDetails (id : number) : Observable<CharacterDetails> {
let keys : SpreadsheetKeys = this.characters[id];
if (!keys) {
return null;
}
let frontUrl : string = `${BASE_URL}/${CELLS}/${keys.spreadsheetKey}/${keys.frontWorksheetKey}/${OPTIONS}`;
let backUrl : string = `${BASE_URL}/${CELLS}/${keys.spreadsheetKey}/${keys.backWorksheetKey}/${OPTIONS}`;
return Observable.forkJoin(
this.http.get(frontUrl).map(response => response.json()),
this.http.get(backUrl).map(response => response.json()))
.map((response : any[]) => {
let character : CharacterDetails = EMPTY_MODEL;
console.log(JSON.stringify(character)); // The 1st console.log
applyFrontSheetToCharacter(response[0], character);
applyBackSheetToCharacter(response[1], character);
console.log(JSON.stringify(character)); // The 2nd console.log
return character;
});
}
Expected behavior:
我打电话getCharacterDetails(1).subscribe((details) => { this.details = details }));
第一个console.log打印我的EMPTY_MODEL .
第二个console.log打印ID 1的字符模型 .
然后我打电话getCharacterDetails(2).subscribe((details) => { this.details = details });
第一个console.log打印我的EMPTY_MODEL .
第二个console.log打印ID 2的字符模型 .
Actual behavior:
我打电话getCharacterDetails(1).subscribe((details) => { this.details = details }));
第一个console.log打印我的EMPTY_MODEL .
第二个console.log打印ID 1的字符模型 .
然后我打电话给 getCharacterDetails(2).subscribe((details) => { this.details = details });
The 1st console.log prints the character model for ID 1. < - 问题
第二个console.log打印ID 2的字符模型 .
Why are the 2 calls not completely independent? How does the 2nd call even know about the data from the first call?
1 回答
问题很简单，我完全滥用了EMPTY_MODEL . 它基本上是我在.map（）函数中编辑的类范围变量 . 一个与RxJS，Angular或HTTP无关的愚蠢错误