http://www.purebasic.com
https://www.purebasic.fr/english/
Code: Select all
TickTimeout.q = ElapsedMilliseconds() + 1000 ; Timeout after 1sec
Request = HTTPRequest(#PB_HTTP_Post, "https://xxxxxxxxx/xxx/xxxService_v1", "Post Data here", #PB_HTTP_Asynchronous | #PB_HTTP_NoSSLCheck)
If Request
Debug "STARTED"
Repeat
Progress = HTTPProgress(Request)
Select Progress
Case #PB_HTTP_Success
Debug "SUCCESS"
Response$ = HTTPInfo(Request, #PB_HTTP_Response)
FinishHTTP(Request)
Break
Case #PB_HTTP_Failed
FinishHTTP(Request)
Response$ = "FAILED"
Debug "FAILED"
Break
Case #PB_HTTP_Aborted
FinishHTTP(Request)
Response$ = "ABORTED"
Debug "ABORTED"
Break
Default
Debug "PROGRESS = " + Progress
EndSelect
If ElapsedMilliseconds() > TickTimeout
Debug "ABOUT TO ABORT"
;--- HARD ABORT WHILE CONNECTING ---
If Progress = 0
FinishHTTP(Request)
Debug "HARD ABORT WHILE CONNECTING"
Break
EndIf
AbortHTTP(Request)
EndIf
Delay(250)
ForEver
EndIf
I would like to cancel the async request after a specific timeout, so after my period of waiting time I manually call AbortHTTP(). This gives the internal HTTP functions the possibility to end and cleanup, I would then expect to receive a #PB_HTTP_Abort status afterwards which lets me quit my loop. This works fine while downloading a file is in progress, but while connecting, this does not work, I do never receive an abort status.Fred wrote:Do you mean it never returned anything ?
That would be perfect!Fred wrote:I will add an HTTPTimeout() function to allow to specify one, as clearing the request as you did can lead to crash when the internal libcurl timeout is reached.
Too bad this feature did not make it into the release. Hopefully another beta will come soon.Fred wrote:I will add an HTTPTimeout() function to allow to specify one, as clearing the request as you did can lead to crash when the internal libcurl timeout is reached.
Fred wrote:Yes, we need to freeze the dev at one point or it will never have any release
