QWebView check on valid load

ThatDude

Well,
perhaps you should wait a bit longer... if you use QWebElement you could wait until that element shows on page.

Are you sure that data really hits your browser - you could use some tools to check that - Fiddler (web debugging proxy) or Wireshark (network protocol analyzer)

bagipro

Ok, i will check (fiddler doesn't fix it because i use proxy, i need to use wireshark or windows network monitor).

But see, there is all correctly?
@QObject::connect(frame,SIGNAL(urlChanged(QUrl)),&loopUrlChanged,SLOT(quit()));
QObject::connect(frame,SIGNAL(loadFinished(bool)),&loopLoad,SLOT(quit()));
QObject::connect(&timer,SIGNAL(timeout()),&loopUrlChanged,SLOT(quit()));
QObject::connect(&timer,SIGNAL(timeout()),&loopLoad,SLOT(quit()));@

bagipro

[quote author="bagipro" date="1374304254"]Ok, i will check (fiddler doesn't fix it because i use proxy, i need to use wireshark or windows network monitor).[/quote]

I remembered! My site using ssl (https), i can't fix it :(

ArcNexus

had to be server :p

some times yes, other times no..... mmmm that sounds at server problems....

bagipro

[quote author="ArcNexus" date="1374324206"] had to be server :p
some times yes, other times no..... mmmm that sounds at server problems....
[/quote]
Heh, I checked yandex.ru (normal size from 160 to 175 KB) using different proxy for each load and we see:
!http://clip2net.com/clip/m0/1374349499-clip-30kb.png(1)!
!http://clip2net.com/clip/m0/1374349970-clip-51kb.png(1)!

ThatDude

As I said...

you do not wait long enough
there is no way of knowing when the page is fully loaded
HTTPS - Webkit has some troubles with https esp. on windows. I never had any trouble with that on linux
Does you site use self-signed certificate? In that case you might want to add to to the list of know i.e. trusted authorities

Can you show us same yandex.ru requests/responses after you change your timer to wait e.g. 3 min per each request i.e. do not rely on loadFInished ...
Even ui->webView->page()->mainFrame() could fire multiple loadFinished signal if it has (static or dynamically created) internal frames or iframes
so sitting and waiting only for the very first loadFinished won't help you much
you might want to investigate different approaches using a timer:

wait for say 3 sec. without any other loadFinished for mainFrame
wait for say 3 sec. without any more data coming for that page/view
QWebPage::totalBytes
QWebPage::bytesReceived () const

bagipro

ThatDude,
Something like that
@void MainWindow::waitLoad(QWebPage* page)
{
qint64 bytes1 = page->bytesReceived(),bytes2;
QTimer timer;
QEventLoop loop;
QObject::connect(&timer,SIGNAL(timeout(QPrivateSignal)),&loop,SLOT(quit()));

do
{
    bytes2 = bytes1;
    timer.start(2000);
    loop.exec&#40;&#41;;
    bytes1 = page->bytesReceived();
}
while(bytes1 != bytes2);

}@

ThatDude

Yep!
That's like a double edged sword though - some pages just constantly load something in the background that's why I tend to prefer waiting for certain amount of time after last loadFinished (each new loadFinished just resets the timer)... this still suffers from things like
@<META HTTP-EQUIV="Refresh" CONTENT="1">@

in such cases a page timeout must be enforced

P.S. Monitor both loadStarted and loadFinished on QWebView i.e. for all frames (it creates during its lifetime) and increment and decrement a counter, when counter reaches 0 wait a few seconds then consider page loading has finished otherwise reset counter(s).
HTH

bagipro

Whats wrong?
@void MainWindow::waitLoad(QWebPage* page)
{
qint64 bytes=0;
QTimer timer;
QEventLoop loop;
QObject::connect(&timer,SIGNAL(timeout()),&loop,SLOT(quit()));

do
{
    qDebug() << bytes << " " << page->totalBytes();
    bytes  = page->totalBytes();
    timer.start(2000);
    loop.exec&#40;&#41;;
}
while( bytes != page->bytesReceived() );

}@

@0 502701
0 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238
50238 50238 @

ThatDude

You compare two different things
page->totalBytes()
AND
page->bytesReceived()
if you are downloading many pages simultaneously - you better keep count of these per-page

bagipro

Oh yeah, it's my fault ;)
But i checked and.. it did not succeed, i have many files like file1_start.html (after loopLoad and before that method, size of the file is 1 kb, 5% from all files) and file1_end.html (after loopLoad and after that method, size of the file is 1 kb) .

Many files have start size 24.1 kb and end size 25kb, but i'm not waiting for that result ;(

bagipro

!http://clip2net.com/clip/m0/1375083759-clip-48kb.png(bug)!

All time i had a trouble.
Idk, but after some hours of work my program crashes

ThatDude

You are lucky - mine crashed every few minutes with Qt5.x
You could try using Qt 4.8.5 instead
Based on my personal observations WebKit in Qt5 is quite buggy and crashes in 'random' places
I've also hit that bug at WTF::randomNumber but with an earlier version pre 5.1 ...and until today I thought they've fixed it in 5.1

The only QWebKit project I've ported to 5.1 I had to write a standalone app that restarts process when it crashes using QProcess etc.

bagipro

That's awful ;(
I planning to write this program on C# using standart liblary. Without dlls, bugs and random exceptions :3

bagipro

But i need to fix the bug
@Exception at 0x13fa54b, code: 0xc0000005: write access violation at: 0x1, flags=0x0 (first chance) in Qt5WebKit!WTF::randomNumber
First chance exceptions are reported before any exception handling.
This exception may be expected and handled.
eax=00000000 ebx=0001f000 ecx=00000000 edx=0012a504 esi=00000000 edi=0242d7d8
eip=013fa54b esp=0012a584 ebp=00010000 iopl=0 nv up ei pl zr na pe nc
cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 efl=00210246
Qt5WebKit!WTF::randomNumber+0x6b:
Missing image name, possible paged-out or corrupt data.
Missing image name, possible paged-out or corrupt data.
Missing image name, possible paged-out or corrupt data.
013fa54b 8935efbeadbb mov dword ptr ds:[0BBADBEEFh],esi ds:0023:bbadbeef=????????

*** ERROR: Symbol file could not be found. Defaulted to export symbols for C:\Qt\Qt5.1.0\5.1.0\msvc2010\bin\Qt5Core.dll -
s
sException at 0x13fa54b, code: 0xc0000005: write access violation at: 0x1, flags=0x0 (first chance)
dNOTE: INFERIOR SPONTANEOUS STOP
sОстановлено.
dState changed from InferiorRunOk(11) to InferiorStopOk(14) [master]
Resolving symbol: Qt5WebKit!WTF::randomNumber...
<x Qt5WebKit!WTF::randomNumber
<l-t
<l-s

Obtained 0x13fa4e0 for Qt5WebKit!WTF::randomNumber (#1)
<u 0x13fa4e0 0x13fa650
Source options are 4:
4/s - List source code at prompt
Source options are 0:
None

dNOTE: INFERIOR RUN REQUESTED
sПотребован запуск...
dState changed from InferiorStopOk(14) to InferiorRunRequested(10) [master]
<g

dNOTE: INFERIOR RUN OK
sВыполняется.
dState changed from InferiorRunRequested(10) to InferiorRunOk(11) [master]
(83c.f08): Access violation - code c0000005 (!!! second chance !!!)
s
sException at 0x13fa54b, code: 0xc0000005: write access violation at: 0x1, flags=0x0 in Qt5WebKit!WTF::randomNumber
eax=00000000 ebx=0001f000 ecx=00000000 edx=0012a504 esi=00000000 edi=0242d7d8
eip=013fa54b esp=0012a584 ebp=00010000 iopl=0 nv up ei pl zr na pe nc
cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 efl=00200246
Qt5WebKit!WTF::randomNumber+0x6b:
Missing image name, possible paged-out or corrupt data.
Missing image name, possible paged-out or corrupt data.
Missing image name, possible paged-out or corrupt data.
013fa54b 8935efbeadbb mov dword ptr ds:[0BBADBEEFh],esi ds:0023:bbadbeef=????????
s
sException at 0x13fa54b, code: 0xc0000005: write access violation at: 0x1, flags=0x0
dNOTE: INFERIOR SPONTANEOUS STOP
sОстановлено.
dState changed from InferiorRunOk(11) to InferiorStopOk(14) [master]
dUsing cached disassembly for 0x13fa54b (0x13fa4e0-0x13fa64f) in 'WTF::randomNumber'/ ''
@

!http://clip2net.com/clip/m0/1375639721-clip-33kb.png(1)!