summaryrefslogtreecommitdiffstats
path: root/src/network/access
Commit message (Collapse)AuthorAgeFilesLines
* QNetworkAccessManager: Backends were tried in wrong orderMarkus Goetz2010-05-141-1/+1
| | | | | | | | We inversed the order. HTTP shall be tried first, file:// last. See https://bugs.webkit.org/show_bug.cgi?id=38935 Reviewed-by: Thiago
* QNAM HTTP: Preemptive anti crash if() statementMarkus Goetz2010-05-131-0/+3
| | | | Task-number: QTBUG-10649
* QNAM HTTP: Only force read when actual bytes availableMarkus Goetz2010-05-121-1/+2
| | | | | | | This should fix the problem where we were showing a warning that didn't matter anyway. Task-number: QTBUG-9619
* QNAM HTTP: Start more requests in once function callMarkus Goetz2010-05-051-15/+13
| | | | Reviewed-by: Peter Hartmann
* QNAM HTTP: Fix invoking a method when being destructed right now (2)Markus Goetz2010-04-261-0/+7
| | | | | Task-number: QTBUG-10171 Reviewed-by: Olivier Goffart
* QNAM: Use a reference in appendDownstreamDataMarkus Goetz2010-04-261-1/+1
|
* QNAM HTTP: Divide QNetworkReplyImplPrivate::appendDownstreamDataMarkus Goetz2010-04-262-25/+53
| | | | | | The goal is to easier add an overload for (const &QByteArray). Reviewed-by: Peter Hartmann
* QNAM HTTP: Avoid one copyMarkus Goetz2010-04-261-2/+1
| | | | | | Even if we have implicit sharing in QByteArray it makes sense. Reviewed-by: joao
* QNAM: Add a code comment related to the cacheMarkus Goetz2010-04-211-0/+5
|
* QNAM HTTP: Fixed a bug when getting empty files with pipeliningMarkus Goetz2010-04-211-1/+1
| | | | | Instead of returning from the socket's readyRead() handler break out of the switch() and continue parsing the data.
* QNAM HTTP: Pipelining improvementMarkus Goetz2010-04-201-1/+5
| | | | | | | Fail earlier. Use constant that was introduced in one of the previous commits. Reviewed-by: Peter Hartmann
* QNAM HTTP: Pipelining changesMarkus Goetz2010-04-153-42/+60
| | | | | | | | | Re-wrote some code to improve pipelining efficiency. Greatly helps with combining into one TCP packet. Task-number: QTBUG-9894 Task-number: QT-3280 Reviewed-by: Peter Hartmann
* QFtp: Fix possible crashMarkus Goetz2010-04-141-0/+6
| | | | Task-number: QTBUG-7359
* QNAM HTTP: Do not pipeline with WebLogic serversMarkus Goetz2010-04-131-0/+2
|
* Document Symbian platform security requirements on Qt APIsFrans Englich2010-04-121-0/+10
| | | | | | | | | Work done jointly by Gareth and me. Yields no qdoc errors. Task-number: QTBUG-9342 Task-number: QTBUG-9120 Reviewed-by: Gareth Stockwell Reviewed-by: David Boddie
* Fix compile error with QT_NO_NETWORKDISKCACHE in QtNetworkTasuku Suzuki2010-03-312-1/+6
| | | | | Merge-request: 2332 Reviewed-by: Peter Hartmann
* Fix compile error on Symbian 9.1, caused in network/access.Frans Englich2010-03-262-14/+33
| | | | | | | | | | | RVCT has trouble when QPointer is instantiated on classes with inlined constructors/functions. Compile fix for 43e1cffb16c2eea54392f5c56210b10abb2f044e, cc22a14f3d2159a8f760b44e3fe1636a3721ce95 and 656c02f128c56177c48b3de47f7b1e17dbbfa4d3 Reviewed-by: Peter Hartmann Reviewed-by: Prasanth
* QNAM HTTP: Symbian compile fixMarkus Goetz2010-03-251-0/+1
|
* QNAM HTTP: Fix crazy crash when exiting applicationMarkus Goetz2010-03-252-2/+2
| | | | | | | | | | | | | This fixes a crash in tst_qnetworkreply and in some QtWebKit based browsers. It was related to adding a QPointer on an already deleted QHttpNetworkConnection object. By converting an existing normal pointer to a QPointer we set it a 0-pointer and are safe. https://bugs.webkit.org/show_bug.cgi?id=36290 Reviewed-by: joao Reviewed-by: gabi
* QNAM HTTP: Do not use TCP_NODELAYMarkus Goetz2010-03-231-1/+8
| | | | Reviewed-by: thiago
* QNAM HTTP: Fix no-headers and HTTP-100 handlingMarkus Goetz2010-03-163-4/+17
| | | | Reviewed-by: Thiago
* QNAM HTTP: Fix invoking a method when being destructed right nowMarkus Goetz2010-03-021-1/+5
| | | | Reviewed-by: Olivier Goffart
* QNAM HTTP: Switch back to QueuedConnection for requeueRequest.Markus Goetz2010-02-241-2/+16
| | | | Reviewed-by: Peter Hartmann
* QNAM HTTP: Set some values after clear() in a QHttpNetworkReplyMarkus Goetz2010-02-241-0/+2
| | | | Reviewed-by: TrustMe
* QNetworkAccessManager: Documentation enhancementsMarkus Goetz2010-02-241-1/+8
| | | | | Reviewed-by: Peter Hartmann Reviewed-by: David Boddie
* QNetworkRequest: Doc enhancementMarkus Goetz2010-02-221-0/+2
| | | | Reviewed-by: David Boddie
* QNAM HTTP: Clean code a bitMarkus Goetz2010-02-182-29/+45
| | | | | | .. and some changes that will help us track down bugs Reviewed-by: Peter Hartmann
* QNAM HTTP: Always set channel.reply to 0 when doneMarkus Goetz2010-02-185-22/+27
| | | | | | | This will help us to track down a sporadic bug that made requests being sent again even if they were already done. Reviewed-by: Peter Hartmann
* HTTP backend: remove commentPeter Hartmann2010-02-171-1/+1
| | | | | | the date header always needs to be updated in the cache. Task-number: QTBUG-7466
* QNAM HTTP: Unify socket EOF handling a bitMarkus Goetz2010-02-162-24/+38
| | | | Reviewed-by: Peter Hartmann
* doc: Fixed several typos.Martin Smith2010-02-121-1/+1
| | | | Task: QTBUG-7997
* QNAM HTTP: Optimize detectPipeliningSupport()Markus Goetz2010-02-111-7/+7
| | | | | | Do cheap checks first. Reviewed-by: TrustMe
* QNAM HTTP: Set channel to IdleState at better placeMarkus Goetz2010-02-111-3/+3
| | | | | | | | | | The readyRead() from the expand() function called from allDone() had a new request triggered for that channel even if it is not really finished yet. Move the assignment to IdleState inside allDone to avoid this error. Reviewed-by: Andreas Kling Reviewed-by: Peter Hartmann
* QNAM HTTP: Forgot to remove a qDebug()Markus Goetz2010-02-091-1/+1
| | | | Reviewed-by: TrustMe
* QNAM HTTP: Reserve bytes for HTTP parsingMarkus Goetz2010-02-091-0/+12
| | | | Reviewed-by: Peter Hartmann
* QNetworkCookie(Jar): fix includesPeter Hartmann2010-02-092-2/+2
| | | | | | we need to include the module preefix in public headers Reviewed-by: Denis Dzyubenko
* QNAM HTTP: Improve readHeader() of the HTTP replyMarkus Goetz2010-02-091-11/+25
| | | | | | | Just read() instead us using peek() and bytesAvailable() Also don't use the obfuscated line ending checking code. Reviewed-by: Peter Hartmann
* QNAM HTTP: Improve parseStatus() of HTTP replyMarkus Goetz2010-02-091-10/+12
| | | | | | Just read() instead us using peek() and bytesAvailable() Reviewed-by: Peter Hartmann
* QNAM HTTP: Fix issue with some serversMarkus Goetz2010-02-091-2/+5
| | | | | | | | | The call to eatWhitespace() must only be done before reading the status, not while reading it. (We can be in ReadingStatusState for multiple loop iterations) Task-number: QTBUG-7673 Reviewed-by: Peter Hartmann
* QNAM HTTP: Fix bug related to re-sending a requestMarkus Goetz2010-02-033-10/+27
| | | | | | | Thanks Nils Jeisecke for report and proposed fix. Task-number: QTBUG-7713 Reviewed-by: thiago
* Fix some foreach usage to use const refsMarkus Goetz2010-01-282-2/+2
| | | | Reviewed-by: joao
* Core classes, examples, demos: Some changes from string to charMarkus Goetz2010-01-281-1/+1
| | | | Reviewed-by: Peter Hartmann
* QFileNetworkReply: Use a QFileEngineMarkus Goetz2010-01-252-19/+38
| | | | | | Slight performance increase. Reviewed-by: Peter Hartmann
* QNAM HTTP: make it give some better error messages in case of socket failure.Thiago Macieira2010-01-213-5/+7
| | | | Reviewed-By: Peter Hartmann
* Enhance QFileNetworkReplyMarkus Goetz2010-01-184-65/+48
| | | | | | | | It is now considered as finished immediatly after creation to avoid an event loop spin. Had to change the auto tests a bit to account for the new behaviour. Reviewed-by: Peter Hartmann
* network internals: fix uploading of dataPeter Hartmann2010-01-141-1/+1
| | | | Reviewed-by: Markus Goetz
* Fix an issue with HTTP headers like "private, max-age=300".Thiago Macieira2010-01-121-1/+1
| | | | | | | | | | | The parsing would ignore the fact that "private" ended at the comma and would instead try to get the continuation as the value. We have to check if equal < comma and it's not -1. Using unsigned comparisons does both things at once for us. Task-number: QTBUG-7060 Reviewed-by: Peter Hartmann
* QNAM HTTP Code: Cache socket state inside ensureConnection()Markus Goetz2010-01-111-4/+6
| | | | Reviewed-by: TrustMe
* QNAM HTTP Code: When starting new request, prefer connected sockets.Markus Goetz2010-01-111-5/+17
| | | | | | | Prefer a QHttpNetworkConnectionChannel which has a connected socket to a channel which doesn't have a connected socket. Reviewed-by: Peter Hartmann
* QNAM HTTP: Don't call d_func() so oftenMarkus Goetz2010-01-111-27/+30
| | | | Reviewed-by: Peter Hartmann
generated by cgit v0.12 at 2025-03-07 07:41:35 (GMT)