Bug 51853 – Add code to encode/decode the back/forward tree

• Status: RESOLVED
• Resolution: FIXED
• Component: Page Loading

[Darin Adler]

Add code to encode/decode the back/forward tree

[Darin Adler]

Created attachment 77851 [details]
Patch

[WebKit Review Bot]

Attachment 77851 [details] did not build on chromium:
Build output: http://queues.webkit.org/results/7231354

[Brady Eidson]

Comment on attachment 77851 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=77851&action=review

> WebCore/history/HistoryItem.cpp:642
> +        encoder->encodeString(child.m_title);

We probably don't need the titles for children.

> WebCore/history/HistoryItem.cpp:644
> +        encoder->encodeString(child.m_urlString);

If we don't need this for children, it'd simplify the code a lot (or a bit (or not at all))

> WebCore/history/HistoryItem.cpp:710
> +            return 0;

As above, probably don't need this.

> WebCore/history/HistoryItem.cpp:715
> +        recursionStack.append(DecodeBackForwardTreeRecursionStackElement(node.release(), i));

Need to store size here, too.

> WebCore/history/HistoryItem.cpp:778
> +        recursionStack.removeLast();

Move this down till you're done using "element"

> WebCore/platform/network/FormData.cpp:338
> +        encoder->encodeBytes(reinterpret_cast<const uint8_t*>(element.m_data.data()), element.m_data.size());

This cast probably isn't needed anymore.

> WebCore/platform/network/FormData.cpp:420
> +}

The encoder doesn't encode everything the decoder decodes....!!!!  Boo-urns

[Darin Adler]

Created attachment 77854 [details]
Patch

[Darin Adler]

Committed r74951: <http://trac.webkit.org/changeset/74951>

[WebKit Review Bot]

Attachment 77854 [details] did not build on chromium:
Build output: http://queues.webkit.org/results/7343188

[Build Bot]

Attachment 77854 [details] did not build on win:
Build output: http://queues.webkit.org/results/7282335

[WebKit Review Bot]

Attachment 77851 [details] did not build on chromium:
Build output: http://queues.webkit.org/results/7220396

[Gyuyoung Kim]

It seems to me that this patch makes a build break in WebKitEFL as below,


[ 99%] Building C object WebCore/CMakeFiles/webcore_efl.dir/platform/network/soup/cache/soup-directory-input-stream.c.o
/home/gyuyoung/webkit/working-source/WebKit-video/WebCore/platform/network/soup/ResourceHandleSoup.cpp: In function ‘bool WebCore::startHttp(WebCore::ResourceHandle*)’:
/home/gyuyoung/webkit/working-source/WebKit-video/WebCore/platform/network/soup/ResourceHandleSoup.cpp:617: error: no matching function for call to ‘WebCore::FormData::flatten(WTF::Vector<char, 0u>&)’
/home/gyuyoung/webkit/working-source/WebKit-video/WebCore/platform/network/FormData.h:114: note: candidates are: void WebCore::FormData::flatten(WTF::Vector<unsigned char, 0u>&) const
[ 99%] [ 99%] Building C object WebCore/CMakeFiles/webcore_efl.dir/platform/network/soup/cache/soup-http-input-stream.c.o
Building C object WebCore/CMakeFiles/webcore_efl.dir/platform/network/soup/cache/soup-request-data.c.o
make[2]: *** [WebCore/CMakeFiles/webcore_efl.dir/platform/network/soup/ResourceHandleSoup.cpp.o] Error 1

[WebKit Review Bot]

http://trac.webkit.org/changeset/74951 might have broken Leopard Intel Debug (Build)

[David Kilzer (:ddkilzer)]

(In reply to comment #10)
> http://trac.webkit.org/changeset/74951 might have broken Leopard Intel Debug (Build)

/Volumes/Big/WebKit-BuildSlave/leopard-intel-release/build/WebCore/history/HistoryItem.cpp:727: warning: implicit conversion shortens 64-bit value into a 32-bit value

A uint64_t variable is passed as a size_t parameter.

[David Kilzer (:ddkilzer)]

(In reply to comment #10)
> http://trac.webkit.org/changeset/74951 might have broken Leopard Intel Debug (Build)

World build fix:  r74952  <http://trac.webkit.org/changeset/74952>
Leopard build fix: r74955  <http://trac.webkit.org/changeset/74955>
Chromium build fix: r74956  <http://trac.webkit.org/changeset/74956>