Revert some changes in PacketListModel::headerData from c5fb4022
to preserve more optimized code.
Change-Id: If708999a6d446d70eca7414670dec0c618190fe0
Reviewed-on: https://code.wireshark.org/review/12058
Reviewed-by: Stig Bjørlykke <stig@bjorlykke.org>
Added get_column_tooltip() to use common code in GTK and Qt.
Change-Id: I2f6ce95e2e129752bbb958a28aec6f42aa81be3d
Reviewed-on: https://code.wireshark.org/review/12047
Petri-Dish: Stig Bjørlykke <stig@bjorlykke.org>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Add a timeShifted signal to TimeShiftDialog and use it to update the
packet list and model. Add drawCurrentPacket to PacketList so that we
can do a more thorough job of redrawing the current packet and tree.
Bug: 11575
Change-Id: I960d8cdbf6872e3f71007cb4d2bbd5457f268257
Reviewed-on: https://code.wireshark.org/review/11068
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>
Petri-Dish: Pascal Quantin <pascal.quantin@gmail.com>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Gerald Combs <gerald@wireshark.org>
In tests here using GTK+ 2.24 and 3.10, GtkTreeView handles multi-line
items by adjusting the height for all rows, but only after the number of
multi-line items exceeds some sort of threshold. For a packet capture
which contains a few DNS packets and a lot of TCP packets, if I change
"Standard query" to "Standard\nquery" in packet-dns.c I get
single-height packet list items. If I change "[TCP segment of a
reassembled PDU]" to "[TCP segment of a\nreassembled PDU]" in
packet-tcp.c (which results in more multi-line column strings) I get
double-height packet list items.
The current Qt code initially sets the uniformRowHeights property then
falls back to variable row heights if we run across a multi-line column
string. This adds a lot of logic which can impact other functionality
(e.g. column widths) and recalculating row heights is painfully slow for
large numbers of packets.
Instead of trying to manage variable row heights, always enable
uniformRowHeights. Track the maximum newline count and trigger a row
height adjustment when it changes. This mimics the GTK+ UI behavior,
although it should be more reliable.
Note that we need to adjust some numbers in RelatedPacketDelegate.
Change-Id: I289e963b6f00338c4374e602fa3fc83d04554519
Ping-Bug: 11515
Ping-Bug: 10924
Reviewed-on: https://code.wireshark.org/review/10628
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Features such as sorting and scroll bar colorization require
fully-dissected packets. We currently do dissection at the wrong time --
*after* the user clicks on a packet list column header or moves the
scrollbar.
Add a timer + slot that dissects packets when the UI is idle so that our
packets are at least partially dissected when we need them.
Change-Id: I024c590af2250d67404a520f118e46ec0c49cd71
Reviewed-on: https://code.wireshark.org/review/10593
Petri-Dish: Gerald Combs <gerald@wireshark.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Gerald Combs <gerald@wireshark.org>
beginInsertRows + endInsertRows is expensive. Instead of calling them
each time we add a packet to the list, queue up a list of visible packets
and flush it during the next UI update.
Assume that none of our column data has newlines. Enable
uniformRowHeights and only disable it when we need to. Note that this
requires further work.
Ping-Bug: 11515
Ping-Bug: 10924
Change-Id: Ifbdd2964b174247a4745d4889ebda5bf3b886ba4
Reviewed-on: https://code.wireshark.org/review/10553
Petri-Dish: Gerald Combs <gerald@wireshark.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Limit PacketListModel::recreateVisibleRows to checking for packets that
passed the display filter. Remove the call to resetColumns, which
invalidates the column string cache.
Call PacketListModel::resetColumns from PacketList::columnsChanged.
Along with invalidating the column string cache, this triggers a widget
update which means we shouldn't need to call recreateVisibleRows or
redrawVisiblePackets. Remove them.
Change-Id: Ic25f139e2c2f8cdebb2a868ccda7eb4127a3013c
Ping-Bug: 11324
Ping-Bug: 11514
Reviewed-on: https://code.wireshark.org/review/10491
Petri-Dish: Gerald Combs <gerald@wireshark.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Pascal Quantin <pascal.quantin@gmail.com>
Reviewed-by: Gerald Combs <gerald@wireshark.org>
This saves a fair amount of memory in tests here. Loading a large
capture file and sorting on a custom column (tcp.window_size) uses 676
MB before the change and 634 after.
Add notes about possble further improvements:
Roll our own replacement for GStringChunks using wmem_tree.
Have PacketListRecord::columnString return a const char * instead of a
const QByteArray.
Change-Id: Icb36194f5ad290828d7106ccc3bf494d07d76d08
Reviewed-on: https://code.wireshark.org/review/10476
Reviewed-by: Gerald Combs <gerald@wireshark.org>
This fixes a serious (and obvious) memory leak in the packet list.
Change-Id: Ie3d21e1a1d09a3481c9649954c25b9ae73b8e32a
Reviewed-on: https://code.wireshark.org/review/10485
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Call columnsChanged (which invalidates cached column strings) when we
change our name resolution preferences from the View menu.
Remove the last sorted column tracking from gf19a173 while we're here
(it didn't work properly).
Bug: 11468
Change-Id: I7fea58d702b283028235d023f27ab0336d8643a4
Reviewed-on: https://code.wireshark.org/review/10301
Petri-Dish: Gerald Combs <gerald@wireshark.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Gerald Combs <gerald@wireshark.org>
The GTK+ UI sequentially dissects and caches column strings for all rows
before sorting a column. Do the same in the Qt UI, which can improve
performance considerably.
Don't colorize packets when sorting in the Qt UI unless it's necessary.
When sorting in the Qt UI, let the user cancel the initial packet
dissection. Note that we'll need to replace std::sort in order to
cancel out of sorting.
Use a pre-allocated and pre-compiled GRexex when we prime columns. Note
that we probably shouldn't parse a regular expression there.
Cache the last result of proto_registrar_get_byname.
Note performance hot spots elsewhere in the code.
To do:
GeoIP in packet-ip.c is pretty slow.
Bug: 11467
Change-Id: Ib34038fee08ef0319261faeffc4eca01e52f4bd3
Reviewed-on: https://code.wireshark.org/review/10275
Petri-Dish: Gerald Combs <gerald@wireshark.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Fix build error:
ui/qt/packet_list_model.cpp:270:5: error: 'sort' is not a member of 'std'
std::sort(visible_rows_.begin(), visible_rows_.end(), recordLessThan);
Change-Id: I3a577a268f6c12e8fd97b7b6fd2429989c28e2f5
Reviewed-on: https://code.wireshark.org/review/10092
Petri-Dish: Michal Labedzki <michal.labedzki@tieto.com>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Alexis La Goutte <alexis.lagoutte@gmail.com>
Process application events as little as possible while sorting.
Change-Id: I9d779aebf6edaafc1e6af3a99b4a0dac682ea6b6
Reviewed-on: https://code.wireshark.org/review/9722
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Rename CaptureFileProgressFrame to ProgressFrame. It's not limited to
capture files. Add "busy" routines there and in MainStatusBar. Show a
busy indicator while sorting columns.
Use toByteArray().constData instead of .data. I'm not sure if it matters
in our case but it's more correct.
Change-Id: Ibe35fee9b9dd3fabb5ff8ddcc21f6bf59dec4af6
Reviewed-on: https://code.wireshark.org/review/9720
Petri-Dish: Gerald Combs <gerald@wireshark.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Inspired by (but not as fancy as) Packet Fence, an enhancement written
for Ethereal a loooong time ago by Martin Visser:
https://www.wireshark.org/lists/ethereal-dev/200011/msg00122.html
Several text editors call this a "minimap".
Color each scrollbar raster line to match the color of up to 7 packets.
Note in the comments why this number was chosen.
If we have any flagged frames (marked, ignored, time ref) indicate them
on either side of the scrolbar.
Handle HiDPI (retina) displays. This means that your window size depends
on the height of your scrollbar *and* your monitor resolution.
Qt's idea of the slider rect doesn't match up with the slider on OS X.
This might be local to my build -- I can replicate it Qt Creator.
Change-Id: Ia089d2d766ce37bab11e22d1a5721b4908935304
Reviewed-on: https://code.wireshark.org/review/8982
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Tested-by: Gerald Combs <gerald@wireshark.org>
Move frame flag (mark, ignore, ref_time) member functions from
PacketList to PacketListModel. They arguably belong there and we can
emit dataChanged to signal updates.
Rename some variables named "index" since they shadow a function name.
Change-Id: I9a731a76e4e63e562b561c29d13915278d5a7dbb
Reviewed-on: https://code.wireshark.org/review/9663
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Be less aggressive about resetting the packet list model so that we
retain our selection when (un)coloring packets or rebuilding columns.
Correctly enable a colorization menu item while we're here.
Change-Id: I8d1c8f26dbb8a814b8344b609695b77632006e4b
Reviewed-on: https://code.wireshark.org/review/9608
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Change-Id: Ib982662db6cf68730a7d121eac60d9bc5ae67429
Reviewed-on: https://code.wireshark.org/review/9195
Petri-Dish: Michael Mann <mmann78@netscape.net>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Reviewed-by: Michael Mann <mmann78@netscape.net>
The size hint changes in g5ab8490 broke packet list column resizing. Add
back a way to force the hinting necessary for sizeHintForColumn to work
properly.
Bug: 11067
Ping-Bug: 10924
Change-Id: Icd525200f7a88f2b6d22a3039cda98ad0527239e
Reviewed-on: https://code.wireshark.org/review/8334
Petri-Dish: Gerald Combs <gerald@wireshark.org>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Remove the setUniformRowHeights(true) call in the PacketList
constructor. This means that the packet list now calculates its height
by querying the Qt::SizeHintRole for every item instead of simply
multiplying item_height * number_of_rows.
Implement SizeHintRole in PacketListModel::data so that size
calculations aren't unbearably slow. We don't have any row text until an
item is drawn (via DisplayRole), so items with multiple lines end up
being rendered twice. Note where we make assumptions on line heights.
Although we call gtk_tree_view_set_fixed_height_mode() in the GTK+ UI we
don't have this problem there. GTK+ is apparently less strict about
enforcing row heights.
Bug: 10924
Change-Id: I98e9f4f5f321c2e03f18498e0a7e7556f88792a1
Reviewed-on: https://code.wireshark.org/review/7430
Petri-Dish: Evan Huus <eapache@gmail.com>
Reviewed-by: Evan Huus <eapache@gmail.com>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Michal Labedzki <michal.labedzki@tieto.com>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Add PacketList::applyRecentColumnWidths which set the packet list
column widths from our recent settings. Make sure it gets called at
startup and when we change profiles.
Save the packet list header state so that we can restore it when we
reset the model (i.e. freezing and thawing) and load a new capture file.
Save the state when the user resizes a column. As a side effect this
works around a weird bug that adjusts the width of column 1 at an
inopportune time.
Add a profileChanging signal so that we can save the main window geometry
in each profile.
Get rid of MainWindow::configurationProfileChanged. It was unused.
Apply saved pane widths and heights. Note that we might want to add a
separate pair of recent settings for the Qt panes.
Use the last opened directory in the capture file dialog.
Git rid of some unneeded Q_UNUSEDs while we're here.
Bug: 10953
Change-Id: I812aff59818cf0b4d1598b580627d32728d2e9d7
Reviewed-on: https://code.wireshark.org/review/7247
Reviewed-by: Michal Labedzki <michal.labedzki@tieto.com>
Tested-by: Michal Labedzki <michal.labedzki@tieto.com>
Petri-Dish: Michal Labedzki <michal.labedzki@tieto.com>
Tested-by: Petri Dish Buildbot <buildbot-no-reply@wireshark.org>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Reselect the current packet when we redraw the packet list.
Don't crash if we try to mark when no frame is selected. Try to
invalidate cached packet list strings when needed. Rename
PacketList::updateAll to redrawVisiblePackets so that its purpose is
more clear. When changing the font size, call redrawVisiblePackets
instead of rebuilding the entire list of visible rows.
Change-Id: I6e7a15067e7063d0efc26082170e1795ae3c0779
Reviewed-on: https://code.wireshark.org/review/4901
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Sorting behavior should be identical to GTK+.
Pass the correct position to beginInsertRows. Add a list of to-do items
to packet_list.cpp.
Change-Id: Ie6ab4b9f2d780a2af430d0f90529edca5485dada
Reviewed-on: https://code.wireshark.org/review/4481
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Plumb signals and slots for changing the text size in the main window.
Remove the bold font code from WiresharkApplication. It was only used in
ByteViewText. Adjust the icons a bit.
Bug: When we change the font preferences the packet list stops drawing
cached strings. I haven't been able to track down the cause.
Change-Id: I609d740c9f26265628fa4b7de1b75b0e56651387
Reviewed-on: https://code.wireshark.org/review/4269
Reviewed-by: Gerald Combs <gerald@wireshark.org>
For each displayed packet list row, save a copy of or a pointer to
column strings similar to ui/gtk/packet_list_store.c. This lets us call
epan_dissect_run only once per row.
Bug: 9511
Change-Id: I17e8ebeb5ed70518c9047413c3b2a46f01e904ef
Reviewed-on: https://code.wireshark.org/review/2752
Reviewed-by: Anders Broman <a.broman58@gmail.com>
Remove some dependencies on having an open capture file.
Change-Id: Iad5d03e0a45ed31c83781861ef1ed7fdeb92e4a5
Reviewed-on: https://code.wireshark.org/review/2352
Reviewed-by: Gerald Combs <gerald@wireshark.org>
Add a dissector table indexed by the file type, and, for the
file-type-specific records, have the frame dissector skip the usual
pseudo-header processing, as the pseudo-header has a file-type-specific
record subtype in it, and call the dissector for that file type's
records.
Change-Id: Ibe97cf6340ffb0dabc08f355891bc346391b91f9
Reviewed-on: https://code.wireshark.org/review/1782
Reviewed-by: Guy Harris <guy@alum.mit.edu>
Add a "record type" field to "struct wtap_pkthdr"; currently, it can be
REC_TYPE_PACKET, for a record containing a packet, or
REC_TYPE_FILE_TYPE_SPECIFIC, for records containing file-type-specific
data.
Modify code that reads packets to be able to handle non-packet records,
even if that just means ignoring them.
Rename some routines to indicate that they handle more than just
packets.
We don't yet have any libwiretap code that supplies records other than
REC_TYPE_PACKET or that supporting writing records other than
REC_TYPE_PACKET, or any code to support plugins for handling
REC_TYPE_FILE_TYPE_SPECIFIC records; this is just the first step for bug
8590.
Change-Id: Idb40b78f17c2c3aea72031bcd252abf9bc11c813
Reviewed-on: https://code.wireshark.org/review/1773
Reviewed-by: Guy Harris <guy@alum.mit.edu>
This reverts commit c0c480d08c.
A better way to do this is to have the record type be part of struct wtap_pkthdr; that keeps the metadata for the record together and requires fewer API changes. That is in-progress.
Change-Id: Ic558f163a48e2c6d0df7f55e81a35a5e24b53bc6
Reviewed-on: https://code.wireshark.org/review/1741
Reviewed-by: Guy Harris <guy@alum.mit.edu>
This is the first step towards implementing the mechanisms requestd in
bug 8590; currently, we don't return any records other than packet
records from libwiretap, and just ignore non-packet records in the rest
of Wireshark, but this at least gets the ball rolling.
Change-Id: I34a45b54dd361f69fdad1a758d8ca4f42d67d574
Reviewed-on: https://code.wireshark.org/review/1736
Reviewed-by: Guy Harris <guy@alum.mit.edu>
This fixes part-1 of bug9931: the uninitialized use of a wtap_pkthdr
struct. The second part of the bug deals with dissectors calling
the Ethernet dissector for ecnapsulated Ethernet packets but using
the wrong dissector handle to do so. That's unrelated to the issue this
commit addresses, so I'm splitting them up.
Change-Id: I87be7b736f82dd74d8c261062f88143372b5344c
Reviewed-on: https://code.wireshark.org/review/848
Reviewed-by: Hadriel Kaplan <hadrielk@yahoo.com>
Reviewed-by: Anders Broman <a.broman58@gmail.com>
(Using sed : sed -i '/^ \* \$Id\$/,+1 d')
Fix manually some typo (in export_object_dicom.c and crc16-plain.c)
Change-Id: I4c1ae68d1c4afeace8cb195b53c715cf9e1227a8
Reviewed-on: https://code.wireshark.org/review/497
Reviewed-by: Anders Broman <a.broman58@gmail.com>
not finding it, I finally found it in column_info.h
Renamed column_info.h to column-info.h to have consistency
with the column*h files.
svn path=/trunk/; revision=52667
column. Conversation spans (setup frame to last frame) are shown with a
square bracket. Linked frames are shown with a circle.
Use correct column justifications in Qt. Move common
justification-related packet list code to ui/packet_list_utils.[ch].
Add a last_frame element to conversation_t.
svn path=/trunk/; revision=50447
byte view.
Move the packet_char_enc enum from packet.h to frame_data.h. Make the
encoding flag a packet_char_enc and make it one bit.
Get rid of the "cfile" global in a few places. C++-ize some of the font
code. Clean up some variable names.
svn path=/trunk/; revision=45838
Pascal Quantin