subsurface/profile-widget/diveprofileitem.cpp

1019 lines
32 KiB
C++
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0
#include "profile-widget/diveprofileitem.h"
#include "qt-models/diveplotdatamodel.h"
#include "profile-widget/divecartesianaxis.h"
#include "profile-widget/divetextitem.h"
#include "profile-widget/animationfunctions.h"
#include "core/profile.h"
#ifndef SUBSURFACE_MOBILE
#include "desktop-widgets/preferences/preferencesdialog.h"
#endif
#include "qt-models/diveplannermodel.h"
#include "core/qthelper.h"
#include "core/settings/qPrefTechnicalDetails.h"
#include "libdivecomputer/parser.h"
#include "profile-widget/profilewidget2.h"
AbstractProfilePolygonItem::AbstractProfilePolygonItem() : QObject(), QGraphicsPolygonItem(), hAxis(NULL), vAxis(NULL), dataModel(NULL), hDataColumn(-1), vDataColumn(-1)
{
setCacheMode(DeviceCoordinateCache);
#ifndef SUBSURFACE_MOBILE
connect(PreferencesDialog::instance(), SIGNAL(settingsChanged()), this, SLOT(settingsChanged()));
#endif
}
void AbstractProfilePolygonItem::settingsChanged()
{
}
void AbstractProfilePolygonItem::setVisible(bool visible)
{
QGraphicsPolygonItem::setVisible(visible);
}
void AbstractProfilePolygonItem::setHorizontalAxis(DiveCartesianAxis *horizontal)
{
hAxis = horizontal;
connect(hAxis, SIGNAL(sizeChanged()), this, SLOT(modelDataChanged()));
modelDataChanged();
}
void AbstractProfilePolygonItem::setHorizontalDataColumn(int column)
{
hDataColumn = column;
modelDataChanged();
}
void AbstractProfilePolygonItem::setModel(DivePlotDataModel *model)
{
dataModel = model;
connect(dataModel, SIGNAL(dataChanged(QModelIndex, QModelIndex)), this, SLOT(modelDataChanged(QModelIndex, QModelIndex)));
connect(dataModel, SIGNAL(rowsAboutToBeRemoved(QModelIndex, int, int)), this, SLOT(modelDataRemoved(QModelIndex, int, int)));
modelDataChanged();
}
void AbstractProfilePolygonItem::modelDataRemoved(const QModelIndex&, int, int)
{
setPolygon(QPolygonF());
qDeleteAll(texts);
texts.clear();
}
void AbstractProfilePolygonItem::setVerticalAxis(DiveCartesianAxis *vertical)
{
vAxis = vertical;
connect(vAxis, SIGNAL(sizeChanged()), this, SLOT(modelDataChanged()));
connect(vAxis, SIGNAL(maxChanged()), this, SLOT(modelDataChanged()));
modelDataChanged();
}
void AbstractProfilePolygonItem::setVerticalDataColumn(int column)
{
vDataColumn = column;
modelDataChanged();
}
bool AbstractProfilePolygonItem::shouldCalculateStuff(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
if (!hAxis || !vAxis)
return false;
if (!dataModel || dataModel->rowCount() == 0)
return false;
if (hDataColumn == -1 || vDataColumn == -1)
return false;
if (topLeft.isValid() && bottomRight.isValid()) {
if ((topLeft.column() >= vDataColumn || topLeft.column() >= hDataColumn) &&
(bottomRight.column() <= vDataColumn || topLeft.column() <= hDataColumn)) {
return true;
}
}
return true;
}
void AbstractProfilePolygonItem::modelDataChanged(const QModelIndex&, const QModelIndex&)
{
// Calculate the polygon. This is the polygon that will be painted on screen
// on the ::paint method. Here we calculate the correct position of the points
// regarting our cartesian plane ( made by the hAxis and vAxis ), the QPolygonF
// is an array of QPointF's, so we basically get the point from the model, convert
// to our coordinates, store. no painting is done here.
QPolygonF poly;
for (int i = 0, modelDataCount = dataModel->rowCount(); i < modelDataCount; i++) {
qreal horizontalValue = dataModel->index(i, hDataColumn).data().toReal();
qreal verticalValue = dataModel->index(i, vDataColumn).data().toReal();
QPointF point(hAxis->posAtValue(horizontalValue), vAxis->posAtValue(verticalValue));
poly.append(point);
}
setPolygon(poly);
qDeleteAll(texts);
texts.clear();
}
DiveProfileItem::DiveProfileItem() : show_reported_ceiling(0), reported_ceiling_in_red(0)
{
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::dcceilingChanged, this, &DiveProfileItem::settingsToggled);
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::redceilingChanged, this, &DiveProfileItem::settingsToggled);
}
void DiveProfileItem::settingsToggled(bool)
{
settingsChanged();
}
void DiveProfileItem::paint(QPainter *painter, const QStyleOptionGraphicsItem *option, QWidget *widget)
{
if (polygon().isEmpty())
return;
painter->save();
// This paints the Polygon + Background. I'm setting the pen to QPen() so we don't get a black line here,
// after all we need to plot the correct velocities colors later.
setPen(Qt::NoPen);
QGraphicsPolygonItem::paint(painter, option, widget);
// Here we actually paint the boundaries of the Polygon using the colors that the model provides.
// Those are the speed colors of the dives.
QPen pen;
pen.setCosmetic(true);
pen.setWidth(2);
QPolygonF poly = polygon();
// This paints the colors of the velocities.
for (int i = 1, count = dataModel->rowCount(); i < count; i++) {
QModelIndex colorIndex = dataModel->index(i, DivePlotDataModel::COLOR);
pen.setBrush(QBrush(colorIndex.data(Qt::BackgroundRole).value<QColor>()));
painter->setPen(pen);
if (i < poly.count())
painter->drawLine(poly[i - 1], poly[i]);
}
painter->restore();
}
int DiveProfileItem::maxCeiling(int row)
{
int max = -1;
plot_data *entry = dataModel->data().entry + row;
for (int tissue = 0; tissue < 16; tissue++) {
if (max < entry->ceilings[tissue])
max = entry->ceilings[tissue];
}
return max;
}
void DiveProfileItem::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
bool eventAdded = false;
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
AbstractProfilePolygonItem::modelDataChanged(topLeft, bottomRight);
if (polygon().isEmpty())
return;
show_reported_ceiling = prefs.dcceiling;
reported_ceiling_in_red = prefs.redceiling;
profileColor = getColor(DEPTH_BOTTOM);
#ifdef SUBSURFACE_MOBILE
Q_UNUSED(eventAdded);
#else
int currState = qobject_cast<ProfileWidget2 *>(scene()->views().first())->currentState;
if (currState == ProfileWidget2::PLAN) {
plot_data *entry = dataModel->data().entry;
for (int i = 0; i < dataModel->rowCount(); i++, entry++) {
int max = maxCeiling(i);
// Don't scream if we violate the ceiling by a few cm
if (entry->depth < max - 100 && entry->sec > 0) {
profileColor = QColor(Qt::red);
if (!eventAdded) {
add_event(&displayed_dive.dc, entry->sec, SAMPLE_EVENT_CEILING, -1, max / 1000,
QT_TRANSLATE_NOOP("gettextFromC", "planned waypoint above ceiling"));
eventAdded = true;
}
}
}
}
#endif
/* Show any ceiling we may have encountered */
if (prefs.dcceiling && !prefs.redceiling) {
QPolygonF p = polygon();
plot_data *entry = dataModel->data().entry + dataModel->rowCount() - 1;
for (int i = dataModel->rowCount() - 1; i >= 0; i--, entry--) {
if (!entry->in_deco) {
/* not in deco implies this is a safety stop, no ceiling */
p.append(QPointF(hAxis->posAtValue(entry->sec), vAxis->posAtValue(0)));
} else {
p.append(QPointF(hAxis->posAtValue(entry->sec), vAxis->posAtValue(qMin(entry->stopdepth, entry->depth))));
}
}
setPolygon(p);
}
// This is the blueish gradient that the Depth Profile should have.
// It's a simple QLinearGradient with 2 stops, starting from top to bottom.
QLinearGradient pat(0, polygon().boundingRect().top(), 0, polygon().boundingRect().bottom());
pat.setColorAt(1, profileColor);
pat.setColorAt(0, getColor(DEPTH_TOP));
setBrush(QBrush(pat));
int last = -1;
for (int i = 0, count = dataModel->rowCount(); i < count; i++) {
struct plot_data *pd = dataModel->data().entry;
struct plot_data *entry = pd + i;
// "min/max" are the 9-minute window min/max indices
struct plot_data *min_entry = pd + entry->min;
struct plot_data *max_entry = pd + entry->max;
if (entry->depth < 2000)
continue;
if ((entry == max_entry) && entry->depth / 100 != last) {
plot_depth_sample(entry, Qt::AlignHCenter | Qt::AlignBottom, getColor(SAMPLE_DEEP));
last = entry->depth / 100;
}
if ((entry == min_entry) && entry->depth / 100 != last) {
plot_depth_sample(entry, Qt::AlignHCenter | Qt::AlignTop, getColor(SAMPLE_SHALLOW));
last = entry->depth / 100;
}
if (entry->depth != last)
last = -1;
}
}
void DiveProfileItem::settingsChanged()
{
//TODO: Only modelDataChanged() here if we need to rebuild the graph ( for instance,
// if the prefs.dcceiling are enabled, but prefs.redceiling is disabled
// and only if it changed something. let's not waste cpu cycles repoloting something we don't need to.
modelDataChanged();
}
void DiveProfileItem::plot_depth_sample(struct plot_data *entry, QFlags<Qt::AlignmentFlag> flags, const QColor &color)
{
DiveTextItem *item = new DiveTextItem(this);
item->setPos(hAxis->posAtValue(entry->sec), vAxis->posAtValue(entry->depth));
item->setText(get_depth_string(entry->depth, true));
item->setAlignment(flags);
item->setBrush(color);
texts.append(item);
}
DiveHeartrateItem::DiveHeartrateItem()
{
QPen pen;
pen.setBrush(QBrush(getColor(::HR_PLOT)));
pen.setCosmetic(true);
pen.setWidth(1);
setPen(pen);
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::hrgraphChanged, this, &DiveHeartrateItem::setVisible);
}
void DiveHeartrateItem::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
int last = -300, last_printed_hr = 0, sec = 0;
struct {
int sec;
int hr;
} hist[3] = {};
// We don't have enougth data to calculate things, quit.
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
qDeleteAll(texts);
texts.clear();
2017-02-20 08:29:20 +00:00
// Ignore empty values. a heart rate of 0 would be a bad sign.
QPolygonF poly;
for (int i = 0, modelDataCount = dataModel->rowCount(); i < modelDataCount; i++) {
int hr = dataModel->index(i, vDataColumn).data().toInt();
if (!hr)
continue;
sec = dataModel->index(i, hDataColumn).data().toInt();
QPointF point(hAxis->posAtValue(sec), vAxis->posAtValue(hr));
poly.append(point);
if (hr == hist[2].hr)
// same as last one, no point in looking at printing
continue;
hist[0] = hist[1];
hist[1] = hist[2];
hist[2].sec = sec;
hist[2].hr = hr;
// don't print a HR
// if it's not a local min / max
// if it's been less than 5min and less than a 20 beats change OR
// if it's been less than 2min OR if the change from the
// last print is less than 10 beats
// to test min / max requires three points, so we now look at the
// previous one
sec = hist[1].sec;
hr = hist[1].hr;
if ((hist[0].hr < hr && hr < hist[2].hr) ||
(hist[0].hr > hr && hr > hist[2].hr) ||
((sec < last + 300) && (abs(hr - last_printed_hr) < 20)) ||
(sec < last + 120) ||
(abs(hr - last_printed_hr) < 10))
continue;
last = sec;
createTextItem(sec, hr);
last_printed_hr = hr;
}
setPolygon(poly);
if (texts.count())
texts.last()->setAlignment(Qt::AlignLeft | Qt::AlignBottom);
}
void DiveHeartrateItem::createTextItem(int sec, int hr)
{
DiveTextItem *text = new DiveTextItem(this);
text->setAlignment(Qt::AlignRight | Qt::AlignBottom);
text->setBrush(getColor(HR_TEXT));
text->setPos(QPointF(hAxis->posAtValue(sec), vAxis->posAtValue(hr)));
text->setScale(0.7); // need to call this BEFORE setText()
text->setText(QString("%1").arg(hr));
texts.append(text);
}
void DiveHeartrateItem::paint(QPainter *painter, const QStyleOptionGraphicsItem*, QWidget*)
{
if (polygon().isEmpty())
return;
painter->save();
painter->setPen(pen());
painter->drawPolyline(polygon());
painter->restore();
}
DivePercentageItem::DivePercentageItem(int i)
{
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::percentagegraphChanged, this, &DivePercentageItem::setVisible);
tissueIndex = i;
settingsChanged();
}
void DivePercentageItem::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
int sec = 0;
// We don't have enougth data to calculate things, quit.
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
2017-02-20 08:29:20 +00:00
// Ignore empty values. a heart rate of 0 would be a bad sign.
QPolygonF poly;
for (int i = 0, modelDataCount = dataModel->rowCount(); i < modelDataCount; i++) {
sec = dataModel->index(i, hDataColumn).data().toInt();
QPointF point(hAxis->posAtValue(sec), vAxis->posAtValue(64 - 4 * tissueIndex));
poly.append(point);
}
setPolygon(poly);
if (texts.count())
texts.last()->setAlignment(Qt::AlignLeft | Qt::AlignBottom);
}
QColor DivePercentageItem::ColorScale(double value, int inert)
{
QColor color;
double scaledValue = value / (AMB_PERCENTAGE * inert) * 1000.0;
if (scaledValue < 0.8) // grade from cyan to blue to purple
color.setHsvF(0.5 + 0.25 * scaledValue / 0.8, 1.0, 1.0);
else if (scaledValue < 1.0) // grade from magenta to black
color.setHsvF(0.75, 1.0, (1.0 - scaledValue) / 0.2);
else if (value < AMB_PERCENTAGE) // grade from black to bright green
color.setHsvF(0.333, 1.0, (value - AMB_PERCENTAGE * inert / 1000.0) / (AMB_PERCENTAGE - AMB_PERCENTAGE * inert / 1000.0));
else if (value < 65) // grade from bright green (0% M) to yellow-green (30% M)
color.setHsvF(0.333 - 0.133 * (value - AMB_PERCENTAGE) / (65.0 - AMB_PERCENTAGE), 1.0, 1.0);
else if (value < 85) // grade from yellow-green (30% M) to orange (70% M)
color.setHsvF(0.2 - 0.1 * (value - 65.0) / 20.0, 1.0, 1.0);
else if (value < 100) // grade from orange (70% M) to red (100% M)
color.setHsvF(0.1 * (100.0 - value) / 15.0, 1.0, 1.0);
else if (value < 120) // M value exceeded - grade from red to white
color.setHsvF(0.0, 1 - (value - 100.0) / 20.0, 1.0);
else // white
color.setHsvF(0.0, 0.0, 1.0);
return color;
}
void DivePercentageItem::paint(QPainter *painter, const QStyleOptionGraphicsItem*, QWidget*)
{
if (polygon().isEmpty())
return;
painter->save();
QPen mypen;
mypen.setCapStyle(Qt::FlatCap);
mypen.setCosmetic(false);
QPolygonF poly = polygon();
for (int i = 1, modelDataCount = dataModel->rowCount(); i < modelDataCount; i++) {
if (i < poly.count()) {
double value = dataModel->index(i, vDataColumn).data().toDouble();
struct gasmix gasmix = gasmix_air;
const struct event *ev = NULL;
int sec = dataModel->index(i, DivePlotDataModel::TIME).data().toInt();
gasmix = get_gasmix(&displayed_dive, displayed_dc, sec, &ev, gasmix);
int inert = 1000 - get_o2(gasmix);
mypen.setBrush(QBrush(ColorScale(value, inert)));
painter->setPen(mypen);
painter->drawLine(poly[i - 1], poly[i]);
}
}
painter->restore();
}
DiveAmbPressureItem::DiveAmbPressureItem()
{
QPen pen;
pen.setBrush(QBrush(getColor(::AMB_PRESSURE_LINE)));
pen.setCosmetic(true);
pen.setWidth(2);
setPen(pen);
settingsChanged();
}
void DiveAmbPressureItem::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
int sec = 0;
// We don't have enougth data to calculate things, quit.
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
2017-02-20 08:29:20 +00:00
// Ignore empty values. a heart rate of 0 would be a bad sign.
QPolygonF poly;
for (int i = 0, modelDataCount = dataModel->rowCount(); i < modelDataCount; i++) {
int hr = dataModel->index(i, vDataColumn).data().toInt();
if (!hr)
continue;
sec = dataModel->index(i, hDataColumn).data().toInt();
QPointF point(hAxis->posAtValue(sec), vAxis->posAtValue(hr));
poly.append(point);
}
setPolygon(poly);
if (texts.count())
texts.last()->setAlignment(Qt::AlignLeft | Qt::AlignBottom);
}
void DiveAmbPressureItem::paint(QPainter *painter, const QStyleOptionGraphicsItem*, QWidget*)
{
if (polygon().isEmpty())
return;
painter->save();
painter->setPen(pen());
painter->drawPolyline(polygon());
painter->restore();
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::percentagegraphChanged, this, &DiveAmbPressureItem::setVisible);
}
DiveGFLineItem::DiveGFLineItem()
{
QPen pen;
pen.setBrush(QBrush(getColor(::GF_LINE)));
pen.setCosmetic(true);
pen.setWidth(2);
setPen(pen);
settingsChanged();
}
void DiveGFLineItem::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
int sec = 0;
// We don't have enougth data to calculate things, quit.
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
2017-02-20 08:29:20 +00:00
// Ignore empty values. a heart rate of 0 would be a bad sign.
QPolygonF poly;
for (int i = 0, modelDataCount = dataModel->rowCount(); i < modelDataCount; i++) {
int hr = dataModel->index(i, vDataColumn).data().toInt();
if (!hr)
continue;
sec = dataModel->index(i, hDataColumn).data().toInt();
QPointF point(hAxis->posAtValue(sec), vAxis->posAtValue(hr));
poly.append(point);
}
setPolygon(poly);
if (texts.count())
texts.last()->setAlignment(Qt::AlignLeft | Qt::AlignBottom);
}
void DiveGFLineItem::paint(QPainter *painter, const QStyleOptionGraphicsItem*, QWidget*)
{
if (polygon().isEmpty())
return;
painter->save();
painter->setPen(pen());
painter->drawPolyline(polygon());
painter->restore();
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::percentagegraphChanged, this, &DiveAmbPressureItem::setVisible);
}
DiveTemperatureItem::DiveTemperatureItem()
{
QPen pen;
pen.setBrush(QBrush(getColor(::TEMP_PLOT)));
pen.setCosmetic(true);
pen.setWidth(2);
setPen(pen);
}
void DiveTemperatureItem::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
int last = -300, last_printed_temp = 0, sec = 0, last_valid_temp = 0;
// We don't have enougth data to calculate things, quit.
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
qDeleteAll(texts);
texts.clear();
// Ignore empty values. things do not look good with '0' as temperature in kelvin...
QPolygonF poly;
for (int i = 0, modelDataCount = dataModel->rowCount(); i < modelDataCount; i++) {
int mkelvin = dataModel->index(i, vDataColumn).data().toInt();
if (!mkelvin)
continue;
last_valid_temp = mkelvin;
sec = dataModel->index(i, hDataColumn).data().toInt();
QPointF point(hAxis->posAtValue(sec), vAxis->posAtValue(mkelvin));
poly.append(point);
/* don't print a temperature
* if it's been less than 5min and less than a 2K change OR
* if it's been less than 2min OR if the change from the
* last print is less than .4K (and therefore less than 1F) */
if (((sec < last + 300) && (abs(mkelvin - last_printed_temp) < 2000)) ||
(sec < last + 120) ||
(abs(mkelvin - last_printed_temp) < 400))
continue;
last = sec;
if (mkelvin > 200000)
createTextItem(sec, mkelvin);
last_printed_temp = mkelvin;
}
setPolygon(poly);
/* it would be nice to print the end temperature, if it's
* different or if the last temperature print has been more
* than a quarter of the dive back */
if (last_valid_temp > 200000 &&
((abs(last_valid_temp - last_printed_temp) > 500) || ((double)last / (double)sec < 0.75))) {
createTextItem(sec, last_valid_temp);
}
if (texts.count())
texts.last()->setAlignment(Qt::AlignLeft | Qt::AlignBottom);
}
void DiveTemperatureItem::createTextItem(int sec, int mkelvin)
{
temperature_t temp;
temp.mkelvin = mkelvin;
DiveTextItem *text = new DiveTextItem(this);
text->setAlignment(Qt::AlignRight | Qt::AlignBottom);
text->setBrush(getColor(TEMP_TEXT));
text->setPos(QPointF(hAxis->posAtValue(sec), vAxis->posAtValue(mkelvin)));
text->setScale(0.8); // need to call this BEFORE setText()
text->setText(get_temperature_string(temp, true));
texts.append(text);
}
void DiveTemperatureItem::paint(QPainter *painter, const QStyleOptionGraphicsItem*, QWidget*)
{
if (polygon().isEmpty())
return;
painter->save();
painter->setPen(pen());
painter->drawPolyline(polygon());
painter->restore();
}
DiveMeanDepthItem::DiveMeanDepthItem()
{
QPen pen;
pen.setBrush(QBrush(getColor(::HR_AXIS)));
pen.setCosmetic(true);
pen.setWidth(2);
setPen(pen);
lastRunningSum = 0.0;
settingsChanged();
}
void DiveMeanDepthItem::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
double meandepthvalue = 0.0;
// We don't have enougth data to calculate things, quit.
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
QPolygonF poly;
plot_data *entry = dataModel->data().entry;
for (int i = 0, modelDataCount = dataModel->rowCount(); i < modelDataCount; i++, entry++) {
// Ignore empty values
if (entry->running_sum == 0 || entry->sec == 0)
continue;
meandepthvalue = entry->running_sum / entry->sec;
QPointF point(hAxis->posAtValue(entry->sec), vAxis->posAtValue(meandepthvalue));
poly.append(point);
}
lastRunningSum = meandepthvalue;
setPolygon(poly);
createTextItem();
}
void DiveMeanDepthItem::paint(QPainter *painter, const QStyleOptionGraphicsItem*, QWidget*)
{
if (polygon().isEmpty())
return;
painter->save();
painter->setPen(pen());
painter->drawPolyline(polygon());
painter->restore();
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::show_average_depthChanged, this, &DiveAmbPressureItem::setVisible);
}
void DiveMeanDepthItem::createTextItem() {
plot_data *entry = dataModel->data().entry;
int sec = entry[dataModel->rowCount()-1].sec;
qDeleteAll(texts);
texts.clear();
DiveTextItem *text = new DiveTextItem(this);
text->setAlignment(Qt::AlignRight | Qt::AlignTop);
text->setBrush(getColor(TEMP_TEXT));
text->setPos(QPointF(hAxis->posAtValue(sec) + 1, vAxis->posAtValue(lastRunningSum)));
text->setScale(0.8); // need to call this BEFORE setText()
text->setText(get_depth_string(lrint(lastRunningSum), true));
texts.append(text);
}
void DiveGasPressureItem::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
// We don't have enougth data to calculate things, quit.
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
int plotted_cyl[MAX_CYLINDERS] = { false, };
int last_plotted[MAX_CYLINDERS] = { 0, };
QPolygonF poly[MAX_CYLINDERS];
QPolygonF boundingPoly;
polygons.clear();
for (int i = 0, count = dataModel->rowCount(); i < count; i++) {
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
struct plot_data *entry = dataModel->data().entry + i;
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
for (int cyl = 0; cyl < MAX_CYLINDERS; cyl++) {
int mbar = GET_PRESSURE(entry, cyl);
int time = entry->sec;
if (!mbar)
continue;
QPointF point(hAxis->posAtValue(time), vAxis->posAtValue(mbar));
boundingPoly.push_back(point);
if (plotted_cyl[cyl]) {
/* Have we used this culinder in the last two minutes? Continue */
if (time - last_plotted[cyl] <= 2*60) {
poly[cyl].push_back(point);
last_plotted[cyl] = time;
continue;
}
/* Finish the previous one, start a new one */
polygons.append(poly[cyl]);
poly[cyl] = QPolygonF();
}
plotted_cyl[cyl] = true;
last_plotted[cyl] = time;
poly[cyl].push_back(point);
}
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
}
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
for (int cyl = 0; cyl < MAX_CYLINDERS; cyl++) {
if (!plotted_cyl[cyl])
continue;
polygons.append(poly[cyl]);
}
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
setPolygon(boundingPoly);
qDeleteAll(texts);
texts.clear();
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
int seen_cyl[MAX_CYLINDERS] = { false, };
int last_pressure[MAX_CYLINDERS] = { 0, };
int last_time[MAX_CYLINDERS] = { 0, };
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
// These are offset values used to print the gas lables and pressures on a
// dive profile at appropriate Y-coordinates. We alternate aligning the
// label and the gas pressure above and under the pressure line.
// The values are historical, and we could try to pick the over/under
// depending on whether this pressure is higher or lower than the average.
// Right now it's just strictly alternating when you have multiple gas
// pressures.
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
QFlags<Qt::AlignmentFlag> alignVar = Qt::AlignTop;
QFlags<Qt::AlignmentFlag> align[MAX_CYLINDERS];
double axisRange = (vAxis->maximum() - vAxis->minimum())/1000; // Convert axis pressure range to bar
double axisLog = log10(log10(axisRange));
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
for (int i = 0, count = dataModel->rowCount(); i < count; i++) {
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
struct plot_data *entry = dataModel->data().entry + i;
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
for (int cyl = 0; cyl < MAX_CYLINDERS; cyl++) {
int mbar = GET_PRESSURE(entry, cyl);
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
if (!mbar)
continue;
if (!seen_cyl[cyl]) {
double value_y_offset, label_y_offset;
// Magic Y offset depending on whether we're aliging
// the top of the text or the bottom of the text to
// the pressure line.
value_y_offset = -0.5;
if (alignVar & Qt::AlignTop) {
label_y_offset = 5 * axisLog;
} else {
label_y_offset = -7 * axisLog;
}
plotPressureValue(mbar, entry->sec, alignVar, value_y_offset);
plotGasValue(mbar, entry->sec, displayed_dive.cylinder[cyl].gasmix, alignVar, label_y_offset);
seen_cyl[cyl] = true;
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
/* Alternate alignment as we see cylinder use.. */
align[cyl] = alignVar;
alignVar ^= Qt::AlignTop | Qt::AlignBottom;
}
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
last_pressure[cyl] = mbar;
last_time[cyl] = entry->sec;
}
}
Profile support for multiple concurrent pressure sensors This finally handles multiple cylinder pressures, both overlapping and consecutive, and it seems to work on the nasty cases I've thrown at it. Want to just track five different cylinders all at once, without any pesky gas switch events? Sure, you can do that. It will show five different gas pressures for your five cylinders, and they will go down as you breathe down the cylinders. I obviously don't have any real data for that case, but I do have a test file with five actual cylinders that all have samples over the whole course of the dive. The end result looks messy as hell, but what did you expect? HOWEVER. The only way to do this sanely was - actually make the "struct plot_info" have all the cylinder pressures (so no "sensor index and pressure" - every cylinder has a pressure for every plot info entry) This obviously makes the plot_info much bigger. We used to have MAX_CYLINDERS be a fairly generous 8, which seems sane. The planning code made that 8 be 20. That seems questionable. But whatever. The good news is that the plot-info should hopefully get freed, and only be allocated one dive at a time, so the fact that it is big and nasty shouldn't be a scaling issue, though. - the "populate_pressure_information()" function had to be rewritten quite a bit. The good news is that it's actually simpler now, although I would not go so far as to really call it simple. It's still complicated and suble, but now it explicitly just does one cylinder at a time. It *used* to have this insanely complicated "keep track of the pressure ranges for every cylinder at once". I just couldn't stand that model and keep my sanity, so it now just tracks one cylinder at a time, and doesn't have an array of live data, instead the caller will just call it for each cylinder. - get rid of some of our hackier stuff, like the code that populates the plot_info data code with the currently selected cylinder number, and clears out any other pressures. That obviously does *not* work when you may not have a single primary cylinder any more. Now, the above sounds like all good things. Yeah, it mostly is. BUT. There's a few big downsides from the above: - there's no sane way to do this as a series of small changes. The change to make the plot_info take an array of cylinder pressures rather than the sensor+pressure model really isn't amenable to "fix up one use at a time". When you switch over to the new data structure model, you have to switch over to the new way of populating the pressure ranges. The two just go hand in hand. - Some of our code *depended* on the "sensor+pressure" model. I fixed all the ones I could sanely fix. There was one particular case that I just couldn't sanely fix, and I didn't care enough about it to do something insane. So the only _known_ breakage is the "TankItem" profile widget. That's the bar at the bottom of the profile that shows which cylinder is in use right now. You'd think that would be trivial to fix up, and yes it would be - I could just use the regular model of firstcyl = explicit_first_cylinder(dive, dc) .. then iterate over the gas change events to see the others .. but the problem with the "TankItem" widget is that it does its own model, and it has thrown away the dive and the dive computer information. It just doesn't even know. It only knows what cylinders there are, and the plot_info. And it just used to look at the sensor number in the plot_info, and be done with that. That number no longer exists. - I have tested it, and I think the code is better, but hey, it's a fairly large patch to some of the more complex code in our code base. That "interpolate missing pressure fields" code really isn't pretty. It may be prettier, but.. Anyway, without further ado, here's the patch. No sign-off yet, because I do think people should look and comment. But I think the patch is fine, and I'll fix anythign that anybody can find, *except* for that TankItem thing that I will refuse to touch. That class is ugly. It needs to have access to the actual dive. Note how it actually does remove more lines than it adds, and that's despite added comments etc. The code really is simpler, but there may be cases in there that need more work. Known missing pieces that don't currently take advantage of concurrent cylinder pressure data: - the momentary SAC rate coloring for dives will need more work - dive merging (but we expect to generally normally not merge dive computers, which is the main source of sensor data) - actually taking advantage of different sensor data from different dive computers But most of all: Testing. Lots and lots of testing to find all the corner cases. Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Dirk Hohndel <dirk@hohndel.org>
2017-07-27 17:17:05 +00:00
// For each cylinder, on right hand side of profile, write cylinder pressure
for (int cyl = 0; cyl < MAX_CYLINDERS; cyl++) {
if (last_time[cyl]) {
double value_y_offset = -0.5;
plotPressureValue(last_pressure[cyl], last_time[cyl], align[cyl] | Qt::AlignLeft, value_y_offset);
}
}
}
void DiveGasPressureItem::plotPressureValue(int mbar, int sec, QFlags<Qt::AlignmentFlag> align, double pressure_offset)
{
const char *unit;
int pressure = get_pressure_units(mbar, &unit);
DiveTextItem *text = new DiveTextItem(this);
text->setPos(hAxis->posAtValue(sec), vAxis->posAtValue(mbar) + pressure_offset );
text->setText(QString("%1%2").arg(pressure).arg(unit));
text->setAlignment(align);
text->setBrush(getColor(PRESSURE_TEXT));
texts.push_back(text);
}
void DiveGasPressureItem::plotGasValue(int mbar, int sec, struct gasmix gasmix, QFlags<Qt::AlignmentFlag> align, double gasname_offset)
{
QString gas = get_gas_string(gasmix);
DiveTextItem *text = new DiveTextItem(this);
text->setPos(hAxis->posAtValue(sec), vAxis->posAtValue(mbar) + gasname_offset );
text->setText(gas);
text->setAlignment(align);
text->setBrush(getColor(PRESSURE_TEXT));
texts.push_back(text);
}
void DiveGasPressureItem::paint(QPainter *painter, const QStyleOptionGraphicsItem*, QWidget*)
{
if (polygon().isEmpty())
return;
QPen pen;
pen.setCosmetic(true);
pen.setWidth(2);
painter->save();
struct plot_data *entry;
Q_FOREACH (const QPolygonF &poly, polygons) {
entry = dataModel->data().entry;
for (int i = 1, count = poly.count(); i < count; i++, entry++) {
if (!in_planner()) {
if (entry->sac)
pen.setBrush(getSacColor(entry->sac, displayed_dive.sac));
else
pen.setBrush(MED_GRAY_HIGH_TRANS);
} else {
pen.setBrush(getPressureColor(entry->density));
}
painter->setPen(pen);
painter->drawLine(poly[i - 1], poly[i]);
}
}
painter->restore();
}
DiveCalculatedCeiling::DiveCalculatedCeiling(ProfileWidget2 *widget) :
profileWidget(widget),
is3mIncrement(false)
{
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::calcceilingChanged, this, &DiveCalculatedCeiling::setVisible);
setVisible(prefs.calcceiling);
settingsChanged();
}
void DiveCalculatedCeiling::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
connect(profileWidget, SIGNAL(dateTimeChangedItems()), this, SLOT(recalc()), Qt::UniqueConnection);
// We don't have enougth data to calculate things, quit.
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
AbstractProfilePolygonItem::modelDataChanged(topLeft, bottomRight);
// Add 2 points to close the polygon.
QPolygonF poly = polygon();
if (poly.isEmpty())
return;
QPointF p1 = poly.first();
QPointF p2 = poly.last();
poly.prepend(QPointF(p1.x(), vAxis->posAtValue(0)));
poly.append(QPointF(p2.x(), vAxis->posAtValue(0)));
setPolygon(poly);
QLinearGradient pat(0, polygon().boundingRect().top(), 0, polygon().boundingRect().bottom());
pat.setColorAt(0, getColor(CALC_CEILING_SHALLOW));
pat.setColorAt(1, getColor(CALC_CEILING_DEEP));
setPen(QPen(QBrush(Qt::NoBrush), 0));
setBrush(pat);
}
void DiveCalculatedCeiling::paint(QPainter *painter, const QStyleOptionGraphicsItem *option, QWidget *widget)
{
if (polygon().isEmpty())
return;
QGraphicsPolygonItem::paint(painter, option, widget);
}
DiveCalculatedTissue::DiveCalculatedTissue(ProfileWidget2 *widget) : DiveCalculatedCeiling(widget)
{
settingsChanged();
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::calcalltissuesChanged, this, &DiveCalculatedTissue::setVisible);
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::calcceilingChanged, this, &DiveCalculatedTissue::setVisible);
}
void DiveCalculatedTissue::setVisible(bool)
{
settingsChanged();
}
void DiveCalculatedTissue::settingsChanged()
{
DiveCalculatedCeiling::setVisible(prefs.calcalltissues && prefs.calcceiling);
}
DiveReportedCeiling::DiveReportedCeiling()
{
connect(qPrefTechnicalDetails::instance(), &qPrefTechnicalDetails::dcceilingChanged, this, &DiveReportedCeiling::setVisible);
setVisible(prefs.dcceiling);
settingsChanged();
}
void DiveReportedCeiling::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
QPolygonF p;
p.append(QPointF(hAxis->posAtValue(0), vAxis->posAtValue(0)));
plot_data *entry = dataModel->data().entry;
for (int i = 0, count = dataModel->rowCount(); i < count; i++, entry++) {
if (entry->in_deco && entry->stopdepth) {
p.append(QPointF(hAxis->posAtValue(entry->sec), vAxis->posAtValue(qMin(entry->stopdepth, entry->depth))));
} else {
p.append(QPointF(hAxis->posAtValue(entry->sec), vAxis->posAtValue(0)));
}
}
setPolygon(p);
QLinearGradient pat(0, p.boundingRect().top(), 0, p.boundingRect().bottom());
// does the user want the ceiling in "surface color" or in red?
if (prefs.redceiling) {
pat.setColorAt(0, getColor(CEILING_SHALLOW));
pat.setColorAt(1, getColor(CEILING_DEEP));
} else {
pat.setColorAt(0, getColor(BACKGROUND_TRANS));
pat.setColorAt(1, getColor(BACKGROUND_TRANS));
}
setPen(QPen(QBrush(Qt::NoBrush), 0));
setBrush(pat);
}
void DiveCalculatedCeiling::recalc()
{
#ifndef SUBSURFACE_MOBILE
dataModel->calculateDecompression();
#endif
}
void DiveCalculatedCeiling::settingsChanged()
{
if (dataModel && is3mIncrement != prefs.calcceiling3m) {
// recalculate that part.
recalc();
}
is3mIncrement = prefs.calcceiling3m;
}
void DiveReportedCeiling::paint(QPainter *painter, const QStyleOptionGraphicsItem *option, QWidget *widget)
{
if (polygon().isEmpty())
return;
QGraphicsPolygonItem::paint(painter, option, widget);
}
void PartialPressureGasItem::modelDataChanged(const QModelIndex &topLeft, const QModelIndex &bottomRight)
{
//AbstractProfilePolygonItem::modelDataChanged();
if (!shouldCalculateStuff(topLeft, bottomRight))
return;
plot_data *entry = dataModel->data().entry;
QPolygonF poly;
QPolygonF alertpoly;
alertPolygons.clear();
double threshold_min = 100.0; // yes, a ridiculous high partial pressure
double threshold_max = 0.0;
if (thresholdPtrMax)
threshold_max = *thresholdPtrMax;
if (thresholdPtrMin)
threshold_min = *thresholdPtrMin;
bool inAlertFragment = false;
for (int i = 0; i < dataModel->rowCount(); i++, entry++) {
double value = dataModel->index(i, vDataColumn).data().toDouble();
int time = dataModel->index(i, hDataColumn).data().toInt();
QPointF point(hAxis->posAtValue(time), vAxis->posAtValue(value));
poly.push_back(point);
if (thresholdPtrMax && value >= threshold_max) {
if (inAlertFragment) {
alertPolygons.back().push_back(point);
} else {
alertpoly.clear();
alertpoly.push_back(point);
alertPolygons.append(alertpoly);
inAlertFragment = true;
}
} else if (thresholdPtrMin && value <= threshold_min) {
if (inAlertFragment) {
alertPolygons.back().push_back(point);
} else {
alertpoly.clear();
alertpoly.push_back(point);
alertPolygons.append(alertpoly);
inAlertFragment = true;
}
} else {
inAlertFragment = false;
}
}
setPolygon(poly);
/*
createPPLegend(trUtf8("pN" UTF8_SUBSCRIPT_2),getColor(PN2), legendPos);
*/
}
void PartialPressureGasItem::paint(QPainter *painter, const QStyleOptionGraphicsItem*, QWidget*)
{
const qreal pWidth = 0.0;
painter->save();
painter->setPen(QPen(normalColor, pWidth));
painter->drawPolyline(polygon());
QPolygonF poly;
painter->setPen(QPen(alertColor, pWidth));
Q_FOREACH (const QPolygonF &poly, alertPolygons)
painter->drawPolyline(poly);
painter->restore();
}
void PartialPressureGasItem::setThresholdSettingsKey(const double *prefPointerMin, const double *prefPointerMax)
{
thresholdPtrMin = prefPointerMin;
thresholdPtrMax = prefPointerMax;
}
PartialPressureGasItem::PartialPressureGasItem() :
thresholdPtrMin(NULL),
thresholdPtrMax(NULL)
{
}
void PartialPressureGasItem::setColors(const QColor &normal, const QColor &alert)
{
normalColor = normal;
alertColor = alert;
}