Method Details

• createGraph

  void createGraph() throws IOException

  Create graph

  Throws:

  IOException - if the source RRD cannot be read while fetching data or the rendered image cannot be written out

• collectInfo

  private void collectInfo()

  Collect info

• saveImage

  private void saveImage() throws IOException

  Save image

  Throws:

  IOException - if the image file cannot be written

• drawOverlay

  private void drawOverlay() throws IOException

  Draw overlay

  Throws:

  IOException - if the overlay image file cannot be read

• gator

  private void gator()

  Draws the vertical graph signature along the left edge.

• countPlottedSeries

  static int countPlottedSeries(List<PlotElement> plotElements)

  Counts the independently plotted series in a graph, which decides whether the lines are drawn solid or dotted.

  Counts Line and Area together. An area counts because it is drawn with an outline, so a graph made of a filled total plus a subset line has two lines on it and the subset must be dotted to match. Stack stays excluded: a stack is one quantity split into bands, not several series.

  Parameters:

  plotElements - the graph's plot elements, may be null

  Returns:

  number of plotted series, zero for an empty or null list

• drawRulesAndSpans

  private void drawRulesAndSpans()

  Draw rules and spans

• drawText

  private void drawText()

  Draw text

• drawGrid

  private void drawGrid()

  Draw grid

• valueShadePaint

  private Paint valueShadePaint(Color[] stops)

  Build the vertical gradient that shades a series line by its value.

  The gradient runs from the bottom of the plot area to the top, which is what makes it read by value rather than by position: a point's height above the baseline is its magnitude, so the paint at that height is the colour for that magnitude. Three or more stops need LinearGradientPaint, the only one of the two gradient types that holds an arbitrary number; either way the stops are spread evenly over the span.

  Parameters:

  stops - bottom-to-top colour stops, or null for a flat line

  Returns:

  the gradient paint, or null when no shading was asked for

• drawData

  private void drawData()

  Draw data

• drawAxis

  private void drawAxis()

  Draw axis

• drawBackground

  private void drawBackground() throws IOException

  Draw background

  Throws:

  IOException - if a background or canvas image file cannot be read

• createImageWorker

  private void createImageWorker()

  Create image worker

• placeLegends

  private void placeLegends()

  Place legends

• initializeLimits

  private void initializeLimits()

  Initialize limits

• removeOutOfRangeRules

  private void removeOutOfRangeRules()

  Remove out of range rules

• removeOutOfRangeSpans

  private void removeOutOfRangeSpans()

  Remove out of range spans

• expandValueRange

  private void expandValueRange()

  Expand value range

• identifySiUnit

  private void identifySiUnit()

  Identify si unit

• findMinMaxValues

  private void findMinMaxValues()

  Find min max values

• calculatePlotValues

  private void calculatePlotValues()

  Calculate plot values

• resolveTextElements

  private void resolveTextElements()

  Resolve text elements

• fetchData

  private void fetchData() throws IOException

  Fetch data

  Throws:

  IOException - if a source's RRD file cannot be read

• lazyCheck

  private boolean lazyCheck() throws IOException

  Lazy check

  Returns:

  true if the existing image is fresh enough that the graph need not be redrawn

  Throws:

  IOException - if the image file's last-modified time cannot be read

• drawLegend

  private void drawLegend()

  Draw legend

• getFontHeight

  double getFontHeight(RrdGraphConstants.FontTag fonttag)

  Return the font height

  Parameters:

  fonttag - the text element whose configured font is measured

  Returns:

  the height in pixels of a line of that element's font

• getFontCharWidth

  double getFontCharWidth(RrdGraphConstants.FontTag fonttag)

  Return the font char width

  Parameters:

  fonttag - the text element whose configured font is measured

  Returns:

  the width in pixels of one average character of that font

• getInterlegendSpace

  double getInterlegendSpace()

  Return the inter-legend space

  Returns:

  the blank gap in pixels left between inter-legend rows

• getLeading

  double getLeading()

  The vertical advance applied between legend rows.

  Returns:

  the line advance in pixels applied between legend rows

• getSmallLeading

  double getSmallLeading()

  Line spacing for small legend text.

  Returns:

  the legend font height times the small leading factor

• getBoxSpace

  double getBoxSpace()

  Space around the legend box.

  Returns:

  the legend font height times the box space factor, rounded up

• getBox

  private double getBox()

  Chooses a round grid spacing for the axis.

  Returns:

  the spacing between grid lines

• xtr

  private double[] xtr(long[] timestamps)

  Converts timestamps to pixel coordinates, duplicating each step.

  Parameters:

  timestamps - the sample timestamps to map, at least one

  Returns:

  the pixel x of each timestamp, each repeated so the polyline steps

• xtrDistinct

  private double[] xtrDistinct(long[] timestamps)

  Converts timestamps to pixel x with one point per sample.

  xtr(long[]) deliberately writes each sample's x twice so that the polyline is forced into a staircase: a horizontal run to each sample, then a vertical jump to the next. That is what gives the plots their stepped look, but it also means every value change lands on a zero-width edge, leaving no horizontal room for a curve to cross. The smoothed renderer needs the undoubled positions so a transition spans the pixels between samples.

  Parameters:

  timestamps - sample timestamps, one per data point and not pre-doubled

  Returns:

  one pixel x per timestamp

• ytr

  private double[] ytr(double[] values)

  Converts values to pixel coordinates, duplicating each step.

  Parameters:

  values - the sample values to map, NaN marks a gap, at least one

  Returns:

  the pixel y of each value, each repeated so the polyline steps

• ytrDistinct

  private double[] ytrDistinct(double[] values)

  Converts values to pixel y with one point per sample.

  The undoubled counterpart of ytr(double[]); see xtrDistinct(long[]) for why the doubled form exists and why the smoothed renderer must not use it. Gaps stay NaN so PathIterator still splits the series where the data does.

  Parameters:

  values - sample values, possibly containing NaN

  Returns:

  one pixel y per value