I'm developing a gxt application that provides a view for svn-like versioning: A TreePanel that displays a file system structure with files and folders having different states (new, modified, deleted, etc.).
I want to display a small overlay icon on each of the items to reflect its state. What I could do, is to create an icon for each state, but I don't want to do that, because I would end up in creating a large bundle of redundant images.
What would be the best way to implement overlay icon capabilities in trees or grids?
After checking the dom of the tree structure, I figured out that a tree icon is displayed using the background-url css attribute. To ensure the image is being displayed with correct size, the src attribute of the element is filled with a placeholder image url.
The trick is, to replace the placeholder image with the overlay icon, since it is displayed above the actual image.
To accomplish this, I extended ClippedImagePrototype to inject the url of the overlay image:
public class OverlayImagePrototype extends ClippedImagePrototype {
protected String overlayImageUrl = null;
public static AbstractImagePrototype create(ImageResource resource, String overlayUrl) {
return new OverlayImagePrototype(resource.getSafeUri(),
resource.getLeft(), resource.getTop(), resource.getWidth(),
resource.getHeight(), overlayUrl);
}
private OverlayImagePrototype(SafeUri url, int left, int top, int width,
int height, String overlayUrl) {
super(url, left, top, width, height);
this.overlayImageUrl = overlayUrl;
}
public ImagePrototypeElement createElement() {
ImagePrototypeElement imgEl = super.createElement();
if (overlayImageUrl != null) {
imgEl.setAttribute("src", overlayImageUrl);
}
return imgEl;
}
}
This is how I use OverlayImagePrototype in the IconProvider implementation:
tree.setIconProvider(new ModelIconProvider<ModelData>() {
public AbstractImagePrototype getIcon(ModelData model) {
return OverlayImagePrototype.create(model.get("icon"), model.get("overlayIconUrl"));
}
});
Related
I'm using Itext 7 and their html2Pdf lib.
Is there a way to implement for example cmyk colors.
.wootWorkingCMYK-color{
color: cmyk( 1 , 0.69 , 0.08 , 0.54);
}
I know the itext core part pretty good, looking for away to use the html2Pdf side. I'm aware of the CssApplierFactory but this seems to be to far up the chain.
Well, of course there is a way of processing custom CSS properties like cmyk colors, but unfortunately the code would be quite bulky and you will need to write quite some code for different cases. I will show how to apply custom color for font, but e.g. for backgrounds, borders or other cases you will need to write separate code in a similar way. Reason behind it is that iText layout structure, although designed with HTML/CSS in mind, is not 100% similar and has some differences you have to code around.
Having that said, if you can fork, build and use your custom version from sources, this is the way I would advice to go. Although it has drawbacks like having to rebase to get updates, the solution would be simpler and more generic. To do that, search for usages of CssUtils.parseRgbaColor in pdfHTML module, and you will find that it is used in BackgroundApplierUtil, BorderStyleApplierUtil, FontStyleApplierUtil, OutlineApplierUtil. There you will find code like
if (!CssConstants.TRANSPARENT.equals(cssColorPropValue)) {
float[] rgbaColor = CssUtils.parseRgbaColor(cssColorPropValue);
Color color = new DeviceRgb(rgbaColor[0], rgbaColor[1], rgbaColor[2]);
float opacity = rgbaColor[3];
transparentColor = new TransparentColor(color, opacity);
} else {
transparentColor = new TransparentColor(ColorConstants.BLACK, 0f);
}
Which I belive you can tweak to process cmyk as well, knowing that you know core part pretty well.
Now, the solution without custom pdfHTML version is to indeed start with implementing ICssApplierFactory, or subclassing default implementation DefaultCssApplierFactory. We are mostly interested in customizing implementation of SpanTagCssApplier and BlockCssApplier, but you can consult with DefaultTagCssApplierMapping to get the full list of appliers and cases they are used in, so that you can decide which of them you want to process in your code.
I will show you how to add support for custom color space for font color in the two main applier classes I mentioned and you can work from there.
private static class CustomCssApplierFactory implements ICssApplierFactory {
private static final ICssApplierFactory DEFAULT_FACTORY = new DefaultCssApplierFactory();
#Override
public ICssApplier getCssApplier(IElementNode tag) {
ICssApplier defaultApplier = DEFAULT_FACTORY.getCssApplier(tag);
if (defaultApplier instanceof SpanTagCssApplier) {
return new CustomSpanTagCssApplier();
} else if (defaultApplier instanceof BlockCssApplier) {
return new CustomBlockCssApplier();
} else {
return defaultApplier;
}
}
}
private static class CustomSpanTagCssApplier extends SpanTagCssApplier {
#Override
protected void applyChildElementStyles(IPropertyContainer element, Map<String, String> css, ProcessorContext context, IStylesContainer stylesContainer) {
super.applyChildElementStyles(element, css, context, stylesContainer);
String color = css.get("color2");
if (color != null) {
color = color.trim();
if (color.startsWith("cmyk")) {
element.setProperty(Property.FONT_COLOR, new TransparentColor(parseCmykColor(color)));
}
}
}
}
private static class CustomBlockCssApplier extends BlockCssApplier {
#Override
public void apply(ProcessorContext context, IStylesContainer stylesContainer, ITagWorker tagWorker) {
super.apply(context, stylesContainer, tagWorker);
IPropertyContainer container = tagWorker.getElementResult();
if (container != null) {
String color = stylesContainer.getStyles().get("color2");
if (color != null) {
color = color.trim();
if (color.startsWith("cmyk")) {
container.setProperty(Property.FONT_COLOR, new TransparentColor(parseCmykColor(color)));
}
}
}
}
}
// You might want a safer implementation with better handling of corner cases
private static DeviceCmyk parseCmykColor(String color) {
final String delim = "cmyk(), \t\r\n\f";
StringTokenizer tok = new StringTokenizer(color, delim);
float[] res = new float[]{0, 0, 0, 0};
for (int k = 0; k < 3; ++k) {
if (tok.hasMoreTokens()) {
res[k] = Float.parseFloat(tok.nextToken());
}
}
return new DeviceCmyk(res[0], res[1], res[2], res[3]);
}
Having that custom code, you should configure the ConverterProperties accordingly and pass it to HtmlConverter:
ConverterProperties properties = new ConverterProperties();
properties.setCssApplierFactory(new CustomCssApplierFactory());
HtmlConverter.convertToPdf(..., properties);
You might have noticed that I used color2 instead of color, and this is for a reason. pdfHTML has a mechanism of CSS property validation (as browsers do as well), to discard invalid CSS properties when calculating effective properties for an element. Unfortunately, there is no mechanism of customizing this validation logic currently and of course it treats cmyk colors as invalid declarations at the moment. Thus, if you really want to have custom color property, you will have to preprocess your HTML and replace declarations like color: cmyk... to color2: cmyk.. or whatever the property name you might want to use.
As I mentioned at the start of the answer, my recommendation is to build your own custom version :)
In a GWT web app, I am using a DataGrid to manage elements from a database. I represent a list of elements as rows, the columns being editable fields of their characteristics (id, name, description). I am mostly using the EditTextCell class.
I now want to create a custom cell, for a column that has to represent a list of "tags" that can be attached to every element. From this cell, tags could be added, using a + button (that makes a drop-down menu appear or something), and deleted. Each tag should be a kind of button, or interactive widget (I later want to display pop-up with info, trigger actions, etc).
Actually, it would not be so different from the "tags" bar on the Stack Overflow website...
So I have been looking for a solution:
I thought this would be easy to do. I imagined just putting a FlowPanel in the cell, adding/removing Buttons/Widgets dynamically. But it turns out that in GWT Widgets and Cells and very different objects apparently..
I read making use of the AbstractCell class to create a custom cell allows to do anything, but its working is very low level and obscure to me.
I saw that CompositeCell allows to combine various cell widgets into one cell, but I have not found if it is possible to do it dynamically, or if the widgets are going to be the same for all lines throughout a column. I mostly saw examples about, for instance, how to put two Buttons in every cell of a single column.
What is the easiest way to implement what I need?
EDIT:
So, after some tests, I am going for Andrei's suggestion and going "low-level", creating a custom cell extending AbstractCell<>. I could create an appropriate "render" function, that generates a list of html "button", and also attaches Javascript calls to my Java functions when triggering a Javascript event (onclick, onmouseover, onmouseout...).
It is working pretty well. For instance, by clicking the "+" button at the end a tag list, it calls a MenuBar widget that presents the list of tags that can be added.
But I am struggling to find a way to update the underlying data when adding a tag.
To sum up:
I have a CustomData class that represents the data I want to display in each line of the table. It also contains the list of tags as a Set.
ModelTable (extends DataGrid) is my table.
CustomCell (extends AbstractCell) can renders the list of tags as several buttons on a line.
A click on a "+" button in a cell makes a AddTagMenu popup drop down, from which I can click on the tag to add.
How do I update the content of the cell?
I tried playing around with onBrowserEvent, onEnterKeyDown, bus events... with no success. At best I can indeed add a tag element to the underlying object, but the table is not updated.
It's not possible to meet your requirements without going really "low-level", as you call it.
It's relatively easy to create a cell that would render tags exactly as you want them. Plus icon is also easy, if this is the only action on the cell. However, it is very difficult to make every tag within a cell an interactive widget, because the DataGrid will not let you attach handlers to HTML rendered within a cell. You will need to supply your own IDs to these widgets, and then attach handlers to them in your code. The problem, however, is that when the DataGrid refreshes/re-renders, your handlers will most likely be lost. So you will have to attach them again to every tag in every cell on every change in the DataGrid.
A much simpler approach is to create a composite widget that represents a "row", and then add these "rows" to a FlowPanel. You can easily make it look like a table with CSS, and supply your own widget that looks like a table header. You will need to recreate some of the functionality of the DataGrid, e.g. sorting when clicked on "column" header - if you need this functionality, of course.
As you have already noted, using CompositeCell could be a way to get what you want.
The idea is to create a cell for every tag and then (during rendering) decide which one should be shown (rendered). Finally combine all those cells into one by creating a CompositeCell.
The main disadvantage of this solution is that you need to know all possible tags before you create a DataGrid.
So, if you have a fixed list of possible tags or can get a list of all existing tags and this list is reasonably small, here is a solution.
First, we need to know which tag is represented by a column so I extended a Column class to keep information about a tag. Please, note that TagColumn uses ButtonCell and also handles update when the button is clicked:
public class TagColumn extends Column<DataType, String> {
private TagEnum tag;
public TagColumn(TagEnum tag) {
super(new ButtonCell());
this.tag = tag;
setFieldUpdater(new FieldUpdater<DataType, String>() {
#Override
public void update(int index, DataType object, String value) {
Window.alert("Tag " + getTag().getName() + " clicked");
}
});
}
public TagEnum getTag() {
return tag;
}
#Override
public String getValue(DataType object) {
return tag.getName();
}
}
Then create a cell for each tag (I have hard-coded all tags in a TagEnum):
List<HasCell<DataType, ?>> tagColumns = new ArrayList<HasCell<DataType, ?>>();
for(TagEnum tag : TagEnum.values())
tagColumns.add(new TagColumn(tag));
Now, the most important part: decide either to show the tag or not - overwrite render method of the CompositeCell:
CompositeCell<DataType> tagsCell = new CompositeCell<DataType>(tagColumns) {
#Override
protected <X> void render(Context context, DataType value, SafeHtmlBuilder sb, HasCell<DataType, X> hasCell) {
if(value.getTagList().contains(((TagColumn) hasCell).getTag()))
super.render(context, value, sb, hasCell);
else
sb.appendHtmlConstant("<span></span>");
}
};
This is important to always render any element (for example empty span when the tag should not be shown). Otherwise the CompositeCell's implemantation will get confused when accessing sibling elements.
Finally, full, working example code:
private DataGrid<DataType> getGrid() {
DataGrid<DataType> grid = new DataGrid<DataType>();
List<HasCell<DataType, ?>> tagColumns = new ArrayList<HasCell<DataType, ?>>();
for(TagEnum tag : TagEnum.values())
tagColumns.add(new TagColumn(tag));
CompositeCell<DataType> tagsCell = new CompositeCell<DataType>(tagColumns) {
#Override
protected <X> void render(Context context, DataType value, SafeHtmlBuilder sb, HasCell<DataType, X> hasCell) {
if(value.getTagList().contains(((TagColumn) hasCell).getTag()))
super.render(context, value, sb, hasCell);
else
sb.appendHtmlConstant("<span></span>");
}
};
Column<DataType, DataType> tagsColumn = new Column<DataType, DataType>(tagsCell) {
#Override
public DataType getValue(DataType object) {
return object;
}
};
grid.addColumn(tagsColumn, "Tags");
grid.setRowData(Arrays.asList(
new DataType(Arrays.asList(TagEnum.gwt)),
new DataType(Arrays.asList(TagEnum.table, TagEnum.datagrid)),
new DataType(Arrays.asList(TagEnum.datagrid, TagEnum.widget, TagEnum.customCell)),
new DataType(Arrays.asList(TagEnum.gwt, TagEnum.table, TagEnum.widget, TagEnum.customCell)),
new DataType(Arrays.asList(TagEnum.gwt, TagEnum.customCell)),
new DataType(Arrays.asList(TagEnum.gwt, TagEnum.table, TagEnum.datagrid, TagEnum.widget, TagEnum.customCell))
)
);
return grid;
}
public class TagColumn extends Column<DataType, String> {
private TagEnum tag;
public TagColumn(TagEnum tag) {
super(new ButtonCell());
this.tag = tag;
setFieldUpdater(new FieldUpdater<DataType, String>() {
#Override
public void update(int index, DataType object, String value) {
Window.alert("Tag " + getTag().getName() + " clicked");
}
});
}
public TagEnum getTag() {
return tag;
}
#Override
public String getValue(DataType object) {
return tag.getName();
}
}
public class DataType {
List<TagEnum> tagList;
public DataType(List<TagEnum> tagList) {
this.tagList = tagList;
}
public List<TagEnum> getTagList() {
return tagList;
}
}
public enum TagEnum {
gwt ("gwt"),
table ("table"),
datagrid ("datagrid"),
widget ("widget"),
customCell ("custom-cell");
private String name;
private TagEnum(String name) {
this.name = name;
}
public String getName() {
return name;
}
}
I want to fetch an image from a remote server (not my app's server, but I don't think that matters) and display it in an existing Image widget. The existing widget displays its original content, from a ClientBundle ImageResource , OK.
In UiBinder template:
<g:Image ui:field='myImage' resource='{res.anImage}'/>
In code:
#UiField Image myImage;
...
int width = Window.getClientWidth();
int height = Window.getClientHeight();
String url = ...;
myImage.addErrorHandler(new ErrorHandler() {
public void onError(ErrorEvent event) {
Window.alert("Error getting image data: " + event);
}
});
myImage.addLoadHandler(new LoadHandler() {
public void onLoad(LoadEvent event) {
Window.alert("LoadEvent: " + event);
}
});
myImage.setUrlAndVisibleRect(url, 0, 0, width, height);
As far as I can tell setUrlAndVisibleRect is a no-op. FireBug reports no network activity -- no request to the server specified by the URL. What am I overlooking? In my extended thrashing about trying to get this working I have inferred that it may have something to do with myImage not being "logically attached", but I'm not entirely sure what that means and I've no idea how to correct it if that is the problem.
EDIT with SUMMARY of SOLUTION:
My initial hunch was right. Because I had chosen to implement the image code within a second pseudo-widget (...extends Composite) that shared my UiBinder template with the main pseudo-widget that implements most of my app's UI, I got into trouble. I neglected to add this second pseudo-widget to the RootPanel as is normally done in the class that implements EntryPoint. This left my Image widget unattached to the widget chain, because its parent, the second pseudo-widget, was unattached. And an Image widget must be attached to work. What I ended up doing is moving the Image code back into my main app/GUI class, i.e., into the first and now only pseudo-widget and abandoning the special class for the Image code. I did that because it's simpler and the Image code turns out not to be as long as I had originally thought.
Adding image to the DOM is little tricky,the below code which supports all the browsers(setVisibility trick added to support IE also,as It has a different way to image rendering).
I did'nt use setUrlAndVisibleRect before and AFAIK,Image must render to the DOM inorder to resize it.Just try the below codes.
image.addLoadHandler(new LoadHandler() {
#Override
public void onLoad(LoadEvent event) {
//Do your operations on image .//resize ..etc
image.getElement().getStyle().setVisibility
(Style.Visibility.Visible);
}
});
image.getElement().getStyle().setVisibility(Style.Visibility.HIDDEN);
RootPanel.get().add(image);
image.setUrl(url);
You are using ClientBundle ImageResource which is compile time. You cannot change it unless you replace the new image with exact name in the exact position of prev one. One of the possible hack which is possible is, place the image in a div with its ID set ( getElement().setId("your ID") ). Once you get you new image you use RootPanel.get("Your Id") and do your job.
The problem is on the drawing by click or resizing browser. I have TabPanel placed with RowData, two TabItems with Chart (Google Vizualization) on one and Table with the same Data on the next. I create them on the page loading.
Then I click on Load Data (button) from DB, I redraw this two:
public void reDraw(final List<Double> slices, final String[] devices)
{
pcPie.draw(createTable(slices,devices),createOptions("По автомобилям"));
tPie.draw(createTable(slices, devices),CreateTableOptions());
}
That's work only for active TabItem and replace the drawing space from behind with this size (400px;200px) in generated HTML and I find that Data isn't changed at the behind section.
Also, when I resized the browser, Charts and Tables aren't resizing. I've tryed to use some of Layout, they don't work. May be I don't understand exactly how can use them correctly.
So,
How can I resize my Charts and Tables correct in the both of the
section (active and behind)?
How can I resize my Charts and Tables
on the browser resizing events?
Our first problem came from this: when you use the TabPanel component with some TabItems, behind TabItems aren't being created exactly, and you can not redraw them, cause object isn't created. So we change our code in activated section:
public void run() {
tpLineCharts.setBorders(true);
TabItem tiGraph = new TabItem("График");
tableData = createTable();
lcLines = new LineChart(tableData,
createOptions("По компании"));
lcLines.addSelectHandler(createSelectHandler(lcLines));
tiGraph.setLayout(new FitLayout());
tiGraph.add(lcLines);
tpLineCharts.add(tiGraph);
TabItem tiTable = new TabItem("Таблица");
tLine = new Table(tableData, CreateTableOptions());
tiTable.add(tLine);
tiTable.addListener(Events.Select, new Listener<BaseEvent>()
{
#Override
public void handleEvent(BaseEvent be) {
tLine.draw(tableData);
}
});
tpLineCharts.add(tiTable);
}}, CoreChart.PACKAGE, Table.PACKAGE);
where tableData - AbstractTableData. After this modification we can redraw our components:
public void reDrawLineChart(final ArrayList <Double> sumCompanyTraffic,
final ArrayList<Integer> axisName, String title)
{
tableData =createTable(sumCompanyTraffic, axisName);
tLine.draw(tableData, CreateTableOptions());
lcLines.draw(tableData, createOptions(title));
}
Also you need to add this options:
private Options createOptions(String title)
{
Options options = Options.create();
options.setTitleX("Период");
options.setTitle(title);
if(tpLineCharts.isRendered())
options.setSize(tpLineCharts.getWidth(true),
tpLineCharts.getHeight(true));
return options;
}
I am trying to drag and drop a cell in GWT and customize the drag image using dataTransfer.setDragImage().
My image is an ImageResource and here is what I have:
Here is how I construct my "draggable cell"
public CheckBoxDragCell(boolean dependsOnSelection, boolean handlesSelection,ImageResource icon) {
this(dependsOnSelection,handlesSelection);
this.icon = AbstractImagePrototype.create(icon);
}
Here is the eventHandler for the drag:
#Override
public void onBrowserEvent(Context context, Element parent, Boolean value,
NativeEvent event, ValueUpdater<Boolean> valueUpdater) {
String type = event.getType();
boolean enterPressed = "keydown".equals(type) && event.getKeyCode() == KeyCodes.KEY_ENTER;
if ("dragstart".equals(type)){
DataTransfer dataTransfer = event.getDataTransfer();
dataTransfer.setData("text", String.valueOf(context.getIndex()));
if (icon != null){
Image image = icon.createImage();
image.setVisible(true);
dataTransfer.setDragImage(ImageElement.as(image.getElement()), 25, 15);
}
But somehow the image never shows up. To debug, I created a dialogbox and displayed the image there when the eventHandler is called and sure enough it show up there... but not when I call the dataTransfer.setDragImage(....) function.
Any help will be enormously appreciated.
The reason the image did not show up in the first case is because the ImageElement is not added to the DOM.
Try adding the ImageElement.as(image.getElement()) to the DOM, then it would show.
In this case it could be an issue of browser support?
I've tried this HTML5 native drag and drop demo: http://html5doctor.com/native-drag-and-drop/
The custom drag image part of it didn't work for me (Chrome 17.0.932.0 dev-m).