| 1 | <?xml version="1.0" encoding="UTF-8"?> | 
|---|
| 2 | <?xml-stylesheet type="text/css" | 
|---|
| 3 | href="eclipseos2-xxe.css" | 
|---|
| 4 | ?> | 
|---|
| 5 | <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" | 
|---|
| 6 | "xhtml1-strict.dtd"> | 
|---|
| 7 | <html> | 
|---|
| 8 | <head> | 
|---|
| 9 | <link href="eclipseos2.css" rel="stylesheet" type="text/css" /> | 
|---|
| 10 |  | 
|---|
| 11 | <title>Eclipse for OS/2 Transitional Project Notes</title> | 
|---|
| 12 | </head> | 
|---|
| 13 |  | 
|---|
| 14 | <body> | 
|---|
| 15 | <h1>SWT Step 3. Basics of GUI, part 2</h1> | 
|---|
| 16 |  | 
|---|
| 17 | <h2>Objective</h2> | 
|---|
| 18 |  | 
|---|
| 19 | <p>Implement basic visualization facilities such as primitive drawing. In | 
|---|
| 20 | particular, this means the implementation of | 
|---|
| 21 | <code>o.e.swt.graphics.GC</code> [#Base] class. The test example of this | 
|---|
| 22 | step should be able to create an SWT top window and draw some graphic | 
|---|
| 23 | primitives in it.</p> | 
|---|
| 24 |  | 
|---|
| 25 | <h2>Task notes</h2> | 
|---|
| 26 |  | 
|---|
| 27 | <h3><a name="FrameClientWindow">Frame client window (obsolete)</a></h3> | 
|---|
| 28 |  | 
|---|
| 29 | <p>In OS/2, as opposed to Windows (at least from the API point of view), | 
|---|
| 30 | the client part of the frame window (the area where the custom drawing | 
|---|
| 31 | takes place) is the separate window with its own window class and | 
|---|
| 32 | procedure. SWT uses the Windows approach and therefore the frame window | 
|---|
| 33 | (i.e. <code>Decorations</code> and <code>Shell</code> widgets) should | 
|---|
| 34 | encapsulate the client window functionality to be as a whole for the user. | 
|---|
| 35 | To achieve this the client window, an instance of the special | 
|---|
| 36 | <code>ClientArea</code> class (which is derived from the | 
|---|
| 37 | <code>Composite</code>), is anonymously created by the | 
|---|
| 38 | <code>Decorations.createHandle()</code> method. "Anonymously" means that | 
|---|
| 39 | this instance is unaccessible anywhere outside the | 
|---|
| 40 | <code>Decorations</code> class. The special feature of the | 
|---|
| 41 | <code>ClientArea</code> class is that it uses its parent's (i.e. the | 
|---|
| 42 | <code>Decorations</code>) event table and, when sending or posting | 
|---|
| 43 | messages, sets the <code>Event.widget</code> field to the value of | 
|---|
| 44 | <code>parent</code> rather than <code>this</code>. This means that all | 
|---|
| 45 | events (specifically, the <code>SWT.Paint</code> event) generated by the | 
|---|
| 46 | client window look as <code>Decorations</code> events to the user (and | 
|---|
| 47 | directed to listeners registered on the <code>Decorations</code> | 
|---|
| 48 | instance).</p> | 
|---|
| 49 |  | 
|---|
| 50 | <h3>Window presentation space</h3> | 
|---|
| 51 |  | 
|---|
| 52 | <p>The presentation space handle to do low level painting in the window is | 
|---|
| 53 | obtained via the <code>Control.internal_new_GC()</code> method and must be | 
|---|
| 54 | freed by the <code>Control.internal_dispose_GC()</code> method when no | 
|---|
| 55 | more needed. The logic of work of these two methods is as follows:</p> | 
|---|
| 56 |  | 
|---|
| 57 | <ul> | 
|---|
| 58 | <li>if <code>data</code> argument is <code>null</code> or | 
|---|
| 59 | <code>data.rcl</code> is <code>null</code> then <code>WinGetPS()</code> | 
|---|
| 60 | is used to obtain the presentation space handle and | 
|---|
| 61 | <code>WinReleasePS()</code> -- to release it. Otherwise | 
|---|
| 62 | <code>WinBeginPaint()</code> obtains a handle, reusing | 
|---|
| 63 | <code>data.hps</code> (if not <code>null)</code> and filling | 
|---|
| 64 | <code>data.rcl</code> with the updating rectangle, and | 
|---|
| 65 | <code>WinEndPaint()</code> releases the presentation space | 
|---|
| 66 | obtained.</li> | 
|---|
| 67 |  | 
|---|
| 68 | <li><code>WinGetPS()</code> and <code>WinBeginPaint()</code> use | 
|---|
| 69 | <code>data.hwnd</code>, if it is not zero, as the window handle to | 
|---|
| 70 | obtain the presentation space for. Otherwise <code>Control.handle</code> | 
|---|
| 71 | is used. The latter is the normal behavior (i.e. <code>data.hwnd</code> | 
|---|
| 72 | should normally be zero), the former can be used to substitute the | 
|---|
| 73 | handle in subclasses when obtaining the presentation space handle (it is | 
|---|
| 74 | not currently used and may be removed in the future).</li> | 
|---|
| 75 |  | 
|---|
| 76 | <li>Also, If <code>data</code> is not <code>null</code>, | 
|---|
| 77 | <code>data.device</code> is assigned to the return value of | 
|---|
| 78 | <code>getDevice()</code> and <code>data.hdc</code> is filled with the | 
|---|
| 79 | device context handle returned by <code>GpiQueryDevice()</code> for the | 
|---|
| 80 | presentation space obtained above.</li> | 
|---|
| 81 | </ul> | 
|---|
| 82 |  | 
|---|
| 83 | <h4>WM_PAINT message</h4> | 
|---|
| 84 |  | 
|---|
| 85 | <p>In the Windows version, the <code>WM_PAINT</code> handler uses the | 
|---|
| 86 | device context obtained via the <code>BeginPaint()</code> API call, which | 
|---|
| 87 | initializes it with default attributes. To setup the correct graphics | 
|---|
| 88 | context (a pen, brush, etc.) the <code>GC.win32_new()</code> method is | 
|---|
| 89 | called every time when the <code>WM_PAINT</code> message occurs. Although | 
|---|
| 90 | the same technique can be used in OS/2 too (and the existing code is ready | 
|---|
| 91 | to do it) at the present time the "long-term" presentation space is used | 
|---|
| 92 | to save attributes between <code>WM_PAINT</code> messages (and for some | 
|---|
| 93 | other purposes). This presentation space is allocated only once (lazily in | 
|---|
| 94 | the <code>WM_PAINT</code> handler) for every <code>Control</code> widget | 
|---|
| 95 | which has at least one <code>SWT.Paint</code> event listener registered, | 
|---|
| 96 | and it is released when the widget is being destroyed. However, the | 
|---|
| 97 | <code>Composite</code> class (and its descendants, of course) allocates | 
|---|
| 98 | such presentation space even if no <code>SWT.Paint</code> listeners are | 
|---|
| 99 | registered, to draw the background (this behavior may change in the | 
|---|
| 100 | future). This approach should lead to a better performance since no calls | 
|---|
| 101 | to setup attributes on every paint message (except the first one) are made | 
|---|
| 102 | (the <code>GCData.doInit</code> field is used by <code>GC.pm_new()</code> | 
|---|
| 103 | and the <code>WM_PAINT</code> handler to indicate the need to setup the | 
|---|
| 104 | presentation space). At the same time, this should not cause the increase | 
|---|
| 105 | of memory usage because the presentation space is obtained via the | 
|---|
| 106 | <code>WinGetPS()</code> call, which returns the cached presentation | 
|---|
| 107 | space.</p> | 
|---|
| 108 |  | 
|---|
| 109 | <p>Besides other things, the long-term presentation space is used to store | 
|---|
| 110 | the default view transformation matrix to do the automatic coordinate | 
|---|
| 111 | space flipping when using <code>Gpi*</code> functions for painting (see | 
|---|
| 112 | also <a href="swt002.html#ScreenCoordinates">here</a>). It allows to | 
|---|
| 113 | update the matrix only when the window is resized instead of doing it on | 
|---|
| 114 | every paint message.</p> | 
|---|
| 115 |  | 
|---|
| 116 | <h3><a name="GraphicPrimitives">Graphic primitives (obsolete)</a></h3> | 
|---|
| 117 |  | 
|---|
| 118 | <p>There is some muddle with dimensions of graphic primitives in the Win32 | 
|---|
| 119 | version of SWT: for example, the javadoc says that | 
|---|
| 120 | <code>GC.drawRoundRectangle()</code>draws a rectangle, those left and | 
|---|
| 121 | right edges are at <kbd>x</kbd> and <kbd>x + width</kbd>, top and bottom | 
|---|
| 122 | edges are at <kbd>y</kbd> and <kbd>y + height</kbd>. But it's not the case | 
|---|
| 123 | -- actually, the right edge is at <kbd>x + width - 1</kbd>, and the bottom | 
|---|
| 124 | one is at <kbd>y + height - 1</kbd>, i.e. the horizontal and vertical size | 
|---|
| 125 | is one pixel smaller than it stated. Besides this, all <code>fill*</code> | 
|---|
| 126 | versions of primitives draw figures one pixel smaller in size than their | 
|---|
| 127 | <code>draw*</code> versions, i.e. the rectangle drawn by | 
|---|
| 128 | <code>GC.fillRectangle(0,0,10,10)</code> is not the same in size as by | 
|---|
| 129 | <code>GC.drawRectangle(0,0,10,10)</code> -- the size of the former is 11 | 
|---|
| 130 | by 11 pixels while the size of the latter is 10 by 10. The docs just say | 
|---|
| 131 | that <code>fill*</code> versions fill the interior of a primitive, but we | 
|---|
| 132 | see no reasons for the interior to be one pixel smaller than the outline | 
|---|
| 133 | (although it's an expected behavior of the corresponding Win32 API calls, | 
|---|
| 134 | but not OS/2 ones). Also, there are errors when drawing very small figures | 
|---|
| 135 | (few pixels in size) and figures with the negative width and/or height | 
|---|
| 136 | (see the <code>SWT003_01</code> example). Most of these things look just | 
|---|
| 137 | like bugs, so in OS/2 version we decided to fix them until any | 
|---|
| 138 | incompatibility issue appears. "To fix" here means that the bounding | 
|---|
| 139 | rectangle of all primitives in the OS/2 version of SWT always follows the | 
|---|
| 140 | rule stated in the javadoc for <code>GC.drawRectangle()</code> about the | 
|---|
| 141 | edges and covers an area <kbd>width + 1</kbd> pixels wide by <kbd>height + | 
|---|
| 142 | 1</kbd> pixels tall.</p> | 
|---|
| 143 |  | 
|---|
| 144 | <h3>Colors</h3> | 
|---|
| 145 |  | 
|---|
| 146 | <p>In the Win32 version of SWT, if the device is in the palette mode (that | 
|---|
| 147 | is, when it can display only 256 distinct colors simultaneously) and we | 
|---|
| 148 | try to create new colors inside the paint event listener (as we draw in a | 
|---|
| 149 | window) these colors will not appear on the screen until the next window | 
|---|
| 150 | invalidation (see the <code>SWT003_02</code> example). It is because new | 
|---|
| 151 | colors are being added to the physical palette when it's already selected | 
|---|
| 152 | to the device context at the beginning of the paint event handling and | 
|---|
| 153 | won't be selected again until the next paint event. In the OS/2 version we | 
|---|
| 154 | solve this problem by posting the <code>WM_REALIZEPALETTE</code> message | 
|---|
| 155 | to every <code>Shell</code> widget when the new color is created, which | 
|---|
| 156 | causes widgets to invalidate themselves.</p> | 
|---|
| 157 |  | 
|---|
| 158 | <h3>Styled lines</h3> | 
|---|
| 159 |  | 
|---|
| 160 | <p>In Windows, when drawing styled lines (dotted, dashed etc.) intervals | 
|---|
| 161 | between sytle segments are considered to be the "background" of the line | 
|---|
| 162 | and therefore are drawn by the current brush (see the SWT003_01 example). | 
|---|
| 163 | In OS/2 they are considered to be transparent and this behavior cannot be | 
|---|
| 164 | altered. Due to the inefficiency of the only possible application-level | 
|---|
| 165 | implementation of this functionality it is left as is at the present.</p> | 
|---|
| 166 |  | 
|---|
| 167 | <h2>Step checklist</h2> | 
|---|
| 168 |  | 
|---|
| 169 | <table> | 
|---|
| 170 | <col width="40%" /> | 
|---|
| 171 |  | 
|---|
| 172 | <col /> | 
|---|
| 173 |  | 
|---|
| 174 | <col width="50%" /> | 
|---|
| 175 |  | 
|---|
| 176 | <thead> | 
|---|
| 177 | <tr> | 
|---|
| 178 | <th>Operation</th> | 
|---|
| 179 |  | 
|---|
| 180 | <th>Status</th> | 
|---|
| 181 |  | 
|---|
| 182 | <th>Remarks</th> | 
|---|
| 183 | </tr> | 
|---|
| 184 | </thead> | 
|---|
| 185 |  | 
|---|
| 186 | <tr> | 
|---|
| 187 | <td>Implement <code>OS.WinWindowFromID()</code> and <code>FID_*</code> | 
|---|
| 188 | constants</td> | 
|---|
| 189 |  | 
|---|
| 190 | <td>Done [dmik]</td> | 
|---|
| 191 |  | 
|---|
| 192 | <td></td> | 
|---|
| 193 | </tr> | 
|---|
| 194 |  | 
|---|
| 195 | <tr> | 
|---|
| 196 | <td>Implement the frame client window creation and the message routing | 
|---|
| 197 | for it</td> | 
|---|
| 198 |  | 
|---|
| 199 | <td>Done [dmik]</td> | 
|---|
| 200 |  | 
|---|
| 201 | <td></td> | 
|---|
| 202 | </tr> | 
|---|
| 203 |  | 
|---|
| 204 | <tr> | 
|---|
| 205 | <td>Add <code>OS.GpiCreateRegion()</code>, | 
|---|
| 206 | <code>WinQueryUpdateRegion()</code>, | 
|---|
| 207 | <code>WinInvalidateRegion()</code> and <code>RGN_* | 
|---|
| 208 | </code>constants</td> | 
|---|
| 209 |  | 
|---|
| 210 | <td>Done [dmik]</td> | 
|---|
| 211 |  | 
|---|
| 212 | <td>The <code>RECTL</code> structure (which is originally passed as a | 
|---|
| 213 | pointer) is passed as an <code>int</code> array (with four elements) | 
|---|
| 214 | from Java to simplify access to it from the native | 
|---|
| 215 | implementation.</td> | 
|---|
| 216 | </tr> | 
|---|
| 217 |  | 
|---|
| 218 | <tr> | 
|---|
| 219 | <td>Add <code>OS.GpiQueryRegionRects()</code>, <code>RGNRECT</code> | 
|---|
| 220 | class, <code>RECTDIR_*</code> constants</td> | 
|---|
| 221 |  | 
|---|
| 222 | <td>Done [dmik]</td> | 
|---|
| 223 |  | 
|---|
| 224 | <td></td> | 
|---|
| 225 | </tr> | 
|---|
| 226 |  | 
|---|
| 227 | <tr> | 
|---|
| 228 | <td>Implement <code>WM_PAINT</code> message handling and | 
|---|
| 229 | <code>SWT.Paint</code> event firing</td> | 
|---|
| 230 |  | 
|---|
| 231 | <td>Done [dmik]</td> | 
|---|
| 232 |  | 
|---|
| 233 | <td></td> | 
|---|
| 234 | </tr> | 
|---|
| 235 |  | 
|---|
| 236 | <tr> | 
|---|
| 237 | <td>Implement <code>OS.WinGetMaxPosition()</code></td> | 
|---|
| 238 |  | 
|---|
| 239 | <td>Done [dmik]</td> | 
|---|
| 240 |  | 
|---|
| 241 | <td>Used to implement the <code>Display.getClientArea()</code> method, | 
|---|
| 242 | which returns the size and position of maximized windows of the | 
|---|
| 243 | desktop. These values can differ from | 
|---|
| 244 | <code>Display.getBounds()</code>, e.g. when XCenter is running. Note: | 
|---|
| 245 | XCenter 1.0.1 has a strange feature -- it reduces the maximum height | 
|---|
| 246 | (reported by <code>WinGetMaxPosition()</code>) by the value of its own | 
|---|
| 247 | height plus additional 5 pixels, therefore the reported maximum height | 
|---|
| 248 | is 5 pixels less than the actual. Possibly, it's a bug.</td> | 
|---|
| 249 | </tr> | 
|---|
| 250 |  | 
|---|
| 251 | <tr> | 
|---|
| 252 | <td>Implement <code>Device.getBounds()</code>, | 
|---|
| 253 | <code>getClientArea()</code> for <code>Device</code>, | 
|---|
| 254 | <code>Display</code> and controls that support this</td> | 
|---|
| 255 |  | 
|---|
| 256 | <td>Done [dmik]</td> | 
|---|
| 257 |  | 
|---|
| 258 | <td><code>Scrollable.getClientArea()</code> code is temporary (should | 
|---|
| 259 | take the scrollers into account)</td> | 
|---|
| 260 | </tr> | 
|---|
| 261 |  | 
|---|
| 262 | <tr> | 
|---|
| 263 | <td>Add <code>OS.GpiMove</code>, <code>OS.GpiBox</code>, | 
|---|
| 264 | <code>GPI_*</code> and <code>DRO_*</code> constants, implement | 
|---|
| 265 | <code>GC.{fill|draw}[Round]Rectangle()</code> methods</td> | 
|---|
| 266 |  | 
|---|
| 267 | <td>Done [dmik]</td> | 
|---|
| 268 |  | 
|---|
| 269 | <td><code>POINTL</code> structures are passed as arrays of | 
|---|
| 270 | <code>int</code> with two elements (x and y).</td> | 
|---|
| 271 | </tr> | 
|---|
| 272 |  | 
|---|
| 273 | <tr> | 
|---|
| 274 | <td>Add <code>WM_CALCVALIDRECTS</code> message handling, | 
|---|
| 275 | <code>OS.CVR_*</code> constants</td> | 
|---|
| 276 |  | 
|---|
| 277 | <td>Done [dmik]</td> | 
|---|
| 278 |  | 
|---|
| 279 | <td><code>Composite</code> instance returns <code>CVR_REDRAW</code> on | 
|---|
| 280 | <code>WM_CALCVALIDRECTS</code> if it has the <code>CANVAS</code> state | 
|---|
| 281 | and no <code>SWT.NO_REDRAW_RESIZE</code> style to completely | 
|---|
| 282 | invalidate itself when resizing.</td> | 
|---|
| 283 | </tr> | 
|---|
| 284 |  | 
|---|
| 285 | <tr> | 
|---|
| 286 | <td>Create the draft <code>SWT003</code> testcase</td> | 
|---|
| 287 |  | 
|---|
| 288 | <td>Done [dmik]</td> | 
|---|
| 289 |  | 
|---|
| 290 | <td></td> | 
|---|
| 291 | </tr> | 
|---|
| 292 |  | 
|---|
| 293 | <tr> | 
|---|
| 294 | <td>Fix the bug of <code>GpiBox()</code></td> | 
|---|
| 295 |  | 
|---|
| 296 | <td>Done [dmik]</td> | 
|---|
| 297 |  | 
|---|
| 298 | <td><code>GpiBox()</code> draws the rectangle 1 pixel higher than it | 
|---|
| 299 | should do, when rounded corners and <code>DRO_FILL</code> or | 
|---|
| 300 | <code>DRO_OUTLINEFILL</code> are specified.</td> | 
|---|
| 301 | </tr> | 
|---|
| 302 |  | 
|---|
| 303 | <tr> | 
|---|
| 304 | <td>Add the long-term presentation space creation for a | 
|---|
| 305 | <code>Control</code> object</td> | 
|---|
| 306 |  | 
|---|
| 307 | <td>Done [dmik]</td> | 
|---|
| 308 |  | 
|---|
| 309 | <td></td> | 
|---|
| 310 | </tr> | 
|---|
| 311 |  | 
|---|
| 312 | <tr> | 
|---|
| 313 | <td>Implement <code>OS.GpiSetDefaultViewMatrix()</code>, | 
|---|
| 314 | <code>TRANSFORM_*</code> constants</td> | 
|---|
| 315 |  | 
|---|
| 316 | <td>Done [dmik]</td> | 
|---|
| 317 |  | 
|---|
| 318 | <td>Necessary for automatic coordinate space flipping.</td> | 
|---|
| 319 | </tr> | 
|---|
| 320 |  | 
|---|
| 321 | <tr> | 
|---|
| 322 | <td>Add <code>OS.SHORT{1|2}FROMMP</code> pseudo macros</td> | 
|---|
| 323 |  | 
|---|
| 324 | <td>Done [dmik]</td> | 
|---|
| 325 |  | 
|---|
| 326 | <td></td> | 
|---|
| 327 | </tr> | 
|---|
| 328 |  | 
|---|
| 329 | <tr> | 
|---|
| 330 | <td>Add the <code>o.e.swt.widgets.ClientArea</code> class</td> | 
|---|
| 331 |  | 
|---|
| 332 | <td>Done [dmik]</td> | 
|---|
| 333 |  | 
|---|
| 334 | <td>This class represents the frame client window.</td> | 
|---|
| 335 | </tr> | 
|---|
| 336 |  | 
|---|
| 337 | <tr> | 
|---|
| 338 | <td>Implement <code>OS.DevQueryCaps()</code> and <code>CAPS_*</code> | 
|---|
| 339 | constants</td> | 
|---|
| 340 |  | 
|---|
| 341 | <td>Done [dmik]</td> | 
|---|
| 342 |  | 
|---|
| 343 | <td>Currently, constants are almost fully commented -- uncomment when | 
|---|
| 344 | needed.</td> | 
|---|
| 345 | </tr> | 
|---|
| 346 |  | 
|---|
| 347 | <tr> | 
|---|
| 348 | <td>Overload <code>OS.WinSendMsg()</code> with one, that takes | 
|---|
| 349 | <code>OS.RECTL</code> as the first message parameter</td> | 
|---|
| 350 |  | 
|---|
| 351 | <td>Done [dmik]</td> | 
|---|
| 352 |  | 
|---|
| 353 | <td>Currently used to send <code>WM_CALCFRAMERECT</code> message.</td> | 
|---|
| 354 | </tr> | 
|---|
| 355 |  | 
|---|
| 356 | <tr> | 
|---|
| 357 | <td>Add <code>OS.GpiCreatePalette()</code>, | 
|---|
| 358 | <code>GpiDeletePalette()</code>, <code>LCOL_*</code> and | 
|---|
| 359 | <code>LCOLF_*</code> constants, | 
|---|
| 360 | <code>GpiQueryNearestPaletteIndex()</code>, | 
|---|
| 361 | <code>GpiQueryPaletteInfo()</code>, | 
|---|
| 362 | <code>GpiSetPaletteEntries()</code>, <code>GpiSelectPalette()</code>, | 
|---|
| 363 | <code>GpiCreateLogColorTable()</code>, | 
|---|
| 364 | <code>WinRealizePalette()</code></td> | 
|---|
| 365 |  | 
|---|
| 366 | <td>Done [dmik]</td> | 
|---|
| 367 |  | 
|---|
| 368 | <td>Note: <code>GpiQueryNearestPaletteIndex()</code> is undocumented. | 
|---|
| 369 | We hope it presents in Warp4 and earlier OS/2 versions...</td> | 
|---|
| 370 | </tr> | 
|---|
| 371 |  | 
|---|
| 372 | <tr> | 
|---|
| 373 | <td>Add <code>OS.GpiQueryColor()</code>, <code>GpiSetColor()</code>, | 
|---|
| 374 | <code>GpiQueryBackColor()</code>, <code>GpiSetBackColor()</code>, | 
|---|
| 375 | <code>GpiQieryMix()</code>, <code>GpiSetMix()</code>, | 
|---|
| 376 | <code>FM_*</code> constants, <code>GpiQueryBackMix()</code>, | 
|---|
| 377 | <code>GpiSetBackMix()</code>, <code>BM_*</code> constants, | 
|---|
| 378 | <code>GpiQueryPattern()</code>, <code>GpiSetPattern()</code>, | 
|---|
| 379 | <code>PATSYM_*</code> constants</td> | 
|---|
| 380 |  | 
|---|
| 381 | <td>Done [dmik]</td> | 
|---|
| 382 |  | 
|---|
| 383 | <td></td> | 
|---|
| 384 | </tr> | 
|---|
| 385 |  | 
|---|
| 386 | <tr> | 
|---|
| 387 | <td>Add <code>WM_REALIZEPALETTE</code> message handling, implement | 
|---|
| 388 | <code>{get|set}{Foreground|Background}()</code> methods of | 
|---|
| 389 | <code>Control</code> and <code>GC</code> classes, | 
|---|
| 390 | <code>OS.WinQuerySysColor()</code> and <code>SYSCLR_*</code> | 
|---|
| 391 | constants</td> | 
|---|
| 392 |  | 
|---|
| 393 | <td>Done [dmik]</td> | 
|---|
| 394 |  | 
|---|
| 395 | <td></td> | 
|---|
| 396 | </tr> | 
|---|
| 397 |  | 
|---|
| 398 | <tr> | 
|---|
| 399 | <td>Add the <code>GC.Color</code> class and implement color | 
|---|
| 400 | handling</td> | 
|---|
| 401 |  | 
|---|
| 402 | <td>Done [dmik]</td> | 
|---|
| 403 |  | 
|---|
| 404 | <td></td> | 
|---|
| 405 | </tr> | 
|---|
| 406 |  | 
|---|
| 407 | <tr> | 
|---|
| 408 | <td>Add <code>OS.GpiLine()</code>, <code>GpiBeginArea()</code>, | 
|---|
| 409 | <code>GpiEndArea()</code>, <code>GpiPartialArc()</code>, | 
|---|
| 410 | <code>GpiSetArcParams()</code>, | 
|---|
| 411 | <code>GpiQueryCurrentPosition()</code>, implement | 
|---|
| 412 | <code>GC.drawLine()</code>, <code>{draw|fill}Arc()</code>, | 
|---|
| 413 | <code>{draw|fill}Oval()</code> methods</td> | 
|---|
| 414 |  | 
|---|
| 415 | <td>Done [dmik]</td> | 
|---|
| 416 |  | 
|---|
| 417 | <td><code>ARCPARAMS</code> structure is passed to | 
|---|
| 418 | <code>GpiSetArcParams()</code> as an array of | 
|---|
| 419 | <code>int[4]</code>.</td> | 
|---|
| 420 | </tr> | 
|---|
| 421 |  | 
|---|
| 422 | <tr> | 
|---|
| 423 | <td>Create <code>TestCase</code> and <code>SWTTestCase</code> classes | 
|---|
| 424 | to simplify test case writing, remove the <code>SWT003</code> tescase | 
|---|
| 425 | and create <code>SWT003_01</code> and <code>SWT003_02</code> instead, | 
|---|
| 426 | update the <code>SWT004</code> testcase</td> | 
|---|
| 427 |  | 
|---|
| 428 | <td>Done [dmik]</td> | 
|---|
| 429 |  | 
|---|
| 430 | <td>If the step requires several different testcases each with its own | 
|---|
| 431 | top shell window, it's possible to do so, naming them as | 
|---|
| 432 | <code>SWTXXX_YY</code> where XXX is the step number and YY is the | 
|---|
| 433 | number of the testcase. Such testcases can be run as usual, specifying | 
|---|
| 434 | the step name in the format above via the <kbd>test.step</kbd> | 
|---|
| 435 | property in the <kbd>build.prp</kbd> file or on the command line.</td> | 
|---|
| 436 | </tr> | 
|---|
| 437 |  | 
|---|
| 438 | <tr> | 
|---|
| 439 | <td>Add <code>OS.GpiPolyLine()</code>, implement | 
|---|
| 440 | <code>GC.getBackground()</code>, <code>getForeground()</code>, | 
|---|
| 441 | <code>drawPolyline()</code>, <code>drawPolygon()</code>, | 
|---|
| 442 | <code>fillPolygon()</code></td> | 
|---|
| 443 |  | 
|---|
| 444 | <td>Done [dmik]</td> | 
|---|
| 445 |  | 
|---|
| 446 | <td></td> | 
|---|
| 447 | </tr> | 
|---|
| 448 |  | 
|---|
| 449 | <tr> | 
|---|
| 450 | <td>Add <code>OS.Gpi{Query|Set}LineType()</code>, | 
|---|
| 451 | <code>LINETYPE_*</code> constants, implement | 
|---|
| 452 | <code>GC.{get|set}LineStyle()</code>, <code>drawFocus()</code></td> | 
|---|
| 453 |  | 
|---|
| 454 | <td>Done [dmik]</td> | 
|---|
| 455 |  | 
|---|
| 456 | <td></td> | 
|---|
| 457 | </tr> | 
|---|
| 458 |  | 
|---|
| 459 | <tr> | 
|---|
| 460 | <td>Add <code>OS.Gpi{Begin|End|Stoke}Path()</code>, | 
|---|
| 461 | <code>Gpi{Query|Set}LineWidthGeom()</code>, | 
|---|
| 462 | <code>GpiSetLine{End|Join}</code>, implement | 
|---|
| 463 | <code>GC.{get|set}LineWidth</code>, | 
|---|
| 464 | <code>{get|set}XORMode()</code></td> | 
|---|
| 465 |  | 
|---|
| 466 | <td>Done [dmik]</td> | 
|---|
| 467 |  | 
|---|
| 468 | <td></td> | 
|---|
| 469 | </tr> | 
|---|
| 470 |  | 
|---|
| 471 | <tr> | 
|---|
| 472 | <td>Readress the <code>Decorations.redraw()</code> methods to the | 
|---|
| 473 | client window; add <code>OS.WinEnableWindowUpdate()</code>, implement | 
|---|
| 474 | <code>Control.setRedraw()</code>; add | 
|---|
| 475 | <code>OS.WinIsWindowShowing()</code>, implement | 
|---|
| 476 | <code>Control.isShowing()</code></td> | 
|---|
| 477 |  | 
|---|
| 478 | <td>Done [dmik]</td> | 
|---|
| 479 |  | 
|---|
| 480 | <td></td> | 
|---|
| 481 | </tr> | 
|---|
| 482 |  | 
|---|
| 483 | <tr> | 
|---|
| 484 | <td>Add <code>OS.GpiQueryClipBox()</code>, | 
|---|
| 485 | <code>GpiSetClipRegion()</code>, <code>GpiQueryRegionBox()</code>, | 
|---|
| 486 | <code>GpiCombineRegion()</code>, <code>GpiPtInRegion()</code> and | 
|---|
| 487 | <code>PRGN_*</code> constants, <code>GpiRectInRegion()</code> abd | 
|---|
| 488 | <code>RRGN_*</code> constants; implement the o.e.swt.graphics.Region | 
|---|
| 489 | class, implement <code>GC.{get|set}Clipping()</code>; add testcase | 
|---|
| 490 | <code>SWT003_03</code> to test clipping</td> | 
|---|
| 491 |  | 
|---|
| 492 | <td>Done [dmik]</td> | 
|---|
| 493 |  | 
|---|
| 494 | <td><code>Region</code> class is fully implemented. However, there's a | 
|---|
| 495 | potential question about using regions with other devices than the | 
|---|
| 496 | screen (since OS/2 requires the presentation space to be specified | 
|---|
| 497 | when operating regions, the screen PS is used for this purpose).</td> | 
|---|
| 498 | </tr> | 
|---|
| 499 |  | 
|---|
| 500 | <tr> | 
|---|
| 501 | <td>Move the frame class and window procedure definition code from the | 
|---|
| 502 | <code>Shell</code> to the <code>Decorations</code> class</td> | 
|---|
| 503 |  | 
|---|
| 504 | <td>Will be done on next steps</td> | 
|---|
| 505 |  | 
|---|
| 506 | <td><code>Decorations</code> is logically a frame too. And if it will | 
|---|
| 507 | be used/subclassed directly (i.e. not by the <code>Shell</code>) then | 
|---|
| 508 | we need this code in it.</td> | 
|---|
| 509 | </tr> | 
|---|
| 510 |  | 
|---|
| 511 | <tr> | 
|---|
| 512 | <td>Try to re-enable <code>SWT.CLIP_SIBLINGS</code> and | 
|---|
| 513 | <code>SWT.CLIP_CHILDREN</code> styles recognition in | 
|---|
| 514 | <code>Control.widgetStyle()</code>, and remove forcing them in | 
|---|
| 515 | <code>Composite.WM_PAINT()</code> and | 
|---|
| 516 | <code>Composite.widgetStyle()</code></td> | 
|---|
| 517 |  | 
|---|
| 518 | <td>Will be done on next steps</td> | 
|---|
| 519 |  | 
|---|
| 520 | <td>In Windows version it is intentionally commented because: "When | 
|---|
| 521 | strict clipping was not enforced on all widgets, poorly written | 
|---|
| 522 | application code could draw outside of the control". Is it the same in | 
|---|
| 523 | OS/2? Possibly, not.</td> | 
|---|
| 524 | </tr> | 
|---|
| 525 | </table> | 
|---|
| 526 | </body> | 
|---|
| 527 | </html> | 
|---|