Commit Graph

  • cb13a6d053 "GeoRSS Format Parsing fails with items with no geo-feature". Patch from Edgemaster to fix this failure mode. r=me (Closes #1258) crschmidt 2008-01-11 01:16:18 +00:00
  • 454026baf8 remove test number from wfs tile test. no functional change here. euzuro 2008-01-10 17:03:39 +00:00
  • 11c1ef3f68 The pixelTolerance check on the click handler only works if mousedown events get to the handler. This keeps the handler from failing in cases where mousedown is stopped and pixelTolerance is set. r=crschmidt (closes #1246) Tim Schaub 2008-01-09 09:55:22 +00:00
  • 5b31ee3c0a Giving the drag handler a stopDown property. Set this to true in the constructor options to let mousedown events propagate. r=crschmidt (closes #1247) Tim Schaub 2008-01-09 09:50:21 +00:00
  • f176d12dc1 Fix redrawing of tiles in Safari, Opera, which don't fire onlod events unless the img.src changes. crschmidt 2008-01-09 08:36:41 +00:00
  • d864adeee0 Adding OpenLayers.Number.format for string formatted numbers. Thanks for initiating this Andreas. Nice pairing with you. r=me (closes #1253) Tim Schaub 2008-01-08 18:22:39 +00:00
  • dc77bd5d5f Removing methods copied from the drag handler. Tim Schaub 2008-01-08 17:34:34 +00:00
  • 21f2605b55 click handler's cachedEvent property isn't needed. r=tschaub. (closes #1252) Éric Lemoine 2008-01-07 23:49:59 +00:00
  • 9193ef67d5 Gutting the OverviewMap control to give it an update. Now uses a click handler and a drag handler instead of its own event handling code. In addition, the control now has a minRecSize property. When the extent rectangle is smaller than the specified size, its style is switched. By default, we provide a rectangle replacement graphic. This can be customized with CSS. Finally, I'm tucking in one non-API property. The dragging arg sent to map.setCenter is now stored at map.dragging. This gives easy reference to dragging state where a reference to the navigation control and its drag handler is not convenient. r=pspencer (closes #1244) Tim Schaub 2008-01-07 20:54:28 +00:00
  • 580a44a1b3 Getting explicit about precision. We now support thirteen significant digits in coordinates when testing for geometry intersections. r=pspencer (closes #1245) Tim Schaub 2008-01-07 19:54:44 +00:00
  • ef3ad090c1 If you'd like to try OpenLayers but only have Netscape 2.0, please buy a new browser. Tim Schaub 2008-01-06 22:15:43 +00:00
  • d7cd93abf3 If your browser doesn't support the script tag, you have no business running these tests. Tim Schaub 2008-01-06 22:10:28 +00:00
  • 02ad66f580 keep local value (closes #1248) Tim Schaub 2008-01-05 00:39:33 +00:00
  • d516b63fe3 fix typo in Ajax.js. Thanks david reksten for reporting this. (closes #1242) Éric Lemoine 2008-01-03 09:58:39 +00:00
  • caefde18fe implement support for alternative js minimization in build system, and if jsmin/minimize is not available, just spit out the uncompressed version. crschmidt 2008-01-03 06:40:27 +00:00
  • c63ac2d80f Make buildUncompressed executable crschmidt 2008-01-03 05:52:37 +00:00
  • 6048cdc108 Constructor doc comment for control panel. Tim Schaub 2008-01-03 00:30:33 +00:00
  • d80bb6bd74 Oh please Google, give us some well named methods that solve this issue. We muddle around with obfuscated code trying in vein to solve your pesky setCenter flicker. This brings us a bit closer with even sized viewports but is still embarassingly wrong for users panning across the dateline with odd sized viewport. r=crschmidt (closes #1174) Tim Schaub 2008-01-02 19:36:12 +00:00
  • 4effccbb97 Uncommit unrelated part of r5433. crschmidt 2008-01-02 19:27:14 +00:00
  • af76852841 Giving vector features an onScreen method. By default, this uses geometry.intersects. If a rougher approximation will do, call with boundsOnly set to true. r=crschmidt (closes #1238) Tim Schaub 2008-01-02 19:16:58 +00:00
  • 6791cf6086 Point, path, and poly handlers now clean up before calling callbacks. r=crschmidt (closes #1239) Tim Schaub 2008-01-02 19:06:07 +00:00
  • d643492888 Proper handling of virtual vertices so we don't leave them on the layer. r=crschmidt (closes #1230) Tim Schaub 2008-01-02 18:54:33 +00:00
  • 67e260775e Make MapServer multi-url selection deterministic. (Closes #803) crschmidt 2008-01-02 15:23:45 +00:00
  • 09e0d3ec81 Update copyrights to 2008. crschmidt 2008-01-01 08:14:55 +00:00
  • 7cb4ba4354 adding click handler as dependency to navigation control Tim Schaub 2007-12-31 23:41:36 +00:00
  • b24f50efe3 Spaces only. Tim Schaub 2007-12-31 00:43:51 +00:00
  • 16d9d36da2 Slightly more informative docs. crschmidt 2007-12-28 13:11:36 +00:00
  • e636e82455 reordering tests for a bit easier reading Tim Schaub 2007-12-26 23:49:20 +00:00
  • 9425869c2b Match XHR behavor at 2.5. The browser default content type is application/xml, so we set that as our default contentType option. Looks like there is no browser default for encoding, so we remove that. r=crschmidt (closes #1232) Tim Schaub 2007-12-24 18:44:06 +00:00
  • f4df2a3b2a Changing prototype.js ajax stuff so exceptions aren't swallowed by default. If you want this behavior, set request.options.onException to something that swallows, or use OpenLayers.Ajax.Responders.register for onException. r=crschmidt (closes #1232) Tim Schaub 2007-12-21 23:42:05 +00:00
  • 86116ddd73 Making feature handler call over and out callbacks just once per mouseover and mouseout (of a feature). r=elemoine (closes #1226) Tim Schaub 2007-12-21 08:48:08 +00:00
  • 33e313b059 Separating jsdoc style docs into their own block. The new NaturalDocs doesn't parse these mixed blocks, giving us no class descriptions with the two together. (closes #1229) Tim Schaub 2007-12-21 08:28:09 +00:00
  • 2447848e7f Coding style: dont use local variable when it's not necessary, remove new lines. No functional changes, tests pass on FF2, IE6, IE7 and safari 3.0.4 windows. Frédéric Junod 2007-12-21 07:28:05 +00:00
  • c056118b59 Small patch to removeItem so that all instances of a value in the array are removed, even if they are consecutive by reversing the order in which the array is enumerated. Updated tests and checked in Safari 3, FF2. (closes #1228). Paul Spencer 2007-12-20 20:22:36 +00:00
  • 2931cd3065 Add style property "display" to vector feature style. If display is "none" the feature won't be rendered and there's a DOM node for that feature it'll be removed. Commiting this on behalf of ahocevar. Thanks Andreas for the patch. r=crschmidt. (closes #1173). Éric Lemoine 2007-12-20 19:20:23 +00:00
  • 2d1099d60d Abort XMLHttpRequest on tile.destroy for WFS. The loadURL function now returns a request object. Thanks pgiraud for the fix. r=crschmidt (closes #964) Tim Schaub 2007-12-20 17:28:51 +00:00
  • 59dd8b2d99 Fix to "style with no rule returns empty array" via the 'unruly.patch' from tschaub. -10 points for bad pun, +10 points for catching this in his review of the SLD code. r=me (Closes #1223) crschmidt 2007-12-20 13:01:19 +00:00
  • 3268e53ed1 Serious rewrite in the Ajax namespace by pgiraud to keep up with prototype.js 1.6. Thanks for the effort and for suffering numerous revisions. Now, let's cross our fingers... r=crschmidt,me (closes #1170) Tim Schaub 2007-12-20 09:26:53 +00:00
  • 44543b3e36 adding manual test for the feature handler Tim Schaub 2007-12-20 08:49:15 +00:00
  • 36e04a689f Adding OpenLayers.Array.filter to mimic Array.prototype.filter. Adding panel.getControlsBy and related methods that use filter. Reworking map.getBy to use filter as well. r=elemoine (closes #1203) Tim Schaub 2007-12-20 08:12:19 +00:00
  • c44b028a30 Fix ND comment typo Frédéric Junod 2007-12-20 07:37:26 +00:00
  • 52fb422c49 fix sdl broken url Frédéric Junod 2007-12-20 06:57:48 +00:00
  • 72dd4047af doc change only - feature vector not vector feature Tim Schaub 2007-12-20 06:47:28 +00:00
  • 3751402ff8 importNode in XML.Format serializer to prevent reports of adding the node to the wrong DOM element. This fixes tests in Webkit-based browsers and FF3, which enforces this constraint. r=tschaub (Closes #1219) crschmidt 2007-12-20 04:15:26 +00:00
  • 2e507b68b7 Fix for "When Navigation control is inactive, doubleclick zooms in", using the new click handler that Tim and I created over the weekend. (Hooray for handlers!) r=tschaub (Closes #1068) crschmidt 2007-12-20 04:11:52 +00:00
  • a22909bfce correcting url for apidocs in doc readme.txt Tim Schaub 2007-12-20 03:58:36 +00:00
  • 78bf7f5912 correcting url for apidocs in doc readme.txt Tim Schaub 2007-12-20 03:58:36 +00:00
  • 981f00dd6d Removing the explicit maxResolution from the TileCache layer. We now depend on resolutions being properly set on the map or directly on the layer. r=crschmidt (closes #1212). Tim Schaub 2007-12-20 03:50:08 +00:00
  • 3b7e63a488 Silence these errors like we did for Opera. With this and #1219 applied, Safari passes all tests. (Time to write more tests.) crschmidt 2007-12-20 03:23:39 +00:00
  • 5babc73615 stop clicks on the point handler - this means no more clicks sneaking through while editing - if you wanted that behavior, speak up - r=crschmidt (closes #1020) Tim Schaub 2007-12-20 03:02:54 +00:00
  • 8eb682c02f More typos. am blind. crschmidt 2007-12-20 02:49:51 +00:00
  • 1d4b8f8c2a Typos. crschmidt 2007-12-20 00:37:37 +00:00
  • 7011c0f470 Big thanks to Andreas Hocevar for this SLD format class - this continues to extend his Rule and Style work by giving us the ability to read SLD docs - onward with style r=tschaub (closes #533). Tim Schaub 2007-12-20 00:24:20 +00:00
  • acc8cfc0f2 add displayProjection API Property on the Map, and on relevant controls. Spherical mercator example now makes use of this, displaying coordinates in lon/lat instead of meters, and permalink/argparser now work in lon/lat as well. this functionality will make using SphericalMercator easier for a number of applications. r=tschaub (Closes #1036) crschmidt 2007-12-19 23:36:34 +00:00
  • 811e35cff3 tidying up the doc menus - somebody has to be the doc nanny around here Tim Schaub 2007-12-19 22:15:32 +00:00
  • 6d6c08fbef Formats now support reprojection using internalProjection and externalProjection properties. These allow for the reprojection of data -- OpenLayers users with SphericalMercator get this built in for EPSG:900913, and other users can use the external proj4js library available from MapBuilder SVN to add support for any number of projections. This means that featres can be, for example, transformed from a KML doc in 4326 to Spherical Mercator before being added to a layer, making using SphericalMercator slightly more enticing. r=elemoine (Closes #1039) crschmidt 2007-12-19 22:07:12 +00:00
  • 7440556585 Add reprojection support on Geometry classes. Geometries can now be transformed in place by using .transform(source, dest). r=elemoine (Closes #1037) crschmidt 2007-12-19 22:04:30 +00:00
  • 50ec61dfc9 nd comments only - wrapping html in code block to make nd happy Tim Schaub 2007-12-19 21:55:28 +00:00
  • e3406bf095 OpenLayers.Layer.Markers - propagate layer's opacity to the markers. Thanks ahughes for the bug report and crschmidt for review and advices. (Closes #1114) Frédéric Junod 2007-12-19 15:00:49 +00:00
  • 6fb48c0cd7 Enables unselecting selected features by clicking outside any feature. This is the default mode of the modify feature control. If you want to use the old mode where a click on a selected feature unselects that feature, set the toggle option to true. If you really want the old mode and not have selected features be unselected when clicking outside any feature, set the clickout option to false. The patch also adds the properties toggleKey and multipleKey to the modify feature control. With these one can temporarily enable the toggle and multiple mode, respectively. See the select-feature.html example to see all this in action. Thanks to pvalsecc and tschaub for the great colloration on this. (closes #1137) Éric Lemoine 2007-12-19 11:09:44 +00:00
  • 65ad59a277 Don't override tileSize, maxExtent and theme when calling Map.setOptions(). Thanks pspencer for the review. (Closes #1206) Frédéric Junod 2007-12-19 07:30:52 +00:00
  • 5c6934b0b0 Allow the use of custom GMapTypes via the Layer.Google. When passed a type in the Layer.Google constructor, we check if the type we are given is not in the map's list of types, and if so, add it. (Closes #1082) r=tschaub crschmidt 2007-12-19 01:33:20 +00:00
  • cc29ab1085 bring back removal of XML prolog. crschmidt 2007-12-18 17:45:09 +00:00
  • 5d4109a3ab Because IE serializes the processing instruction, we can't just grab the firstChild, we have to actually do a getElementsByTagNameNS to be cross-browser friendly. crschmidt 2007-12-18 17:06:09 +00:00
  • 8572258fd1 no need to process the expected data twice. crschmidt 2007-12-18 16:54:15 +00:00
  • d9230297ec Change test to use importNode if available, and add a test for serializing a node as well as a geometry. (The latter test will fail in FF3 until #1219 is fixed.) crschmidt 2007-12-18 16:52:06 +00:00
  • 546d49fc34 Fix (again) the setSize() test. Frédéric Junod 2007-12-18 15:26:53 +00:00
  • ac4f4dba7c Fix the setSize() test. Regression introduced in r5488 Frédéric Junod 2007-12-18 15:14:23 +00:00
  • 32081348bf Fix tests for Safari 3. Composite styles like border have to be read through their component properties. Invalid styles, like filter, return null regardless of what is assigned to them. Paul Spencer 2007-12-18 14:26:08 +00:00
  • f51315c26d Fix ND comment. No functional change. Frédéric Junod 2007-12-18 13:49:19 +00:00
  • 9200676f89 Fix tests for Safari 3. For some reason, you can set the style.overflow property but you have to read it as overflowX or overflowY. Paul Spencer 2007-12-18 13:47:23 +00:00
  • c1b008dcb3 Remove the 'blinking' phenomenon on IE by setting the display elements to 'absolute'. Thanks to pgiraud for the patch (Closes #965) Frédéric Junod 2007-12-18 13:08:35 +00:00
  • d4ff6a4c8b XML Serialization doesn't wrap nodes in documents. This breaks or changes serialization in some browsers. This fixes the XML Format based tests in Safari. (Closes #1218) crschmidt 2007-12-18 12:17:34 +00:00
  • 93d73acd49 for the autommated tester, make the delay_call timeouts ~3 times as long -- presumably, the tests are less 'interactive', so the longer wait doesn't hurt anyone, and this should help prevent failures due to timeouts.l crschmidt 2007-12-18 12:16:23 +00:00
  • bcc623f444 Fix ND comments and indentation. No functional changes. Frédéric Junod 2007-12-18 10:49:29 +00:00
  • a61e5fb91c fix Class ND comment Frédéric Junod 2007-12-18 10:37:04 +00:00
  • 7b04c305cd Run the four eventPane test in more browsers. If this breaks IE, you have permission to beat me senseless for not being able to test it there first. Note that I'm pretty sure this still doens't work with FF3. crschmidt 2007-12-18 05:15:47 +00:00
  • a50bfb4d4e Remove XML prolog from tests. Browsers can't seem to decide whether they like it or not, and #1218 will cause the behavior to change across the board, most likely, so we'll just be slightly more accepting, since the XML prolog is almost definitely unlikely to not matter. crschmidt 2007-12-18 03:44:21 +00:00
  • b2d12df117 revert previous commit. This is actually caused by a bug in Format.XML. (See #1218) crschmidt 2007-12-18 03:33:01 +00:00
  • 611cb91171 Vector test fix crschmidt 2007-12-18 02:44:17 +00:00
  • 1b40772aaf Safari's XMLSerializer support is broken: it doesn't export namespaces, so the KML comes out namespaceless. Wrap the test up so that it doesn't totally fail, and Paul is adding this to KnownIssues. crschmidt 2007-12-18 02:23:41 +00:00
  • 9e6d2a649a FF and Opera have different ideas about how this test should work. FF2 won't work unless the map is loaded 'onload' -- but in Opera, adding something to onload causes opera to believe that the page never finishes loading within the test framework. FF3 works fine with this test if we don't run it in a new window, but FF2 does not. For the time being, make it work in our most well-supported platform -- FF2 -- and add a FIXME to make it work in Opera. crschmidt 2007-12-17 17:09:42 +00:00
  • 5b2f4a2335 loosen the test for XML write by removing the <?xml ... ?> header before comparing the results (fixes test for Safari 3 and FF3) Paul Spencer 2007-12-17 17:05:38 +00:00
  • 5a11ac8ac6 patches Test.AnotherWay._constructor_name to return an appropriate value using 'typeof' instead of an empty string when the constructor of a value could not be correctly determined using the current logic. This fixes many of the tests in Safari 3 and does not appear to affect FF2 or IE6 since the current logic is working for them. Paul Spencer 2007-12-17 16:44:00 +00:00
  • 50ad4d4090 coding standards: line too long, fix ND comment typo Frédéric Junod 2007-12-17 14:45:34 +00:00
  • dc2cc625f3 Do not call preFeatureInsert() and onFeatureInsert() callbacks when the modify feature control and the regular polygon handler internally adds point geometries to the layer. This is accomplished by adding an 'options' argument to the addFeatures() method in Layer.Vector. If that options argument has the silent property set to true, then the preFeatureInsert() and onFeatureInsert() callbacks aren't called. Thanks tschaub and fredj for your input. Thanks crschmidt for the final review. (closes #1148) Éric Lemoine 2007-12-17 14:36:39 +00:00
  • ccf16fa024 Work around the Opera test failures for now, and log to debug that they fail so that someone who is motivated knows to fix them. crschmidt 2007-12-17 13:39:04 +00:00
  • 8479ecb4a0 Don't test reprojection in Opera, since we get different results, and we still want this to b reported as a success, but log them to debug so that people are aware the test is being skipped. crschmidt 2007-12-17 13:35:07 +00:00
  • 175c401e0f ModifyFeature: enable dragging without enabling vertex modifications. Special thanks to tschaub for the collaboration on all the changes to the modify feature control. And thanks to crschmidt for the review. (closes #1188) Éric Lemoine 2007-12-17 10:12:56 +00:00
  • 8733534ad8 Fix test_Style.html on IE. Thanks Andreas for fixing this in no time. Éric Lemoine 2007-12-17 09:56:50 +00:00
  • a1522246a6 Subclasses of markers layer should destroy, then clear their list of markers. Thanks crschmidt for the review. (closes #1123) Éric Lemoine 2007-12-17 09:03:01 +00:00
  • 772b6452ce fix wrapdateline test crschmidt 2007-12-17 06:33:29 +00:00
  • dc05a91d4a don't open a new window: we don't need to in order to run this test. crschmidt 2007-12-17 06:29:46 +00:00
  • 29f2280fb8 missed one -- this function has two maps initialized. crschmidt 2007-12-17 06:23:09 +00:00
  • 8a6f50496f fix more tests in Opera by adding map.destroy() calls. crschmidt 2007-12-17 06:20:10 +00:00
  • 0fd898859f fix more tests in opera crschmidt 2007-12-17 06:14:00 +00:00
  • 03f6ddd5d0 Fix tests in opera by cleaning up after ourselves and destroying our maps. TMS tests now pass. crschmidt 2007-12-17 06:11:59 +00:00
  • 5667311cba add geometry.intersects method for all geometry types (closes #1072) Tim Schaub 2007-12-17 06:05:35 +00:00
  • 189b12d020 Come closer to passing tests in opera by destroying all map objects we create. If we do not do this, then every time we instantiate a new map using the same div, the clientHeight as reported by the browser grows, and eventually we're doing tests with a totally different sized window than we intended to. crschmidt 2007-12-17 06:03:59 +00:00
  • 80fb92686b Opera serializes attributes before namespace declarations. crschmidt 2007-12-17 05:28:06 +00:00