Parent Directory
| 
Revision Log
| Links to HEAD: | (view) (annotate) |
| Sticky Revision: |
* Thuban/Model/messages.py (LAYER_SHAPESTORE_REPLACED): New message. * Thuban/Model/layer.py (Layer.SetShapeStore): Send LAYER_SHAPESTORE_REPLACED when the shapestore changes. A LAYER_CHANGED will still be sent if the classification changes. * Thuban/UI/classifier.py (Classifier.__init__): Add the map as parameter so we can subscribe to some of its messages (Classifier.__init__): Subscribe to the map's MAP_LAYERS_REMOVED and the layer's LAYER_SHAPESTORE_REPLACED (Classifier.unsubscribe_messages): New. Unsubscribe from message subscribed to in __init__ (Classifier.map_layers_removed) (Classifier.layer_shapestore_replaced): receivers for the messages subscribed to in __init__. Unsubscribe and close the dialog * Thuban/UI/mainwindow.py (MainWindow.OpenLayerProperties): Pass the map to the Classifier dialog * test/test_layer.py (SetShapeStoreTests): Derive from SubscriberMixin as well so we can test messages (SetShapeStoreTests.setUp): Subscribe to some of the layer's messages (SetShapeStoreTests.tearDown): Clear the messages again (SetShapeStoreTests.test_sanity): Expand the doc-string and check for the modified flag too (SetShapeStoreTests.test_set_shape_store_modified_flag): New test to check whether SetShapeStore sets the modified flag (SetShapeStoreTests.test_set_shape_store_different_field_name) (SetShapeStoreTests.test_set_shape_store_same_field) (SetShapeStoreTests.test_set_shape_store_same_field_different_type): Add tests for the messages. This checks both the new LAYER_SHAPESTORE_REPLACED and the older LAYER_CHANGED
(Layer.Destroy): Set all instance variables to None that have direct or indirect references to shapefiles or dbf files to make sure that they do go away and the files are closed.
* Thuban/Model/layer.py (Layer.SetShapeStore): Set the classification to "None" if the type of the field has changed. * test/test_layer.py (SetShapeStoreTests): New. Class with a few test for the Layer.SetShapeStore method
(Layer.TreeInfo): Fixed a bug (a layer does not necessarily have a filename).
(Layer.TreeInfo): Added filename.
* test/test_layer.py (TestLayer.setUp, TestLayer.tearDown): Provide a better way to destroy the layers (TestLayer.test_base_layer, TestLayer.test_arc_layer) (TestLayer.test_point_layer, TestLayer.test_empty_layer) (TestLayer.test_polygon_layer, TestLayer.test_get_field_type): Use the new way to destroy the layers. (TestLayer.test_derived_store): New. Test for using a layer with a DerivedShapeStore * Thuban/Model/layer.py (Layer.SetShapeStore): Only set the filename if the shape store actually has one.
(RasterLayer.__init__): Test if the filename is a valid image file.
Throw IOError otherwise.
(BaseLayer.HasClassification): New.
Defaults to False, but can be overridden by subclasses if they
support classification.
(RasterLayer): New. Defines a new layer that represents an image.
Convert all table users to use the new table interface. This only covers Thuban itself, not GREAT-ER or other applications built on Thuban yet, so the compatibility interface stays in place for the time being but it now issues DeprecationWarnings. Finally, the new Table interface has a new method, HasColumn. * Thuban/Model/table.py (OldTableInterfaceMixin): All methods issue deprecation warnings when they're. The warnings refer to the caller of the method. (OldTableInterfaceMixin.__deprecation_warning): New. Helper method for the deprecation warnings * test/test_table.py: Ignore the deprecation warnings for the old table in the tests in this module. The purpose of the tests is to test the old interface, after all. * test/test_transientdb.py (TestTransientTable.run_iceland_political_tests): Use the constants for the types. Add a test for HasColumn (TestTransientTable.test_transient_joined_table): Adapt to new table interface. Add a test for HasColumn (TestTransientTable.test_transient_table_read_twice): Adapt to new table interface * Thuban/Model/transientdb.py (TransientTableBase.HasColumn) (AutoTransientTable.HasColumn): Implement the new table interface method (AutoTransientTable.ReadRowAsDict, AutoTransientTable.ValueRange) (AutoTransientTable.UniqueValues): Adapt to new table interface * Thuban/Model/layer.py (Layer.SetShapeStore, Layer.GetFieldType): Adapt to new table interface * test/test_layer.py (TestLayer.open_shapefile): Helper method to simplify opening shapefiles a bit easier. (TestLayer.test_arc_layer, TestLayer.test_polygon_layer) (TestLayer.test_point_layer): Use the new helper method (TestLayer.test_get_field_type): New. Test for the GetFieldType method * test/test_dbf_table.py (TestDBFTable.test_has_column): Test for the new table method * test/test_memory_table.py (TestMemoryTable.test_has_column): Test for the new table method HasColumn
(Shape): Since a Shape is immutable only
calculate the bounding box once (the first time compute_bbox() is
called).
(Layer.ShapesBoundingBox): New. Given a list of shape ids, return
the bounding box for the shapes in lat/long coordinates.
(Layer.ShapesInRegion): Fix docstring
to say that the parameter is a tuple of unprojected
points (which is what the callers to this method were assuming).
Also, since the points are unprojected we need to projected them.
(Layer.SetShapeStore): Fixed a bug: don't set the Classification to None if the classfication field is None (ie only a DEFAULT).
(BaseLayer.__init__): Change default value of visible from 1 to True. (Layer.__init__): Change default value of visible from 1 to True.
(Layer.TreeInfo): Add an item to the tree for projection information.
(Layer.GetProjection): New. Needed to implement the interface that the
ProjFrame dialog expects.
First step towards table management. Introduce a simple data
abstraction so that we replace the data a layer uses more easily
in the next step.
* Thuban/Model/data.py: New file with a simple data abstraction
that bundles shapefile and dbffile into one object.
* Thuban/Model/session.py (Session.OpenShapefile): New method to
open shapefiles and return a shape store object
* Thuban/Model/layer.py (Layer.__init__): Pass the data as a store
object instead of a shapefile filename. This introduces a new
instance variable store holding the datastore. For intermediate
backwards compatibility keep the old instance variables.
(open_shapefile): Removed. No longer needed with the shape store.
(Layer.SetShapeStore, Layer.ShapeStore): New methods to set and
get the shape store used by a layer.
(Layer.Destroy): No need to explicitly destroy the shapefile or
table anymore.
* Thuban/UI/mainwindow.py (MainWindow.AddLayer)
(MainWindow.AddLayer): Use the session's OpenShapefile method to
open shapefiles
* Thuban/Model/load.py (ProcessSession.start_layer): Use the
session's OpenShapefile method to open shapefiles
* test/test_classification.py
(TestClassification.test_classification): Use the session's
OpenShapefile method to open shapefiles and build the filename in
a more platform independed way
* test/test_layer.py (TestLayer.setUp, TestLayer.tearDown):
Implement to have a session to use in the tests
(TestLayer.test_arc_layer, TestLayer.test_polygon_layer)
(TestLayer.test_point_layer, TestLayer.test_empty_layer): Use the
session's OpenShapefile method to open shapefiles
(TestLayerLegend.setUp): Instantiate a session so that we can use
it to open shapefiles.
(TestLayerLegend.tearDown): Make sure that all references to
layers and session are removed otherwise we may get a resource
leak
* test/test_map.py (TestMapAddLayer.test_add_layer)
(TestMapWithContents.setUp): Instantiate a session so that we can
use it to open shapefiles.
(TestMapWithContents.tearDown): Make sure that all references to
layers, maps and sessions are removed otherwise we may get a
resource leak
("__main__"): use support.run_tests() so that more info about
uncollected garbage is printed
* test/test_save.py (SaveSessionTest.testSingleLayer): Use the
session's OpenShapefile method to open shapefiles
("__main__"): use support.run_tests() so that more info about
uncollected garbage is printed
* test/test_selection.py (TestSelection.tearDown): Make sure that
all references to the session and the selection are removed
otherwise we may get a resource leak
(TestSelection.get_layer): Instantiate a session so that we can
use it to open shapefiles.
("__main__"): use support.run_tests() so that more info about
uncollected garbage is printed
* test/test_session.py (TestSessionBase.tearDown)
(TestSessionWithContent.tearDown): Make sure that all references
to the session and layers are removed otherwise we may get a
resource leak
(TestSessionWithContent.setUp): Use the session's OpenShapefile
method to open shapefiles
Don't import LAYER_LEGEND_CHANGED. (Layer): Update doc-string since LAYER_LEGEND_CHANGED is not used anymore. (Layer.BoundingBox, Layer.GetFieldType, Layer.NumShapes) (Layer.ShapeType, Layer.Shape): No need to call self.open_shapefile since it's always called in __init__
Rename Color.None to Color.Transparent.
(Layer.ClassChanged): Send a LAYER_CHANGED message, not a LAYER_LEGEND_CHANGED message.
(Layer.SetClassification): Generate a LAYER_CHANGED event instead of a LAYER_LEGEND_CHANGED event.
Handle the cyclic references between
a layer and its classification better, and be sure to disconnect
the classification from the layer when the layer is destroyed
so that we don't maintain a cyclic reference that may not be
garbage collected.
Added ClassChanged() so that the
classification can tell the layer when its data has changed.
(Layer.SetClassification): Accepts None as an arguement to
remove the current classification and correctly handles
adding a new classification.
(SetClassification): prevent recursion between this method and Classification.SetLayer().
(Layer): Add GetFieldType to retreive the kind of data represented by a field.
Fixed name resolution problem.
(Layer): Removed references to fill, stroke,
stroke_width attributes. Made the 'classification' attribute private.
New methods for setting/getting the classification.
* Thuban/Model/layer.py (Layer): Remove the
Set[Fill|Stroke|StrokeWidth]() methods. Code should call the
SetDefault*() methods on the layer's classification object.
(Layer.__init__): Use the new SetDefault*() methods in the
Classification class.
Modified these functions to change the null data in the classification rather than keep these values directly in the Layer class. Menu options to change these values work again.
Added a call to build the tree info for classifications. Commented out unnecessary lines.
Replace user string by _() for i18n.
added classification initializations
(Layer.__init__): Make sure we have an absolute filename.
* Thuban/Model/layer.py (Layer.Destroy): Call the unboundd method with self.
(Layer.Destroy): New. Explicitly close the shapefile and destroy the table.
* Thuban/UI/tree.py (color_string): Removed. No longer used. (SessionTreeCtrl.update_tree, SessionTreeCtrl.add_items): Split update_tree into update_tree and add_items. The tree now uses a more generic way to display the contents of the tree. (SessionTreeCtrl): Add a doc string explaining the TreeInfo method * Thuban/Model/layer.py (Layer.TreeInfo), Thuban/Model/extension.py (Extension.TreeInfo), Thuban/Model/map.py (Map.TreeInfo), Thuban/Model/session.py (Session.TreeInfo): Add TreeInfo methods to implement the new tree view update scheme
* Thuban/Model/layer.py (Layer.open_shapefile): Set bbox to None when the shapefile is empty. (Layer.BoundingBox, Layer.LatLongBoundingBox): Both methods may now return None for empty shapefiles. Update doc-strings.
* Thuban/Model/layer.py (Layer.open_shapefile): Choose a better maximum depth for the tree than shapelib does by default.
* Thuban/Model/layer.py (Layer.ShapesInRegion): Remove some debug prints.
* Thuban/Model/layer.py (Layer.__init__): New instance variable shapetree to hold the shapelib quadtree for the shapefile (Layer.open_shapefile): Create the quad tree. (Layer.ShapesInRegion): New method to return the ids of shapes in a given region using the quad tree.
* Thuban/Model/layer.py (Layer.Shape): list append only takes one argument (python <= 1.5.2 erroneously accepted multiuple arguments)
* Thuban/Model/layer.py (Layer.__init__): New parameter and instance variable stroke_width (Layer.SetStrokeWidth): Set the stroke_width.
Open the shapefile immediately. This will cause an exception in case the file can't be opened and we can display an appropriate message.
import all the source files
This form allows you to request diffs between any two revisions of this file. For each of the two "sides" of the diff, enter a numeric revision.
| [email protected] | ViewVC Help |
| Powered by ViewVC 1.1.26 |