Changeset 3784

Timestamp:

15/04/08 14:27:58 (13 years ago)

Author:

spascoe

Message:

Moved authentication code to just cover MIDAS. This is a hack which will need moving to a configuration file in the future

Location:

DPPP/kml/python/csml2kmlpylon

Files:

• csml2kmlpylon.egg-info/PKG-INFO (modified) (1 diff)
• csml2kmlpylon.egg-info/SOURCES.txt (modified) (3 diffs)
• csml2kmlpylon/controllers/csmlGrapher.py (modified) (4 diffs)

DPPP/kml/python/csml2kmlpylon/csml2kmlpylon.egg-info/PKG-INFO

@@ -1 +1 @@
 Metadata-Version: 1.0
 Name: csml2kmlpylon
-Version: 1.0b
+Version: 1.0.1b
 Summary: Web service providing dynamic content for KML documents generated by the csml2kml package.
 Home-page: UNKNOWN

DPPP/kml/python/csml2kmlpylon/csml2kmlpylon.egg-info/SOURCES.txt

@@ -7 +7 @@
 test.ini
 csml2kmlpylon/__init__.py
-csml2kmlpylon/__init__.pyc
 csml2kmlpylon/websetup.py
 csml2kmlpylon.egg-info/PKG-INFO
@@ -18 +17 @@
 csml2kmlpylon.egg-info/top_level.txt
 csml2kmlpylon/config/__init__.py
-csml2kmlpylon/config/__init__.pyc
 csml2kmlpylon/config/csml2kmlpylon.conf.xml
 csml2kmlpylon/config/environment.py
-csml2kmlpylon/config/environment.pyc
 csml2kmlpylon/config/middleware.py
-csml2kmlpylon/config/middleware.pyc
 csml2kmlpylon/config/routing.py
-csml2kmlpylon/config/routing.pyc
 csml2kmlpylon/controllers/__init__.py
-csml2kmlpylon/controllers/__init__.pyc
+csml2kmlpylon/controllers/auth.py
 csml2kmlpylon/controllers/csmlGrapher.py
-csml2kmlpylon/controllers/csmlGrapher.pyc
 csml2kmlpylon/controllers/error.py
 csml2kmlpylon/controllers/template.py
 csml2kmlpylon/lib/__init__.py
-csml2kmlpylon/lib/__init__.pyc
 csml2kmlpylon/lib/app_globals.py
-csml2kmlpylon/lib/app_globals.pyc
 csml2kmlpylon/lib/base.py
 csml2kmlpylon/lib/helpers.py
-csml2kmlpylon/lib/helpers.pyc
 csml2kmlpylon/model/__init__.py
 csml2kmlpylon/public/index.html
@@ -44 +35 @@
 csml2kmlpylon/tests/test_models.py
 csml2kmlpylon/tests/functional/__init__.py
+csml2kmlpylon/tests/functional/test_auth.py
 docs/index.txt

DPPP/kml/python/csml2kmlpylon/csml2kmlpylon/controllers/csmlGrapher.py

@@ -167 +167 @@
         return csmlPointSeriesFeature
 
-    @authorize(ValidAuthKitUser())
     def plot(self):
         '''
@@ -179 +178 @@
         Response:              An image/png of the time series for the CSML PointSeries feature.
         '''
-        def _plot_feature(feature, dataset):
-            '''
-            Plot a PointSeries feature using matplotlib, into a temporary file.
-            @return: An file-like object which represents the file, which can be read from and also closed
-            as a normal file.
-            @rtype: C{tempfile.NamedTemporaryFile}
-            '''
-            
-            # Make sure that time is considered to be UTC time. This is because all CSML data is supposed to be in UTC,
-            # but actually pylab is "time zone aware" and will interpret the time as local time zone time.
-            # Not quite sure if this would be a problem, but rather enforce it is UTC.
-            def _enforce_UTC_timezone(datestr):
-                if datestr[-1] == 'Z':
-                    return datestr
-                else:
-                    return datestr + 'Z'
-            
-            # Isolate time points
-            times=feature.value.pointSeriesDomain.timePositionList.CONTENT.split()
-            times = map(_enforce_UTC_timezone, times)
-
-            # Make a list of times as float numbers being days passed since start of epoch (here 01-01-0001).
-            # (We need to use the wrapping function csml2kml.utils.parseTimestamp() because unfortunately the matplotlib's
-            #  datetime parser cannot handle timestamps like "2004-02-17T24:00:00" directly.)
-            elapsed_times = map(date2num,
-                                map(csml2kml.utils.parseTimestamp,
-                                    times
-                                    )
-                                )
-
-            # Get values of the measured quantity
-            ql = feature.value.rangeSet.quantityList
-            vals = map(float, ql.CONTENT.split())
-
-            # Make sure the times are sorted in ascending order (and rearrange the values accordingly),
-            # this is because we get them potentially unsorted from GeoServer. Also have to handle the sad fact,
-            # that sometimes we are getting duplicate time-value samples from the MIDAS dataset!
-            sortTimesExplicitly = True
-            if sortTimesExplicitly:
-                reorder = {}
-                for i in range(len(elapsed_times)):
-                    reorder[elapsed_times[i]] = i
-                elapsed_times = reorder.keys()
-                elapsed_times.sort()
-                times2 = []
-                vals2 = []
-                for et in elapsed_times:
-                    times2.append(times[reorder[et]])
-                    vals2.append(vals[reorder[et]])
-                times = times2
-                vals = vals2
-
-            print '---times (' + str(len(times)) + ' of them):\n' + str(times)
-            print '---vals (' + str(len(times)) + ' of them):\n' + str(vals)
-
-            # Note the start and end date (they are actually date+time), and time span between them
-
-            start_date = dates.num2date(elapsed_times[0])
-            end_date = dates.num2date(elapsed_times[-1])
-            span = elapsed_times[-1] - elapsed_times[0]
-
-            # Generate tick locator and formatter; these determine the time axis: locator determines tick positions
-            # and formatter format of time. There will be a total of eight ticks.
-            print '---start_date:' + str(start_date)
-            print '---end_date:' + str(end_date)
-            print '---span: ' + str(span)
-            (tickLocator, tickFormatter) = dates.date_ticker_factory(span, numticks=8)
-
-            # Determine the units of measurement
-            uom=ql.uom.title()
-            if ql.uom.islower():
-                uom = uom.lower()
-            if ql.uom.isupper():
-                uom = uom.upper()
-
-            # Prepare the plot of the figure (matplotlib does the actual plotting actions are postponed on file save)
-            fig = figure()
-            plot_date(elapsed_times, vals, 'b-', xdate=True, lw=1)
-            plot_date(elapsed_times, vals, 'go', markeredgecolor = 'g', xdate=True, lw=2)
-            ax = gca()
-            ax.xaxis.set_major_locator(tickLocator)       
-            ax.xaxis.set_major_formatter(tickFormatter)
-            fig.autofmt_xdate()  # make sure times are shown with no overlapping timestamps
-            time_format = '%d-%b-%Y %H:%M:%S UTC'
-            xlabel('Times between %s and %s ' % (start_date.strftime(time_format), end_date.strftime(time_format)))
-            ylabel('Values [%s]' % uom)
-            title('"%s:%s" (%s station %s)' % (dataset.id, feature.id, dataset.name, feature.description.CONTENT))
-            grid(True)
-
-            # Save the figure to a temporary file
-            tempFile = NamedTemporaryFile(suffix='.png')
-            savefig(tempFile.name)
-            
-            return tempFile
-
-        def _set_response(tempFile):
-            """
-            Set the service's response to an image, containing image read from a temporary location.
-            @param tempFile: An object representing a temporary file
-            @type tempFile: C{tempfile.NamedTemporaryFile}
-            """
-            img = Image.open(tempFile.name)
-            buf = StringIO()
-            img.save(buf, 'PNG')
-            response.content_type = 'image/png'
-            response.content = buf.getvalue()
-            
-        #---------------------- (function main body) -----------------------
 
         # Get parameters from the request object
@@ -296 +187 @@
             raise HTTPBadRequest('Parameters "dataset_id", "feature_id" and "station_name" must be supplied.')
 
+        # Very hacky security because I only want to secure midas
+        secured_datasets = ['midas']
+        if dataset_id in secured_datasets:
+            return self._plot_secured(dataset_id, feature_id, station_name)
+        else:
+            return self._plot2(dataset_id, feature_id, station_name)
+
+    @authorize(ValidAuthKitUser())
+    def _plot_secured(self, dataset_id, feature_id, station_name):
+        return self._plot2(dataset_id, feature_id, station_name)
+
+    def _plot2(self, dataset_id, feature_id, station_name):
         # Try to retrieve the csmlPointSeriesFeature object from the appropriate dataset
         try:
@@ -399 +302 @@
         response.content_type = 'text/html'
         response.content = html
+
+
+
+def _plot_feature(feature, dataset):
+    '''
+    Plot a PointSeries feature using matplotlib, into a temporary file.
+    @return: An file-like object which represents the file, which can be read from and also closed
+    as a normal file.
+    @rtype: C{tempfile.NamedTemporaryFile}
+    '''
+
+    # Make sure that time is considered to be UTC time. This is because all CSML data is supposed to be in UTC,
+    # but actually pylab is "time zone aware" and will interpret the time as local time zone time.
+    # Not quite sure if this would be a problem, but rather enforce it is UTC.
+    def _enforce_UTC_timezone(datestr):
+        if datestr[-1] == 'Z':
+            return datestr
+        else:
+            return datestr + 'Z'
+
+    # Isolate time points
+    times=feature.value.pointSeriesDomain.timePositionList.CONTENT.split()
+    times = map(_enforce_UTC_timezone, times)
+
+    # Make a list of times as float numbers being days passed since start of epoch (here 01-01-0001).
+    # (We need to use the wrapping function csml2kml.utils.parseTimestamp() because unfortunately the matplotlib's
+    #  datetime parser cannot handle timestamps like "2004-02-17T24:00:00" directly.)
+    elapsed_times = map(date2num,
+                        map(csml2kml.utils.parseTimestamp,
+                            times
+                            )
+                        )
+
+    # Get values of the measured quantity
+    ql = feature.value.rangeSet.quantityList
+    vals = map(float, ql.CONTENT.split())
+
+    # Make sure the times are sorted in ascending order (and rearrange the values accordingly),
+    # this is because we get them potentially unsorted from GeoServer. Also have to handle the sad fact,
+    # that sometimes we are getting duplicate time-value samples from the MIDAS dataset!
+    sortTimesExplicitly = True
+    if sortTimesExplicitly:
+        reorder = {}
+        for i in range(len(elapsed_times)):
+            reorder[elapsed_times[i]] = i
+        elapsed_times = reorder.keys()
+        elapsed_times.sort()
+        times2 = []
+        vals2 = []
+        for et in elapsed_times:
+            times2.append(times[reorder[et]])
+            vals2.append(vals[reorder[et]])
+        times = times2
+        vals = vals2
+
+    print '---times (' + str(len(times)) + ' of them):\n' + str(times)
+    print '---vals (' + str(len(times)) + ' of them):\n' + str(vals)
+
+    # Note the start and end date (they are actually date+time), and time span between them
+
+    start_date = dates.num2date(elapsed_times[0])
+    end_date = dates.num2date(elapsed_times[-1])
+    span = elapsed_times[-1] - elapsed_times[0]
+
+    # Generate tick locator and formatter; these determine the time axis: locator determines tick positions
+    # and formatter format of time. There will be a total of eight ticks.
+    print '---start_date:' + str(start_date)
+    print '---end_date:' + str(end_date)
+    print '---span: ' + str(span)
+    (tickLocator, tickFormatter) = dates.date_ticker_factory(span, numticks=8)
+
+    # Determine the units of measurement
+    uom=ql.uom.title()
+    if ql.uom.islower():
+        uom = uom.lower()
+    if ql.uom.isupper():
+        uom = uom.upper()
+
+    # Prepare the plot of the figure (matplotlib does the actual plotting actions are postponed on file save)
+    fig = figure()
+    plot_date(elapsed_times, vals, 'b-', xdate=True, lw=1)
+    plot_date(elapsed_times, vals, 'go', markeredgecolor = 'g', xdate=True, lw=2)
+    ax = gca()
+    ax.xaxis.set_major_locator(tickLocator)       
+    ax.xaxis.set_major_formatter(tickFormatter)
+    fig.autofmt_xdate()  # make sure times are shown with no overlapping timestamps
+    time_format = '%d-%b-%Y %H:%M:%S UTC'
+    xlabel('Times between %s and %s ' % (start_date.strftime(time_format), end_date.strftime(time_format)))
+    ylabel('Values [%s]' % uom)
+    title('"%s:%s" (%s station %s)' % (dataset.id, feature.id, dataset.name, feature.description.CONTENT))
+    grid(True)
+
+    # Save the figure to a temporary file
+    tempFile = NamedTemporaryFile(suffix='.png')
+    savefig(tempFile.name)
+
+    return tempFile
+
+def _set_response(tempFile):
+    """
+    Set the service's response to an image, containing image read from a temporary location.
+    @param tempFile: An object representing a temporary file
+    @type tempFile: C{tempfile.NamedTemporaryFile}
+    """
+    img = Image.open(tempFile.name)
+    buf = StringIO()
+    img.save(buf, 'PNG')
+    response.content_type = 'image/png'
+    response.content = buf.getvalue()
+