Merge ccb8e98 into 845d7d3

README.rst

@@ -80,22 +80,19 @@ itself.
 
 .. code-block:: python
 
-    def execute(mp, resampling="nearest"):
+    def execute(mp, dem, land_polygons, resampling="nearest"):
 
-        # Open elevation model.
-        with mp.open("dem") as src:
-            # Skip tile if there is no data available or read data into a NumPy array.
-            if src.is_empty(1):
-                return "empty"
-            else:
-                dem = src.read(1, resampling=resampling)
+        # Skip tile if there is no data available or read data into a NumPy array.
+        if dem.is_empty(1):
+            return "empty"
+        else:
+            dem_data = dem.read(1, resampling=resampling)
 
         # Create hillshade using a built-in hillshade function.
-        hillshade = mp.hillshade(dem)
+        hillshade = mp.hillshade(dem_data)
 
         # Clip with polygons from vector file and return result.
-        with mp.open("land_polygons") as land_file:
-            return mp.clip(hillshade, land_file.read())
+        return mp.clip(hillshade, land_polygons.read())
 
 
 You can then interactively inspect the process output directly on a map in a

doc/source/index.rst

@@ -61,22 +61,19 @@ vector dataset could look like this:
 
     # content of hillshade.py
 
-    def execute(mp, resampling="nearest"):
+    def execute(mp, dem, land_polygons, resampling="nearest"):
 
-        # Open elevation model.
-        with mp.open("dem") as src:
-            # Skip tile if there is no data available or read data into a NumPy array.
-            if src.is_empty(1):
-                return "empty"
-            else:
-                dem = src.read(1, resampling=resampling)
+        # Skip tile if there is no data available or read data into a NumPy array.
+        if dem.is_empty(1):
+            return "empty"
+        else:
+            dem_data = dem.read(1, resampling=resampling)
 
         # Create hillshade using a built-in hillshade function.
-        hillshade = mp.hillshade(dem)
+        hillshade = mp.hillshade(dem_data)
 
         # Clip with polygons from vector file and return result.
-        with mp.open("land_polygons") as land_file:
-            return mp.clip(hillshade, land_file.read())
+        return mp.clip(hillshade, land_polygons.read())
 
 
 Examine the result in your browser by serving the process by pointing it to

mapchete/_processing.py

@@ -87,18 +87,25 @@ def _execute(self):
         process_func = get_process_func(
             process_path=self.process_path, config_dir=self.config_dir
         )
+        mp_obj = MapcheteProcess(
+            tile=self.tile,
+            params=self.process_func_params,
+            input=self.input,
+            output_params=self.output_params
+        )
         try:
             with Timer() as t:
+                params = [
+                    # magic mp object
+                    mp_obj if name == "mp"
+                    # process input
+                    else mp_obj.input[name] if name in mp_obj.input
+                    # process parameter
+                    else value
+                    for name, value in self.process_func_params.items()
+                ]
                 # Actually run process.
-                process_data = process_func(
-                    MapcheteProcess(
-                        tile=self.tile,
-                        params=self.process_func_params,
-                        input=self.input,
-                        output_params=self.output_params
-                    ),
-                    **self.process_func_params
-                )
+                process_data = process_func(*params)
         except MapcheteNodataTile:
             raise
         except Exception as e:

mapchete/config.py

@@ -381,7 +381,7 @@ def input(self):
             for k, v in raw_inputs.items():
                 # for files and tile directories
                 if isinstance(v, str):
-                    logger.debug("load input reader for simple input %s",  v)
+                    logger.debug("load input reader for simple input %s", v)
                     try:
                         reader = load_input_reader(
                             dict(
@@ -481,10 +481,54 @@ def process_func(self):
         )
 
     def get_process_func_params(self, zoom):
-        return {
-            k: v for k, v in self.params_at_zoom(zoom).items()
-            if k in inspect.signature(self.process_func).parameters
-        }
+        """
+        Return process function parameters for zoom.
+
+        The dictionary returned is a snapshot for given zoom which combines custom process
+        parameters and input datasets.
+
+        This function also checks whether parameter names are also used as input names and
+        raises a MapcheteConfigError if this is the case.
+
+        Parameters
+        ----------
+        zoom : int
+            zoom level
+
+        Returns
+        -------
+        parameter map : dictionary
+        """
+        if zoom not in self.init_zoom_levels:
+            raise ValueError("zoom level not available with current configuration")
+        process_func_params = inspect.signature(self.process_func).parameters
+        all_config_params = self.params_at_zoom(zoom)
+
+        # look for config parameters which map on process function parameters
+        inputs = set(all_config_params["input"].keys()) & set(process_func_params)
+        custom_params = set(all_config_params.keys()) & set(process_func_params)
+
+        # verify no parameter name intersection is configured
+        intersecting = custom_params & inputs
+        if intersecting:
+            raise MapcheteConfigError(
+                "custom parameters and inputs cannot have the same key: %s" % intersecting
+            )
+
+        # bring all together
+        return OrderedDict([
+            # set mp value
+            (k, None) if k == "mp"
+            # input values from configuration
+            else (k, all_config_params["input"][k]) if k in all_config_params["input"]
+            # custom values from configuration
+            else (k, all_config_params[k]) if k in all_config_params
+            # default values from process function
+            else (k, v.default)
+            for k, v in process_func_params.items()
+            # excludes **kwargs from process function
+            if v.kind != v.VAR_KEYWORD
+        ])
 
     def get_inputs_for_tile(self, tile):
 

mapchete/processes/contours.py

@@ -7,6 +7,8 @@
 
 def execute(
     mp,
+    dem=None,
+    clip=None,
     resampling="nearest",
     interval=100,
     field='elev',
@@ -19,17 +21,14 @@ def execute(
     **kwargs
 ):
     """
-    Generate hillshade from DEM.
+    Generate contours from DEM.
 
-    Inputs
-    ------
+    Parameters
+    ----------
     dem
         Input DEM.
     clip (optional)
         Vector data used to clip output.
-
-    Parameters
-    ----------
     resampling : str (default: 'nearest')
         Resampling used when reading from TileDirectory.
     interval : integer
@@ -64,26 +63,25 @@ def execute(
     list of GeoJSON-like features
     """
     # read clip geometry
-    if "clip" in mp.params["input"]:
-        clip_geom = mp.open("clip").read()
+    if clip is None:
+        clip_geom = []
+    else:
+        clip_geom = clip.read()
         if not clip_geom:
             logger.debug("no clip data over tile")
             return "empty"
-    else:
-        clip_geom = []
 
-    with mp.open("dem",) as dem:
-        logger.debug("reading input raster")
-        dem_data = dem.read(
-            resampling=resampling,
-            matching_method=td_matching_method,
-            matching_max_zoom=td_matching_max_zoom,
-            matching_precision=td_matching_precision,
-            fallback_to_higher_zoom=td_fallback_to_higher_zoom
-        )
-        if dem_data.mask.all():
-            logger.debug("raster empty")
-            return "empty"
+    logger.debug("reading input raster")
+    dem_data = dem.read(
+        resampling=resampling,
+        matching_method=td_matching_method,
+        matching_max_zoom=td_matching_max_zoom,
+        matching_precision=td_matching_precision,
+        fallback_to_higher_zoom=td_fallback_to_higher_zoom
+    )
+    if dem_data.mask.all():
+        logger.debug("raster empty")
+        return "empty"
 
     logger.debug("calculate hillshade")
     contours = mp.contours(

mapchete/processes/convert.py

@@ -6,6 +6,8 @@
 
 def execute(
     mp,
+    raster=None,
+    clip=None,
     resampling="nearest",
     band_indexes=None,
     td_matching_method="gdal",
@@ -21,15 +23,12 @@ def execute(
     """
     Convert and optionally clip input raster data.
 
-    Inputs
-    ------
+    Parameters
+    ----------
     raster
         Singleband or multiband data input.
     clip (optional)
         Vector data used to clip output.
-
-    Parameters
-    ----------
     resampling : str (default: 'nearest')
         Resampling used when reading from TileDirectory.
     band_indexes : list
@@ -66,27 +65,26 @@ def execute(
     np.ndarray
     """
     # read clip geometry
-    if "clip" in mp.params["input"]:
-        clip_geom = mp.open("clip").read()
+    if clip is None:
+        clip_geom = []
+    else:
+        clip_geom = clip.read()
         if not clip_geom:
             logger.debug("no clip data over tile")
             return "empty"
-    else:
-        clip_geom = []
 
-    with mp.open("raster",) as raster:
-        logger.debug("reading input raster")
-        raster_data = raster.read(
-            indexes=band_indexes,
-            resampling=resampling,
-            matching_method=td_matching_method,
-            matching_max_zoom=td_matching_max_zoom,
-            matching_precision=td_matching_precision,
-            fallback_to_higher_zoom=td_fallback_to_higher_zoom
-        )
-        if raster_data.mask.all():
-            logger.debug("raster empty")
-            return "empty"
+    logger.debug("reading input raster")
+    raster_data = raster.read(
+        indexes=band_indexes,
+        resampling=resampling,
+        matching_method=td_matching_method,
+        matching_max_zoom=td_matching_max_zoom,
+        matching_precision=td_matching_precision,
+        fallback_to_higher_zoom=td_fallback_to_higher_zoom
+    )
+    if raster_data.mask.all():
+        logger.debug("raster empty")
+        return "empty"
 
     if scale_offset != 0.:
         logger.debug("apply scale offset %s", scale_offset)

mapchete/processes/hillshade.py

@@ -5,6 +5,8 @@
 
 def execute(
     mp,
+    dem=None,
+    clip=None,
     resampling="nearest",
     azimuth=315.0,
     altitude=45.0,
@@ -18,17 +20,14 @@ def execute(
     **kwargs
 ):
     """
-    Extract contour lines from DEM.
+    Calculate hillshade from DEM.
 
-    Inputs
-    ------
+    Parameters
+    ----------
     dem
         Input DEM.
     clip (optional)
         Vector data used to clip output.
-
-    Parameters
-    ----------
     resampling : str (default: 'nearest')
         Resampling used when reading from TileDirectory.
     azimuth : float
@@ -66,26 +65,25 @@ def execute(
     np.ndarray
     """
     # read clip geometry
-    if "clip" in mp.params["input"]:
-        clip_geom = mp.open("clip").read()
+    if clip is None:
+        clip_geom = []
+    else:
+        clip_geom = clip.read()
         if not clip_geom:
             logger.debug("no clip data over tile")
             return "empty"
-    else:
-        clip_geom = []
 
-    with mp.open("dem",) as dem:
-        logger.debug("reading input raster")
-        dem_data = dem.read(
-            resampling=resampling,
-            matching_method=td_matching_method,
-            matching_max_zoom=td_matching_max_zoom,
-            matching_precision=td_matching_precision,
-            fallback_to_higher_zoom=td_fallback_to_higher_zoom
-        )
-        if dem_data.mask.all():
-            logger.debug("raster empty")
-            return "empty"
+    logger.debug("reading input DEM")
+    dem_data = dem.read(
+        resampling=resampling,
+        matching_method=td_matching_method,
+        matching_max_zoom=td_matching_max_zoom,
+        matching_precision=td_matching_precision,
+        fallback_to_higher_zoom=td_fallback_to_higher_zoom
+    )
+    if dem_data.mask.all():
+        logger.debug("raster empty")
+        return "empty"
 
     logger.debug("calculate hillshade")
     hillshade = mp.hillshade(