added pandas process (#111)

* added pandas process

Co-authored-by: Trevor James Smith <10819524+Zeitsperre@users.noreply.github.com>

emu/processes/__init__.py

@@ -19,6 +19,7 @@
 from .wps_ncml import NcMLAgg
 from .wps_translation import Translation
 from .wps_geodata import GeoData
+from .wps_pandas import Pandas
 
 
 processes = [
@@ -43,4 +44,5 @@
     NcMLAgg(),
     Translation(),
     GeoData(),
+    Pandas(),
 ]

emu/processes/wps_pandas.py

@@ -0,0 +1,50 @@
+import logging
+
+from pywps import FORMATS, Format, ComplexInput, ComplexOutput, Process
+from pywps.app.Common import Metadata
+
+LOGGER = logging.getLogger("PYWPS")
+
+
+class Pandas(Process):
+    """
+    Notes:
+
+    Create some statisics with pandas from a CSV file.
+    """
+    def __init__(self):
+        inputs = [
+            ComplexInput('csv', 'CSV document',
+                         abstract='A CSV document',
+                         supported_formats=[Format('text/csv', extension='.csv'), FORMATS.TEXT]), ]
+        outputs = [
+            ComplexOutput('output', 'Output',
+                          as_reference=True,
+                          supported_formats=[FORMATS.JSON]), ]
+
+        super(Pandas, self).__init__(
+            self._handler,
+            identifier='pandas',
+            title='Pandas',
+            abstract="Create statisics using Pandas",
+            version='1.0',
+            metadata=[
+                Metadata('User Guide', 'http://emu.readthedocs.io/en/latest/')
+            ],
+            inputs=inputs,
+            outputs=outputs,
+            store_supported=True,
+            status_supported=True)
+
+    def _handler(self, request, response):
+        # optional dependency
+        import pandas as pd
+        # start
+        response.update_status('Pandas Process started.', 0)
+        # read csv
+        df = pd.read_csv(request.inputs['csv'][0].stream)
+        # convert to json
+        response.outputs['output'].data = df.to_json(orient='records')
+        # done
+        response.update_status('Pandas Process completed.', 100)
+        return response

environment.yml

@@ -5,15 +5,17 @@ channels:
 dependencies:
 - pip
 - python>=3.6
-- pywps>=4.4
+- pywps>=4.4.0
 - jinja2
 - click
 - psutil
 - defusedxml
+# GeoJSON
+- geomet
+# pandas
+- pandas
 # opendap support
 - netcdf4
 - xarray
 # tests
 - pytest
-- pip:
-  - geomet

requirements.txt

@@ -6,3 +6,4 @@ defusedxml
 geomet
 xarray
 netCDF4
+pandas

tests/test_wps_caps.py

@@ -27,6 +27,7 @@ def test_wps_caps():
         'ncml',
         'non.py-id',
         'output_formats',
+        'pandas',
         'poly_centroid',
         'show_error',
         'simple_dry_run',

tests/test_wps_pandas.py

@@ -0,0 +1,49 @@
+import json
+from pywps import Service
+from pywps import get_ElementMakerForVersion
+
+from .common import client_for, resource_file, get_output
+from emu.processes.wps_pandas import Pandas
+
+VERSION = "1.0.0"
+
+WPS, OWS = get_ElementMakerForVersion(VERSION)
+
+
+def test_wps_pandas_embedded():
+    client = client_for(Service(processes=[Pandas()]))
+    text = open(resource_file("penguins.csv")).read()
+    request_doc = WPS.Execute(
+        OWS.Identifier('pandas'),
+        WPS.DataInputs(
+            WPS.Input(
+                OWS.Identifier('csv'),
+                WPS.Data(WPS.ComplexData(text, mimeType='text/csv'))
+            )
+        ),
+        WPS.ResponseForm(
+            WPS.RawDataOutput(
+                OWS.Identifier('output')
+            )
+        ),
+        version='1.0.0'
+    )
+    resp = client.post_xml(doc=request_doc)
+    assert resp.status_code == 200
+    penguins = json.loads(resp.data)
+    assert penguins[0]['species'] == "Adelie"
+
+
+def test_wps_pandas_as_ref():
+    client = client_for(Service(processes=[Pandas()]))
+    datainputs = "csv=@xlink:href={0}".format(
+        "https://raw.githubusercontent.com/mwaskom/seaborn-data/master/penguins.csv")
+    resp = client.get(
+        service='wps', request='execute', version='1.0.0',
+        identifier='pandas',
+        datainputs=datainputs,
+        rawdataoutput='output=@mimetype=application/json'
+    )
+    assert resp.status_code == 200
+    penguins = json.loads(resp.data)
+    assert penguins[0]['species'] == "Adelie"