Keep names of all products + new BinaryProduct

oda_api/api.py

@@ -16,7 +16,13 @@
 except ModuleNotFoundError:
     pass
 
-from .data_products import NumpyDataProduct, BinaryData, ApiCatalog, GWContoursDataProduct, PictureProduct
+from .data_products import (NumpyDataProduct, 
+                            BinaryData, 
+                            BinaryProduct, 
+                            ApiCatalog, 
+                            GWContoursDataProduct, 
+                            PictureProduct,
+                            ODAAstropyTable)
 from oda_api.token import TokenLocation
 
 from builtins import (bytes, str, open, super, range,
@@ -1160,7 +1166,7 @@
             if hasattr(data, 'name'):
                 name = data.name
 
-            if name.strip() == '':
+            if name is None or name.strip() == '':
                 if product is not None:
                     name = '%s' % product
                 elif instrument is not None:
@@ -1250,28 +1256,38 @@
                          for d in res_json['products']['numpy_data_product_list']])
 
         if 'binary_data_product_list' in res_json['products'].keys():
-            data.extend([BinaryData().decode(d)
-                         for d in res_json['products']['binary_data_product_list']])
+            try:
+                data.extend([BinaryProduct.decode(d)
+                             for d in res_json['products']['binary_data_product_list']])
+            except:
+                data.extend([BinaryData().decode(d)
+                             for d in res_json['products']['binary_data_product_list']])
 
         if 'catalog' in res_json['products'].keys():
             data.append(ApiCatalog(
                 res_json['products']['catalog'], name='dispatcher_catalog'))
 
-        if 'astropy_table_product_ascii_list' in res_json['products'].keys():
-            data.extend([ascii.read(table_text['ascii'])
+        if 'astropy_table_product_ascii_list' in res_json['products'].keys():         
+            data.extend([ODAAstropyTable.decode(table_text, use_binary=False)
                          for table_text in res_json['products']['astropy_table_product_ascii_list']])
 
         if 'astropy_table_product_binary_list' in res_json['products'].keys():
-            data.extend([ascii.read(table_binary)
+            data.extend([ODAAstropyTable.decode(table_binary, use_binary=True)
                          for table_binary in res_json['products']['astropy_table_product_binary_list']])
 
         if 'binary_image_product_list' in res_json['products'].keys():
             data.extend([PictureProduct.decode(bin_image_data)
                          for bin_image_data in res_json['products']['binary_image_product_list']])
 
         if 'text_product_list' in res_json['products'].keys():
-            data.extend([text_data
-                         for text_data in res_json['products']['text_product_list']])
+            try:
+                data.extend([{'name': json.loads(text_data)['name'], 
+                              'value': json.loads(text_data)['value'],
+                              'meta_data': json.loads(text_data)['meta_data']}
+                               for text_data in res_json['products']['text_product_list']])
+            except:
+                data.extend([text_data
+                             for text_data in res_json['products']['text_product_list']])
 
         if 'gw_strain_product_list' in res_json['products'].keys():
             data.extend([TimeSeries(strain_data['value'],

oda_api/data_products.py

@@ -85,10 +85,9 @@
 
     return _l
 
-# not used?
 class ODAAstropyTable(object):
 
-    def __init__(self,table_object,name='astropy table', meta_data={}):
+    def __init__(self,table_object,name=None, meta_data={}):
         self.name=name
         self.meta_data=meta_data
         self._table=table_object
@@ -119,7 +118,7 @@
         if hasattr(table, 'meta'):
             meta = table.meta
 
-        return cls(table, meta_data=meta)
+        return cls(table, meta_data=meta, name=name)
 
     def encode(self,use_binary=False,to_json = False):
 
@@ -179,14 +178,51 @@
         _file_b64_md5 = hashlib.md5(_file_binary).hexdigest()
 
         return _file_b64,_file_b64_md5
-
+    
     def decode(self,encoded_obj):
         return base64.urlsafe_b64decode(encoded_obj.encode('ascii', 'ignore'))
-
+
+
+class BinaryProduct:
+    # New implementation of binary data product. 
+    # The meaning of the methods is more in-line with the rest of the products
+    def __init__(self, bin_data, name=None):
+        self.bin_data = bin_data
+        if name == 'None': name = None
+        self.name = name
+
+    def encode(self):
+        return {
+            'name': self.name,
+            'data': base64.urlsafe_b64encode(self.bin_data).decode(),
+            'md5': hashlib.md5(self.bin_data).hexdigest()
+        }
+
+    @classmethod
+    def decode(cls, encoded_obj):
+        if not isinstance(encoded_obj, dict):
+            encoded_obj = json.loads(encoded_obj)
+
+        name = encoded_obj['name']
+        bin_data = base64.urlsafe_b64decode(encoded_obj['data'].encode('ascii', 'ignore'))
+        decoded_md5 = hashlib.md5(bin_data).hexdigest()
+        assert decoded_md5 == encoded_obj['md5']
+
+        return cls(bin_data, name)
+
+    def write_file(self, file_path):
+        with open(file_path, 'wb') as fd:
+            fd.write(self.bin_data)
+
+    @classmethod        
+    def from_file(cls, file_path, name=None):
+        with open(file_path, 'rb') as fd:
+            bin_data = fd.read()
+        return cls(bin_data, name)
 
 class NumpyDataUnit(object):
 
-    def __init__(self, data, data_header={}, meta_data={}, hdu_type=None, name='table', units_dict=None):
+    def __init__(self, data, data_header={}, meta_data={}, hdu_type=None, name=None, units_dict=None):
         self._hdu_type_list_ = ['primary', 'image', 'table', 'bintable']
 
         self.name=name
@@ -240,9 +276,9 @@
 
 
     @classmethod
-    def from_fits_hdu(cls,hdu,name=''):
+    def from_fits_hdu(cls,hdu,name=None):
 
-        if name=='':
+        if not name:
             name=hdu.name
 
         return cls(data=hdu.data,
@@ -433,7 +469,7 @@
     @classmethod
     def from_pandas(cls, 
                     pandas_dataframe, 
-                    name = 'table', 
+                    name = None, 
                     column_names=[], 
                     units_dict={}, 
                     meta_data = {},
@@ -455,7 +491,7 @@
 
 class NumpyDataProduct(object):
 
-    def __init__(self, data_unit, name='', meta_data={}):
+    def __init__(self, data_unit, name=None, meta_data={}):
 
         self.name=name
 
@@ -621,7 +657,7 @@
 class ApiCatalog(object):
 
 
-    def __init__(self,cat_dict,name='catalog'):
+    def __init__(self,cat_dict,name=None):
         self.name=name
         _skip_list=['meta_ID']
         meta = {}
@@ -703,7 +739,7 @@
                     errors = None,
                     units_spec = {}, # TODO: not used yet
                     time_format = None,
-                    name = 'lightcurve'):
+                    name = None):
 
         data_header = {}
         meta_data = {} # meta data could be attached to both NumpyDataUnit and NumpyDataProduct. Decide on this
@@ -767,9 +803,10 @@
                    name = name)            
 
 class PictureProduct:
-    def __init__(self, binary_data, metadata={}, file_path=None, write_on_creation = False):
+    def __init__(self, binary_data, name=None, metadata={}, file_path=None, write_on_creation = False):
         self.binary_data = binary_data
         self.metadata = metadata
+        self.name = name
         if file_path is not None and os.path.isfile(file_path):
             self.file_path = file_path 
             logger.info(f'Image file {file_path} already exist. No automatical rewriting.')
@@ -784,10 +821,10 @@
         self.img_type = tp
 
     @classmethod
-    def from_file(cls, file_path):
+    def from_file(cls, file_path, name=None):
         with open(file_path, 'rb') as fd:
             binary_data = fd.read()
-        return cls(binary_data, file_path=file_path)
+        return cls(binary_data, name=name, file_path=file_path)
 
     def write_file(self, file_path):
         logger.info(f'Creating image file {file_path}.')
@@ -801,6 +838,7 @@
         output_dict['img_type'] = self.img_type
         output_dict['b64data'] = b64data.decode()
         output_dict['metadata'] = self.metadata
+        output_dict['name'] = self.name
         if self.file_path:
             output_dict['filename'] = os.path.basename(self.file_path)
         return output_dict
@@ -815,6 +853,7 @@
         return cls(binary_data, 
                    metadata = _encoded_data['metadata'],
                    file_path = _encoded_data.get('filename'),
+                   name = _encoded_data.get('name'),
                    write_on_creation = write_on_creation)
 
     def show(self):
@@ -826,8 +865,8 @@
 
 class ImageDataProduct(NumpyDataProduct):  
     @classmethod
-    def from_fits_file(cls,filename,ext=None,hdu_name=None,meta_data={},name=''):
-        npdp = super().from_fits_file(filename,ext=None,hdu_name=None,meta_data={},name='')
+    def from_fits_file(cls,filename,ext=None,hdu_name=None,meta_data={},name=None):
+        npdp = super().from_fits_file(filename,ext=ext,hdu_name=hdu_name,meta_data=meta_data,name=name)
 
         contains_image = cls.check_contains_image(npdp)
         if contains_image:

oda_api/json.py

@@ -2,7 +2,7 @@
 import logging
 import numpy as np
 
-from .data_products import NumpyDataProduct, ODAAstropyTable, PictureProduct, BinaryData
+from .data_products import NumpyDataProduct, ODAAstropyTable, PictureProduct, BinaryData, BinaryProduct
 
 from astropy.io.fits.card import Undefined as astropyUndefined
 
@@ -15,7 +15,7 @@ def default(self, obj):
         if isinstance(obj, astropyUndefined):
             return "UNDEFINED"
 
-        if isinstance(obj, (NumpyDataProduct, ODAAstropyTable, PictureProduct)):
+        if isinstance(obj, (NumpyDataProduct, ODAAstropyTable, PictureProduct, BinaryProduct)):
             return obj.encode()
 
         if isinstance(obj, BinaryData):

requirements.txt

@@ -6,7 +6,7 @@ matplotlib
 numpy
 jsonschema
 astroquery
--e git+https://github.com/oda-hub/dispatcher-app.git@list-products-from-given-source#egg=cdci_data_analysis[test]
+-e git+https://github.com/oda-hub/dispatcher-app.git#egg=cdci_data_analysis[test]
 simplejson
 sentry_sdk
 rdflib

tests/test_data_products.py

@@ -9,8 +9,13 @@
 import os
 import typing
 from oda_api.json import CustomJSONEncoder
+import filecmp
 
-from oda_api.data_products import LightCurveDataProduct, NumpyDataProduct, ODAAstropyTable, PictureProduct
+from oda_api.data_products import (LightCurveDataProduct, 
+                                   NumpyDataProduct, 
+                                   ODAAstropyTable, 
+                                   PictureProduct,
+                                   BinaryProduct)
 from astropy import time as atime
 from astropy import units as u
 from astropy.table import Table
@@ -165,4 +170,15 @@ def test_lightcurve_product_from_arrays(times, values, time_format, expected_uni
     lc = LightCurveDataProduct.from_arrays(times = times, fluxes = values, errors = errors, time_format=time_format)
     assert lc.data_unit[1].units_dict in expected_units_dict_variants
     assert all(lc.data_unit[1].data['TIME'].astype('int') == 59630)
+
+def test_new_binary_product():
+    infile = 'tests/test_data/lc.fits'
+    bin_prod = BinaryProduct.from_file(infile, name='binprd')
+    encoded = bin_prod.encode()
+    assert encoded['name'] == 'binprd'
+    decoded = BinaryProduct.decode(encoded)
+    assert decoded.name == 'binprd'
+    decoded.write_file('decbinprd.foo')
+    assert filecmp.cmp(infile, 'decbinprd.foo')
+    os.remove('decbinprd.foo')