@@ -44,95 +44,175 @@ def get_opt_parser():
4444 return p
4545
4646
47- def diff_values (compare1 , compare2 ):
47+ def diff_values (first_item , second_item ):
4848 """Generically compares two values, returns true if different"""
49- if np .any (compare1 != compare2 ):
49+ if np .any (first_item != second_item ): # comparing items that are instances of class np.ndarray
5050 return True
51- elif type (compare1 ) != type (compare2 ):
51+
52+ elif type (first_item ) != type (second_item ): # comparing items that differ in data type
5253 return True
53- else :
54- return compare1 != compare2
5554
55+ else : # all other use cases
56+ return first_item != second_item
57+
58+
59+ def diff_headers (files , fields ):
60+ """Iterates over all header fields of all files to find those that differ
61+
62+ Parameters
63+ ----------
64+ files: a given list of files to be compared
65+ fields: the fields to be compared
66+
67+ Returns
68+ -------
69+ list
70+ header fields whose values differ across files
71+ """
72+
73+ headers = []
74+
75+ for f in range (len (files )): # for each file
76+ for h in fields : # for each header
77+
78+ # each maneuver is encased in a try block after exceptions have previously occurred
79+ # get the particular header field within the particular file
80+
81+ try :
82+ field = files [f ][h ]
83+
84+ except ValueError :
85+ continue
86+
87+ # filter numpy arrays with a NaN value
88+ try :
89+ if np .all (np .isnan (field )):
90+ continue
91+
92+ except TypeError :
93+ pass
94+
95+ # compare current file with other files
96+ for i in files [f + 1 :]:
97+ other_field = i [h ]
98+
99+ # sometimes field.item doesn't work
100+ try :
101+ # converting bytes to be compared as strings
102+ if isinstance (field .item (0 ), bytes ):
103+ field = field .item (0 ).decode ("utf-8" )
104+
105+ # converting np.ndarray to lists to remove ambiguity
106+ if isinstance (field , np .ndarray ):
107+ field = field .tolist ()
108+
109+ if isinstance (other_field .item (0 ), bytes ):
110+ other_field = other_field .item (0 ).decode ("utf-8" )
111+ if isinstance (other_field , np .ndarray ):
112+ other_field = other_field .tolist ()
56113
57- def diff_header_fields (key , inputs ):
58- """Iterates over a single header field of multiple files"""
114+ except AttributeError :
115+ continue
116+
117+ # if the header values of the two files are different, append
118+ if diff_values (field , other_field ):
119+ headers .append (h )
120+
121+ if headers : # return a list of headers for the files whose values differ
122+ return headers
123+
124+
125+ def diff_header_fields (header_field , files ):
126+ """Iterates over a single header field of multiple files
127+
128+ Parameters
129+ ----------
130+ header_field: a given header field
131+ files: the files to be compared
132+
133+ Returns
134+ -------
135+ list
136+ str for each value corresponding to each file's given header field
137+ """
59138
60139 keyed_inputs = []
61140
62- for i in inputs : # stores each file's respective header files
141+ for i in files :
142+
143+ # each maneuver is encased in a try block after exceptions have previously occurred
144+ # get the particular header field within the particular file
145+
63146 try :
64- field_value = i [key ]
147+ field_value = i [header_field ]
65148 except ValueError :
66149 continue
67150
68- try : # filter numpy arrays
69- if np .all (np .isnan (field_value )):
70- continue
71- except TypeError :
72- pass
73-
74- for x in inputs [1 :]: # compare different values, print all as soon as diff is found
151+ # compare different data types, return all values as soon as diff is found
152+ for x in files [1 :]:
75153 try :
76- data_diff = diff_values (str (x [key ].dtype ), str (field_value .dtype ))
154+ data_diff = diff_values (str (x [header_field ].dtype ), str (field_value .dtype ))
77155
78156 if data_diff :
79157 break
80158 except ValueError :
81159 continue
82160
161+ # string formatting of responses
83162 try :
84- if data_diff : # prints data types if they're different and not if they're not
163+
164+ # if differences are found among data types
165+ if data_diff :
166+ # accounting for how to arrange arrays
85167 if field_value .ndim < 1 :
86168 keyed_inputs .append ("{}@{}" .format (field_value , field_value .dtype ))
87169 elif field_value .ndim == 1 :
88170 keyed_inputs .append ("{}@{}" .format (list (field_value ), field_value .dtype ))
171+
172+ # if no differences are found among data types
89173 else :
90174 if field_value .ndim < 1 :
91- keyed_inputs .append ("{}" . format ( field_value ) )
175+ keyed_inputs .append (field_value )
92176 elif field_value .ndim == 1 :
93- keyed_inputs .append ("{}" .format (list (field_value )))
177+ keyed_inputs .append (list (field_value ))
178+
94179 except UnboundLocalError :
95180 continue
96181
97- if keyed_inputs : # sometimes keyed_inputs is empty lol
98- comparison_input = keyed_inputs [0 ]
182+ for i in range ( len ( keyed_inputs )):
183+ keyed_inputs [ i ] = str ( keyed_inputs [i ])
99184
100- for i in keyed_inputs [1 :]:
101- if diff_values (comparison_input , i ):
102- return keyed_inputs
185+ return keyed_inputs
103186
104187
105- def get_headers_diff (files , opts ):
188+ def get_headers_diff (file_headers , headers ):
106189 """Get difference between headers
107190
108191 Parameters
109192 ----------
110- files : list of files
111- opts: any options included from the command line
193+ file_headers : list of actual headers from files
194+ headers: list of header fields that differ
112195
113196 Returns
114197 -------
115198 dict
116- str: list for each header field which differs, return list of
199+ str: list for each header field which differs, return list of
117200 values per each file
118201 """
119-
120- header_list = [nib .load (f ).header for f in files ]
121202 output = OrderedDict ()
122203
123- if opts .header_fields : # will almost always have a header field
124- # signals "all fields"
125- if opts .header_fields == 'all' :
126- # TODO: header fields might vary across file types, thus prior sensing would be needed
127- header_fields = header_list [0 ].keys ()
128- else :
129- header_fields = opts .header_fields .split (',' )
204+ # if there are headers that differ
205+ if headers :
130206
131- for f in header_fields :
132- val = diff_header_fields ( f , header_list )
207+ # for each header
208+ for header in headers :
133209
210+ # find the values corresponding to the files that differ
211+ val = diff_header_fields (header , file_headers )
212+
213+ # store these values in a dictionary
134214 if val :
135- output [f ] = val
215+ output [header ] = val
136216
137217 return output
138218
@@ -164,8 +244,19 @@ def main():
164244 # suppress nibabel format-compliance warnings
165245 nib .imageglobals .logger .level = 50
166246
167- diff = get_headers_diff (files , opts )
247+ file_headers = [nib .load (f ).header for f in files ]
248+
249+ if opts .header_fields : # will almost always have a header field
250+ # signals "all fields"
251+ if opts .header_fields == 'all' :
252+ # TODO: header fields might vary across file types, thus prior sensing would be needed
253+ header_fields = file_headers [0 ].keys ()
254+ else :
255+ header_fields = opts .header_fields .split (',' )
256+ headers = diff_headers (file_headers , header_fields )
257+ diff = get_headers_diff (file_headers , headers )
168258 data_diff = get_data_md5sums (files )
259+
169260 if data_diff :
170261 diff ['DATA(md5)' ] = data_diff
171262
0 commit comments