View GitHub Repository Annif

Schedule Inspection
Subscribe
Completed
Push — master ( 8e90e2...462165 )
by Osma
16s queued 12s
created

CombinedCorpus.set_subject_index()   A

↳ Parent: annif.corpus.combine

Complexity

Conditions 3

Size

Total Lines 7
Code Lines 4

Duplication

Lines 0
Ratio 0 %

Importance

Changes 0
Metric Value
eloc 4
dl 0
loc 7
rs 10
c 0
b 0
f 0
cc 3
nop 2
1 
"""Class for combining multiple corpora so they behave like a single corpus"""
2 

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    3
                

                                    
                                                    
                
                
                
import itertools

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    4
                

                                    
                                                    
                
                
                
from .types import DocumentCorpus

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    5
                

                                    
                                                    
                
                
                

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    6
                

                                    
                                                    
                
                
                

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    7
                

                                    
                                                    
                
                
                
class CombinedCorpus(DocumentCorpus):

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    8
                

                                    
                                                    
                
                
                
    """Class for combining multiple corpora so they behave like a single

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    9
                

                                    
                                                    
                
                
                
    corpus"""

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    10
                

                                    
                                                    
                
                
                

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    11
                

                                    
                                                    
                
                
                
    def __init__(self, corpora):

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    12
                

                                    
                                                    
                
                
                
        self._corpora = corpora

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    13
                

                                    
                                                    
                
                
                

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    14
                

                                    
                                                    
                
                
                
    @property

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    15
                

                                    
                                                    
                
                
                
    def documents(self):

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    16
                

                                    
                                                    
                
                
                
        return itertools.chain.from_iterable(

            


            
                                                                                                            
                            
            
                                    
            
            

                
                    17
                

                                    
                                                    
                
                
                
            [corpus.documents for corpus in self._corpora])

            


            
                                                                                                            
                                                                
            
                                    
            
            

                
                    18
                

                                    
                                                    
                
                
                

            


            
                                                        
            
                                    
            
            

                
                    19
                

                                    
                                                    
                
                
                
    def set_subject_index(self, subject_index):

            


            
                                                        
            
                                    
            
            

                
                    20
                

                                    
                                                    
                
                
                
        """Set a subject index for looking up labels that are necessary for

            


            
                                                        
            
                                    
            
            

                
                    21
                

                                    
                                                    
                
                
                
        conversion"""

            


            
                                                        
            
                                    
            
            

                
                    22
                

                                    
                                                    
                
                
                

            


            
                                                        
            
                                    
            
            

                
                    23
                

                                    
                                                    
                
                
                
        for corpus in self._corpora:

            


            
                                                        
            
                                    
            
            

                
                    24
                

                                    
                                                    
                
                
                
            if hasattr(corpus, 'set_subject_index'):

            


            
                                                        
            
                                    
            
            

                
                    25
                

                                    
                                                    
                
                
                
                corpus.set_subject_index(subject_index)

            


            
                                                        
            
                                    
            
            

                
                    26