tests: update should_get tests to .data 2014.10 format

algo/tests/stanford-data.should_get

 !LAUNCH: ../../vidjil -G ../../germline/IGH -r 5 -d ../../data/Stanford_S22.fasta ; cat out/vidjil.data | sh format-json.sh
 
 $ Number of reads
-1:"reads_total" : [ 13153 ] ,
+1:"total" : [ 13153 ] ,
 
 $ Number of segmented reads
-1:"reads_segmented" : [ 13139 ] ,
+1:"segmented" :  [ 13139 ] 
 
 $ Most abundant window
-1:"window" : "CCACCTATTACTGTACCCGGGAGGAACAATATAGCAGCTGGTACTTTGACTTCTGGGGCC", "size" : [ 8 ]
+1:"id" : "CCACCTATTACTGTACCCGGGAGGAACAATATAGCAGCTGGTACTTTGACTTCTGGGGCC", "reads" : [ 8 ]

algo/tests/stanford-fuse.should_get

@@ -4,7 +4,7 @@ $ Points list
 f1:"point": [ "", "" ]
 
 $ Most abundant window, twice, fused 
-f1:"window": "CCACCTATTACTGTACCCGGGAGGAACAATATAGCAGCTGGTACTTTGACTTCTGGGGCC", "id": .*, "size": [ 8, 8 ]
+f1:"id": "CCACCTATTACTGTACCCGGGAGGAACAATATAGCAGCTGGTACTTTGACTTCTGGGGCC", .*"reads": [ 8, 8 ]
 
 # Fails since 49046ca6b97, no more 'others'
 $ Windows that are not in the top 50