Added script to recompose full corpus prediction after cross validation training

scripts/compileCrossValid.py

+#! /usr/bin/env python3
+
+import sys
+import os
+
+################################################################################
+def printUsageAndExit() :
+  print("USAGE : %s bin/ predOutputDir/ goldOutputDir/"%sys.argv[0], file=sys.stderr)
+  exit(1)
+################################################################################
+
+def getModelModelIndexFilename(d, f) :
+  model = ".".join(d[0].split('.')[:-1]).split('/')[-1]
+  modelIndex = int(model.split('_')[-1])
+  model = "_".join(model.split('_')[:-1])
+  filename = d[0]+"/"+f
+  return model, modelIndex, filename
+
+################################################################################
+if __name__ == "__main__" :
+  if len(sys.argv) != 4 :
+    printUsageAndExit()
+
+  binDir = sys.argv[1]
+  outputDir = sys.argv[2]
+  goldOutputDir = sys.argv[3]
+
+  filesByModel = {}
+
+  for d in os.walk(binDir) :
+    for f in d[2] :
+      if "predicted_eval" in f :
+        model, index, filename = getModelModelIndexFilename(d, f)
+        if model not in filesByModel :
+          filesByModel[model] = []
+        while len(filesByModel[model]) <= index :
+          filesByModel[model].append(["",""])
+        filesByModel[model][index][0] = filename
+      if "test.conllu" in f :
+        model, index, filename = getModelModelIndexFilename(d, f)
+        if model not in filesByModel :
+          filesByModel[model] = []
+        while len(filesByModel[model]) <= index :
+          filesByModel[model].append(["",""])
+        filesByModel[model][index][1] = filename
+
+  with open(goldOutputDir+"/corpus.conllu", "w") as out :
+    for f in filesByModel[list(filesByModel.keys())[0]] :
+      for line in open(f[1], "r") :
+        print(line, end="", file=out)
+
+  for model in filesByModel :
+    outDir = outputDir+"/"+model
+    os.makedirs(outDir, exist_ok=True)
+    with open(outDir+"/corpus.conllu", "w") as out :
+      for f in filesByModel[model] :
+        for line in open(f[0], "r") :
+          print(line, end="", file=out)
+################################################################################