#---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------#

*

def python():

	
#fonction de nettoyage
def clean(string):
  string_bis=""
  for char in string:
    if re.match("\w",char):
      string_bis+=char
  return string_bis

	
def extract(fichier,rel_l,pos_l):


f=open(fic,"r",encoding="UTF-8")
liste=f.readlines()
f.close()


sent_buf = {} 
couples = set()
lemma_pos_count={}
rel_dic={}
rel_buf={}


for rel in rel_l:
  rel_dic[rel]=rel_dic.get(rel,{})
  rel_buf[rel]=rel_buf.get(rel,{}).copy()
  
  
for pos in pos_l:
  lemma_pos_count[pos]=lemma_pos_count.get(pos,{})
  
  
for line in liste:

  if line.startswith(""):
    fields = re.findall("([^<]+)", line)
    idx, word, lemma, tag, _, _, head, rel, _, _ = fields
          
    lemma=clean(lemma)
    
    sent_buf[idx] = (lemma,tag)
    
    if tag in pos_l:
      lemma_pos_count[tag][lemma]=lemma_pos_count[tag].get(lemma,0)+1
    
    
    if rel in rel_l:
      if tag == pos_l[1]:
        rel_buf[rel][(lemma,tag),head]=rel_buf[rel].get(((lemma,tag), head),0)+1  


        
  if re.search("
",line):
    rels=[key for key in rel_buf.keys()]
    for rel in rels:
      couples=[couple for couple in rel_buf[rel].keys()]

      for couple in couples:
        if sent_buf[couple[1]][1]==pos_l[0]:
          
          rel_dic[rel][(couple[0],sent_buf[couple[1]])]=rel_dic[rel].get((couple[0],sent_buf[couple[1]]),0)+rel_buf[rel][couple]
        
        
      rel_buf[rel]={}
        
    sent_buf = {}

return lemma_pos_count, rel_dic

def graphe_print(lemma_pos_count, rel_dic):

  for pos in lemma_pos_count.keys():
    print(f'@{pos}:#lemma,lemma,count\n')
    for lemma in lemma_pos_count[pos].keys():
      print(f'{pos}_{lemma},{lemma},{lemma_pos_count[pos][lemma]}\n')


  for rel in rel_dic.keys():
    print(f'_{rel},weight\n')
    for couple in rel_dic[rel].keys():
      print(f'{couple[1][0]},--,{couple[0][0]},{rel_dic[rel][couple]}\n')

  return True

def bao4(fichier, rel_l, pos_l):
  (points, lignes) = extract(fichier,rel_l,pos_l)

  graphe_print(points, lignes)

  return

  if __name__ == "__main__":
    pos_l=sys.argv[2:]
    rel_l=[sys.argv[1]]
    fic = "./sortieudpipe-slurp_3208-2021.xml"
    bao4(fic,rel_l,pos_l)

python3 bao4.py analyse_BAO2 relation GOV DEP | curl -X POST -H 'Content-Type: text/csv' --data-binary @- "https://padagraph.magistry.fr/post_csv/nom_de_graphe"

return script