introduce a company filtering in cleanFields function

close #51
yoannsculo · Nov 22, 2013 · 4f2e049 · yscialom · Nov 26, 2013 · 4f2e049
1 parent d19c5e8
commit 4f2e049
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 0 deletions.
diff --git a/jobcatcher.py b/jobcatcher.py
@@ -976,9 +976,14 @@ def loadFromHtml(self, filename):
 
     def cleanFields(self):
         self.cleanContract()
+        self.cleanCompany()
         self.cleanLocation()
         self.cleanSalary()
 
+    def cleanCompany(self):
+        self.company = utilities.filter_company_fr(self.company)
+        return
+
     def cleanContract(self):
         self.contract = utilities.filter_contract_fr(self.contract)
         return

diff --git a/utilities.py b/utilities.py
@@ -325,6 +325,10 @@ def filter_contract_fr(contract):
 
     return contract
 
+def filter_company_fr(company):
+    company = re.sub(ur'^\(confidentiel\)$', 'NA', company, re.IGNORECASE)
+    company = re.sub(ur'^confidentiel$', 'NA', company, re.IGNORECASE)
+    return company
 
 def filter_location_fr(location):
     location = re.sub(ur'IDF', "Île-de-France", location)