Fix file write (use UTF-8 instead of bytes)

Signed-off-by: Elsie Hupp <github@elsiehupp.com>

wikiteam3/dumpgenerator/greeter.py

@@ -7,11 +7,16 @@ def welcome():
     message = ""
     """Opening message"""
     message += "#" * 73
-    message += """
-# Welcome to DumpGenerator %s by WikiTeam (GPL v3)                   #
-# More info at: https://github.com/WikiTeam/wikiteam                    #""" % (
+    message += "\n"
+    welcome_string = "# Welcome to DumpGenerator %s by WikiTeam (GPL v3)" % (
         getVersion()
     )
+    welcome_string += " " * (73 - len(welcome_string) - 1) + "#"
+    message += welcome_string
+    message += "\n"
+    message += (
+        "# More info at: https://github.com/elsiehupp/wikiteam3                  #"
+    )
     message += "\n"
     message += "#" * 73
     message += "\n"

wikiteam3/dumpgenerator/index_php.py

@@ -15,5 +15,5 @@ def saveIndexPHP(config={}, session=None):
         raw = r.text
         delay(config=config, session=session)
         raw = removeIP(raw=raw)
-        with open("%s/index.html" % (config["path"]), "wb") as outfile:
-            outfile.write(bytes(raw, "utf-8"))
+        with open("%s/index.html" % (config["path"]), "w", encoding="utf-8") as outfile:
+            outfile.write(str(raw))

wikiteam3/dumpgenerator/page_special_version.py

@@ -17,5 +17,7 @@ def saveSpecialVersion(config={}, session=None):
         raw = r.text
         delay(config=config, session=session)
         raw = removeIP(raw=raw)
-        with open("%s/Special:Version.html" % (config["path"]), "wb") as outfile:
-            outfile.write(bytes(raw, "utf-8"))
+        with open(
+            "%s/Special:Version.html" % (config["path"]), "w", encoding="utf-8"
+        ) as outfile:
+            outfile.write(str(raw))

wikiteam3/dumpgenerator/util.py

@@ -64,8 +64,12 @@ def removeIP(raw=""):
 def cleanXML(xml=""):
     """Trim redundant info from the XML however it comes"""
     # do not touch XML codification, leave AS IS
+    # EDIT 2022: we are making this explicitly Unicode
+    # for Windows compatibility.
+    # If the encoding has to stay as is, we'll have
+    # to change all the file encodings, as well.
     if re.search(r"</siteinfo>\n", xml):
-        xml = xml.split("</siteinfo>\n")[1]
+        xml = xml.split("</siteinfo>\n")[1].encode("utf-8")
     if re.search(r"</mediawiki>", xml):
-        xml = xml.split("</mediawiki>")[0]
+        xml = xml.split("</mediawiki>")[0].encode("utf-8")
     return xml

wikiteam3/dumpgenerator/xml_dump.py

@@ -35,8 +35,10 @@ def generateXMLDump(config={}, titles=[], start=None, session=None):
             )
         else:
             print("Retrieving the XML for every page from the beginning")
-            xmlfile = open("{}/{}".format(config["path"], xmlfilename), "wb")
-            xmlfile.write(header)
+            xmlfile = open(
+                "{}/{}".format(config["path"], xmlfilename), "w", encoding="utf-8"
+            )
+            xmlfile.write(str(header))
         try:
             r_timestamp = "<timestamp>([^<]+)</timestamp>"
             for xml in getXMLRevisions(config=config, session=session, start=start):
@@ -50,6 +52,9 @@ def generateXMLDump(config={}, titles=[], start=None, session=None):
             print(e)
             print("This API library version is not working")
             sys.exit()
+        except UnicodeEncodeError as e:
+            print(e)
+
     else:
         print(
             'Retrieving the XML for every page from "%s"' % (start and start or "start")
@@ -65,7 +70,7 @@ def generateXMLDump(config={}, titles=[], start=None, session=None):
             xmlfile = open(
                 "{}/{}".format(config["path"], xmlfilename), "w", encoding="utf-8"
             )
-            xmlfile.write(header)
+            xmlfile.write(str(header))
             xmlfile.close()
 
         xmlfile = open(
@@ -98,6 +103,6 @@ def generateXMLDump(config={}, titles=[], start=None, session=None):
             # (logged in errors log)
             c += 1
 
-    xmlfile.write(footer)
+    xmlfile.write(str(footer))
     xmlfile.close()
     print("XML dump saved at...", xmlfilename)