Detects missing svx and description files

parsers/caves.py

@@ -1,9 +1,11 @@
 import os
 import re
+from pathlib import Path
 
 from django.conf import settings
 from django.db import transaction
 
+from troggle.settings import SURVEX_DATA, EXPOWEB
 from troggle.core.models.troggle import DataIssue
 from troggle.core.models.caves import Area, Cave, Entrance, CaveSlug, EntranceSlug, CaveAndEntrance
 '''Reads all the cave description data by parsing the xml files (stored as e.g. :EXPOWEB:/cave-data/1623-161.html )
@@ -33,7 +35,7 @@ def readcaves():
 
         
         area_1623 = Area.objects.update_or_create(short_name = "1623", parent = None)
-        # This seems to retrun a tuple, not a single object! i.e. (<Area: 1623>, True)
+        # This seems to return a tuple, not a single object! i.e. (<Area: 1623>, True)
         #print(f' ! - READ CAVES: {area_1623}')
 
         area_1626 = Area.objects.update_or_create(short_name = "1626", parent = None)
@@ -106,7 +108,11 @@ def readentrance(filename):
     context = "in file %s" % filename
     #print("Reading file ENTRANCE {} / {}".format(settings.ENTRANCEDESCRIPTIONS, filename))
     entrancecontentslist = getXML(contents, "entrance", maxItems = 1, context = context)
-    if len(entrancecontentslist) == 1:
+    if len(entrancecontentslist) != 1:
+        message = f'! BAD ENTRANCE at "{filename}"'
+        DataIssue.objects.create(parser='caves', message=message)
+        print(message)
+    else:
         entrancecontents = entrancecontentslist[0]
         non_public = getXML(entrancecontents, "non_public", maxItems = 1, context = context)
         name = getXML(entrancecontents, "name", maxItems = 1, context = context)
@@ -192,7 +198,11 @@ def readcave(filename):
         contents = f.read()
     context = " in file %s" % filename
     cavecontentslist = getXML(contents, "cave", maxItems = 1, context = context)
-    if len(cavecontentslist) == 1:
+    if len(cavecontentslist) != 1:
+        message = f'! BAD CAVE at "{filename}"'
+        DataIssue.objects.create(parser='caves', message=message)
+        print(message)
+    else:
         cavecontents = cavecontentslist[0]
         non_public = getXML(cavecontents, "non_public", maxItems = 1, context = context)
         slugs = getXML(cavecontents, "caveslug", maxItems = 1, context = context)
@@ -281,20 +291,31 @@ def readcave(filename):
                 primary = False
 
             for entrance in entrances:
-                slug = getXML(entrance, "entranceslug", maxItems = 1, context = context)[0]
+                eslug = getXML(entrance, "entranceslug", maxItems = 1, context = context)[0]
                 letter = getXML(entrance, "letter", maxItems = 1, context = context)[0]
                 try:
-                    if slug in entrances_xslug:
+                    if eslug in entrances_xslug:
-                        entrance = entrances_xslug[slug]
+                        entrance = entrances_xslug[eslug]
                     else:
-                        entrance = Entrance.objects.get(entranceslug__slug = slug)
+                        entrance = Entrance.objects.get(entranceslug__slug = eslug)
-                        entrances_xslug[slug] = entrance
+                        entrances_xslug[eslug] = entrance
                     ce = CaveAndEntrance.objects.update_or_create(cave = c, entrance_letter = letter, entrance = entrance)
                 except:
-                    message = " ! Entrance setting failure, slug: %s letter: %s" % (slug, letter)
+                    message = f' ! Entrance setting failure, slug:"{slug}" letter:"{letter}" cave:"{c}" filename:"{filename}"'
                     DataIssue.objects.create(parser='caves', message=message)
                     print(message)
-
+                    
+            if survex_file[0]:
+                if not (Path(SURVEX_DATA) / survex_file[0]).is_file():
+                    message = f' ! survex filename does not exist :LOSER:"{survex_file[0]}" in "{filename}"'
+                    DataIssue.objects.create(parser='caves', message=message, url=f'/cave/{slug}/edit/')
+                    print(message)
+                    
+            if description_file[0]:
+                if not (Path(EXPOWEB) / description_file[0]).is_file():
+                    message = f' ! description filename does not exist :{EXPOWEB}:"{description_file[0]}" in "{filename}"'
+                    DataIssue.objects.create(parser='caves', message=message, url=f'/cave/{slug}/edit/')
+                    print(message)
 
 def getXML(text, itemname, minItems = 1, maxItems = None, printwarnings = True, context = ""):
     items = re.findall("<%(itemname)s>(.*?)</%(itemname)s>" % {"itemname": itemname}, text, re.S)
@@ -306,7 +327,7 @@ def getXML(text, itemname, minItems = 1, maxItems = None, printwarnings = True,
         print(message)
         
     if maxItems is not None and len(items) > maxItems and printwarnings:
-        message = " ! %(count)i %(itemname)s found, no more than %(max)i expected" % {"count": len(items),
+        message = " ! %(count)i %(itemname)s found, no more than %(max)i expected in this XML unit " % {"count": len(items),
                                                                                "itemname": itemname,
                                                                                "max": maxItems} + context
         DataIssue.objects.create(parser='caves', message=message)