Changes needed to stop the survex parser having to go through the data twiceFaster-survex-parser

Taken from the Django 1.10 upgrade branch

1 files changed, 178 insertions, 74 deletions

diff --git a/parsers/survex.py b/parsers/survex.py
index 294de73..14bd035 100644
--- a/parsers/survex.py
+++ b/parsers/survex.py
@@ -5,20 +5,26 @@ import troggle.settings as settings
 from subprocess import call, Popen, PIPE
 
 from troggle.parsers.people import GetPersonExpeditionNameLookup
+from django.utils.timezone import get_current_timezone
+from django.utils.timezone import make_aware
+
 import re
 import os
+from datetime import datetime
 
+line_leg_regex = re.compile(r"[\d\-+.]+$")
 
-def LoadSurvexLineLeg(survexblock, stardata, sline, comment):
+def LoadSurvexLineLeg(survexblock, stardata, sline, comment, cave):
+    # The try catches here need replacing as they are relativly expensive
     ls = sline.lower().split()
     ssfrom = survexblock.MakeSurvexStation(ls[stardata["from"]])
     ssto = survexblock.MakeSurvexStation(ls[stardata["to"]])
-    
+
     survexleg = models.SurvexLeg(block=survexblock, stationfrom=ssfrom, stationto=ssto)
     if stardata["type"] == "normal":
         try:
             survexleg.tape = float(ls[stardata["tape"]])
-        except ValueError: 
+        except ValueError:
             print("Tape misread in", survexblock.survexfile.path)
             print("Stardata:", stardata)
             print("Line:", ls)
@@ -53,14 +59,17 @@ def LoadSurvexLineLeg(survexblock, stardata, sline, comment):
                 survexleg.compass = 1000
             survexleg.clino = -90.0
         else:
-            assert re.match(r"[\d\-+.]+$", lcompass), ls
-            assert re.match(r"[\d\-+.]+$", lclino) and lclino != "-", ls
+            assert line_leg_regex.match(lcompass), ls
+            assert line_leg_regex.match(lclino) and lclino != "-", ls
             survexleg.compass = float(lcompass)
             survexleg.clino = float(lclino)
-        
+
+        if cave:
+            survexleg.cave = cave
+
         # only save proper legs
         survexleg.save()
-    
+
     itape = stardata.get("tape")
     if itape:
         try:
@@ -80,96 +89,212 @@ def LoadSurvexEquate(survexblock, sline):
 
 def LoadSurvexLinePassage(survexblock, stardata, sline, comment):
     pass
-    
 
 stardatadefault = {"type":"normal", "t":"leg", "from":0, "to":1, "tape":2, "compass":3, "clino":4}
 stardataparamconvert = {"length":"tape", "bearing":"compass", "gradient":"clino"}
 
+regex_comment = re.compile(r"([^;]*?)\s*(?:;\s*(.*))?\n?$")
+regex_ref     = re.compile(r'.*?ref.*?(\d+)\s*#\s*(\d+)')
+regex_star    = re.compile(r'\s*\*[\s,]*(\w+)\s*(.*?)\s*(?:;.*)?$')
+regex_team    = re.compile(r"(Insts|Notes|Tape|Dog|Useless|Pics|Helper|Disto|Consultant)\s+(.*)$(?i)")
+regex_team_member        = re.compile(r" and | / |, | & | \+ |^both$|^none$(?i)")
+regex_qm      = re.compile(r'^\s*QM(\d)\s+?([a-dA-DxX])\s+([\w\-]+)\.(\d+)\s+(([\w\-]+)\.(\d+)|\-)\s+(.+)$')
+
 def RecursiveLoad(survexblock, survexfile, fin, textlines):
     iblankbegins = 0
     text = [ ]
     stardata = stardatadefault
     teammembers = [ ]
- 
-# uncomment to print out all files during parsing 
-    print("Reading file: " + survexblock.survexfile.path)
-    while True:
-        svxline = fin.readline().decode("latin1")
-        if not svxline:
-            return
-        textlines.append(svxline)
+
+    # uncomment to print out all files during parsing 
+    print("  - Reading file: " + survexblock.survexfile.path)
+    stamp = datetime.now()
+    lineno = 0
+
+    # Try to find the cave in the DB if not use the string as before
+    path_match = re.search(r"caves-(\d\d\d\d)/(\d+|\d\d\d\d-?\w+-\d+)/", survexblock.survexfile.path)
+    if path_match:
+        pos_cave = '%s-%s' % (path_match.group(1), path_match.group(2))
+        # print('Match')
+        # print(pos_cave)
+        cave = models.getCaveByReference(pos_cave)
+        if cave:
+            survexfile.cave = cave
+    svxlines = ''
+    svxlines = fin.read().splitlines()
+    # print('Cave - preloop ' + str(survexfile.cave))
+    # print(survexblock)
+    for svxline in svxlines:
+
+        # print(survexblock)
+
+        # print(svxline)
+        # if not svxline:
+        #     print('   - Not survex')
+        #     return
+        # textlines.append(svxline)
+
+        lineno += 1
+
+        # print('   - Line: %d' % lineno)
 
         # break the line at the comment
-        sline, comment = re.match(r"([^;]*?)\s*(?:;\s*(.*))?\n?$", svxline.strip()).groups()
-        
+        sline, comment = regex_comment.match(svxline.strip()).groups()
         # detect ref line pointing to the scans directory
-        mref = comment and re.match(r'.*?ref.*?(\d+)\s*#\s*(\d+)', comment)
+        mref = comment and regex_ref.match(comment)
         if mref:
             refscan = "%s#%s" % (mref.group(1), mref.group(2))
             survexscansfolders = models.SurvexScansFolder.objects.filter(walletname=refscan)
             if survexscansfolders:
                 survexblock.survexscansfolder = survexscansfolders[0]
                 #survexblock.refscandir = "%s/%s%%23%s" % (mref.group(1), mref.group(1), mref.group(2))
-                survexblock.save()   
+                survexblock.save()
             continue
-        
+
+        # This whole section should be moved if we can have *QM become a proper survex command
+        # Spec of QM in SVX files, currently commented out need to add to survex
+        # needs to match regex_qm
+        # ;Serial number   grade(A/B/C/D/X)  nearest-station  resolution-station description
+        # ;QM1	a	hobnob_hallway_2.42	hobnob-hallway_3.42	junction of keyhole passage
+        # ;QM1	a	hobnob_hallway_2.42	-	junction of keyhole passage
+        qmline = comment and regex_qm.match(comment)
+        if qmline:
+            print(qmline.groups())
+            #(u'1', u'B', u'miraclemaze', u'1.17', u'-', None, u'\tcontinuation of rift')
+            qm_no = qmline.group(1)
+            qm_grade = qmline.group(2)
+            qm_from_section = qmline.group(3)
+            qm_from_station = qmline.group(4)
+            qm_resolve_section = qmline.group(6)
+            qm_resolve_station = qmline.group(7)
+            qm_notes = qmline.group(8)
+
+            print('Cave - %s' % survexfile.cave)
+            print('QM no %d' % int(qm_no))
+            print('QM grade %s' % qm_grade)
+            print('QM section %s' % qm_from_section)
+            print('QM station %s' % qm_from_station)
+            print('QM res section %s' % qm_resolve_section)
+            print('QM res station %s' % qm_resolve_station)
+            print('QM notes %s' % qm_notes)
+
+            # If the QM isn't resolved (has a resolving station) thn load it
+            if not qm_resolve_section or qm_resolve_section is not '-' or qm_resolve_section is not 'None':
+                from_section = models.SurvexBlock.objects.filter(name=qm_from_section)
+                # If we can find a section (survex note chunck, named)
+                if len(from_section) > 0:
+                    print(from_section[0])
+                    from_station = models.SurvexStation.objects.filter(block=from_section[0], name=qm_from_station)
+                    # If we can find a from station then we have the nearest station and can import it
+                    if len(from_station) > 0:
+                        print(from_station[0])
+                        qm = models.QM.objects.create(number=qm_no,
+                                                      nearest_station=from_station[0],
+                                                      grade=qm_grade.upper(),
+                                                      location_description=qm_notes)
+            else:
+                print('QM found but resolved')
+
+        #print('Cave -sline ' + str(cave))
         if not sline:
             continue
-        
+
         # detect the star command
-        mstar = re.match(r'\s*\*[\s,]*(\w+)\s*(.*?)\s*(?:;.*)?$', sline)
+        mstar = regex_star.match(sline)
         if not mstar:
             if "from" in stardata:
-                LoadSurvexLineLeg(survexblock, stardata, sline, comment)
+                # print('Cave ' + str(survexfile.cave))
+                # print(survexblock)
+                LoadSurvexLineLeg(survexblock, stardata, sline, comment, survexfile.cave)
+                # print('   - From: ')
+                #print(stardata)
+                pass
             elif stardata["type"] == "passage":
                 LoadSurvexLinePassage(survexblock, stardata, sline, comment)
+                # print('   - Passage: ')
             #Missing "station" in stardata.
             continue
-        
+
         # detect the star command
         cmd, line = mstar.groups()
         cmd = cmd.lower()
         if re.match("include$(?i)", cmd):
             includepath = os.path.join(os.path.split(survexfile.path)[0], re.sub(r"\.svx$", "", line))
-            includesurvexfile = models.SurvexFile(path=includepath, cave=survexfile.cave)
+            print('   - Include file found including - ' + includepath)
+            # Try to find the cave in the DB if not use the string as before
+            path_match = re.search(r"caves-(\d\d\d\d)/(\d+|\d\d\d\d-?\w+-\d+)/", includepath)
+            if path_match:
+                pos_cave = '%s-%s' % (path_match.group(1), path_match.group(2))
+                # print(pos_cave)
+                cave = models.getCaveByReference(pos_cave)
+                if cave:
+                    survexfile.cave = cave
+            else:
+                print('No match for %s' % includepath)
+            includesurvexfile = models.SurvexFile(path=includepath)
             includesurvexfile.save()
             includesurvexfile.SetDirectory()
             if includesurvexfile.exists():
+                survexblock.save()
                 fininclude = includesurvexfile.OpenFile()
                 RecursiveLoad(survexblock, includesurvexfile, fininclude, textlines)
-        
+
         elif re.match("begin$(?i)", cmd):
-            if line: 
+            if line:
+                newsvxpath = os.path.join(os.path.split(survexfile.path)[0], re.sub(r"\.svx$", "", line))
+                # Try to find the cave in the DB if not use the string as before
+                path_match = re.search(r"caves-(\d\d\d\d)/(\d+|\d\d\d\d-?\w+-\d+)/", newsvxpath)
+                if path_match:
+                    pos_cave = '%s-%s' % (path_match.group(1), path_match.group(2))
+                    print(pos_cave)
+                    cave = models.getCaveByReference(pos_cave)
+                    if cave:
+                        survexfile.cave = cave
+                else:
+                    print('No match for %s' % newsvxpath)
+
                 name = line.lower()
-                survexblockdown = models.SurvexBlock(name=name, begin_char=fin.tell(), parent=survexblock, survexpath=survexblock.survexpath+"."+name, cave=survexblock.cave, survexfile=survexfile, totalleglength=0.0)
+                print('   - Begin found for: ' + name)
+                # print('Block cave: ' + str(survexfile.cave))
+                survexblockdown = models.SurvexBlock(name=name, begin_char=fin.tell(), parent=survexblock, survexpath=survexblock.survexpath+"."+name, cave=survexfile.cave, survexfile=survexfile, totalleglength=0.0)
                 survexblockdown.save()
+                survexblock.save()
+                survexblock = survexblockdown
+                # print(survexblockdown)
                 textlinesdown = [ ]
                 RecursiveLoad(survexblockdown, survexfile, fin, textlinesdown)
             else:
                 iblankbegins += 1
-        
+
         elif re.match("end$(?i)", cmd):
             if iblankbegins:
                 iblankbegins -= 1
             else:
                 survexblock.text = "".join(textlines)
                 survexblock.save()
+                # print('   - End found: ')
+                endstamp = datetime.now()
+                timetaken = endstamp - stamp
+                # print('   - Time to process: ' + str(timetaken))
                 return
-        
+
         elif re.match("date$(?i)", cmd):
             if len(line) == 10:
-                survexblock.date = re.sub(r"\.", "-", line)
+                #print('   - Date found: ' + line)
+                survexblock.date = make_aware(datetime.strptime(re.sub(r"\.", "-", line), '%Y-%m-%d'), get_current_timezone())
                 expeditions = models.Expedition.objects.filter(year=line[:4])
                 if expeditions:
                     assert len(expeditions) == 1
                     survexblock.expedition = expeditions[0]
                     survexblock.expeditionday = survexblock.expedition.get_expedition_day(survexblock.date)
                     survexblock.save()
-                    
+
         elif re.match("team$(?i)", cmd):
-            mteammember = re.match(r"(Insts|Notes|Tape|Dog|Useless|Pics|Helper|Disto|Consultant)\s+(.*)$(?i)", line)
+            pass
+            # print('   - Team found: ')
+            mteammember = regex_team.match(line)
             if mteammember:
-                for tm in re.split(r" and | / |, | & | \+ |^both$|^none$(?i)", mteammember.group(2)):
+                for tm in regex_team_member.split(mteammember.group(2)):
                     if tm:
                         personexpedition = survexblock.expedition and GetPersonExpeditionNameLookup(survexblock.expedition).get(tm.lower())
                         if (personexpedition, tm) not in teammembers:
@@ -179,18 +304,23 @@ def RecursiveLoad(survexblock, survexfile, fin, textlines):
                             if personexpedition:
                                 personrole.person=personexpedition.person
                             personrole.save()
-                            
+
         elif cmd == "title":
-            survextitle = models.SurvexTitle(survexblock=survexblock, title=line.strip('"'), cave=survexblock.cave)
+            #print('   - Title found: ')
+            survextitle = models.SurvexTitle(survexblock=survexblock, title=line.strip('"'), cave=survexfile.cave)
             survextitle.save()
-            
+            pass
+
         elif cmd == "require":
             # should we check survex version available for processing?
             pass
 
         elif cmd == "data":
+            #print('   - Data found: ')
             ls = line.lower().split()
             stardata = { "type":ls[0] }
+            #print('    - Star data: ', stardata)
+            #print(ls)
             for i in range(0, len(ls)):
                 stardata[stardataparamconvert.get(ls[i], ls[i])] = i - 1
             if ls[0] in ["normal", "cartesian", "nosurvey"]:
@@ -199,40 +329,23 @@ def RecursiveLoad(survexblock, survexfile, fin, textlines):
                 stardata = stardatadefault
             else:
                 assert ls[0] == "passage", line
-        
+
         elif cmd == "equate":
+            #print('   - Equate found: ')
             LoadSurvexEquate(survexblock, line)
 
         elif cmd == "fix":
+            #print('   - Fix found: ')
             survexblock.MakeSurvexStation(line.split()[0])
 
         else:
+            #print('   - Stuff')
             if cmd not in ["sd", "include", "units", "entrance", "data", "flags", "title", "export", "instrument",
                            "calibrate", "set", "infer", "alias", "ref", "cs", "declination", "case"]:
                 print("Unrecognised command in line:", cmd, line, survexblock, survexblock.survexfile.path)
-
-
-def ReloadSurvexCave(survex_cave, area):
-    print(survex_cave, area)
-    cave = models.Cave.objects.get(kataster_number=survex_cave, area__short_name=area)
-    print(cave)
-    #cave = models.Cave.objects.get(kataster_number=survex_cave)
-    cave.survexblock_set.all().delete()
-    cave.survexfile_set.all().delete()
-    cave.survexdirectory_set.all().delete()
-    
-    survexfile = models.SurvexFile(path="caves-" + cave.kat_area() + "/" + survex_cave + "/" + survex_cave, cave=cave)
-    survexfile.save()
-    survexfile.SetDirectory()
-    
-    survexblockroot = models.SurvexBlock(name="root", survexpath="caves-" + cave.kat_area(), begin_char=0, cave=cave, survexfile=survexfile, totalleglength=0.0)
-    survexblockroot.save()
-    fin = survexfile.OpenFile()
-    textlines = [ ]
-    RecursiveLoad(survexblockroot, survexfile, fin, textlines)
-    survexblockroot.text = "".join(textlines)
-    survexblockroot.save()
-
+        endstamp = datetime.now()
+        timetaken = endstamp - stamp
+        # print('   - Time to process: ' + str(timetaken))
 
 def LoadAllSurvexBlocks():
 
@@ -258,22 +371,13 @@ def LoadAllSurvexBlocks():
     survexblockroot.save()
     fin = survexfile.OpenFile()
     textlines = [ ]
+    # The real work starts here
     RecursiveLoad(survexblockroot, survexfile, fin, textlines)
+    fin.close()
     survexblockroot.text = "".join(textlines)
     survexblockroot.save()
 
 
-    #Load each cave, 
-    #FIXME this should be dealt with load all above
-    print(" - Reloading all caves")
-    caves = models.Cave.objects.all()
-    for cave in caves:
-        if cave.kataster_number and os.path.isdir(os.path.join(settings.SURVEX_DATA, "caves-" + cave.kat_area(), cave.kataster_number)):
-            if cave.kataster_number not in ['40']:
-                print("loading", cave, cave.kat_area())
-                ReloadSurvexCave(cave.kataster_number, cave.kat_area())
-
-
 poslineregex = re.compile(r"^\(\s*([+-]?\d*\.\d*),\s*([+-]?\d*\.\d*),\s*([+-]?\d*\.\d*)\s*\)\s*([^\s]+)$")
 
 
@@ -286,7 +390,7 @@ def LoadPos():
     posfile = open("%s%s.pos" % (settings.SURVEX_DATA, settings.SURVEX_TOPNAME))
     posfile.readline() #Drop header
     for line in posfile.readlines():
-        r = poslineregex.match(line)                
+        r = poslineregex.match(line)
         if r:
             x, y, z, name = r.groups()
             try: