2
0
mirror of https://expo.survex.com/repositories/troggle/.git synced 2025-04-27 05:01:45 +01:00

start refactor survex import

This commit is contained in:
Philip Sargent 2020-06-24 14:10:13 +01:00
parent dc5a53376d
commit bb69cc073a
4 changed files with 415 additions and 419 deletions

@ -3,6 +3,7 @@ import os
import datetime
import logging
import re
import resource
from subprocess import call
from urllib.parse import urljoin
@ -22,6 +23,11 @@ from django.template import Context, loader
import troggle.core.models_survex
def get_process_memory():
usage=resource.getrusage(resource.RUSAGE_SELF)
return usage[2]/1024.0
def get_related_by_wikilinks(wiki_text):
found=re.findall(settings.QM_PATTERN,wiki_text)
res=[]

@ -81,13 +81,6 @@ class SurvexStation(models.Model):
else:
return r
class SurvexLeg():
"""No longer a models.Model subclass, so no longer a database table
"""
tape = 0.0
compass = 0.0
clino = 0.0
#
# Single SurvexBlock
#

@ -12,7 +12,7 @@ from django.views.generic.list import ListView
from troggle.core.models import Expedition, Person, PersonExpedition
from troggle.core.models_caves import Cave, LogbookEntry
from troggle.core.models_survex import SurvexLeg, SurvexBlock
from troggle.core.models_survex import SurvexBlock
import troggle.settings as settings
@ -106,7 +106,6 @@ def stats(request):
legsbyexpo.append((expedition, {"nsurvexlegs": "{:,}".format(legsyear),
"survexleglength":"{:,.0f}".format(survexleglength)}))
legsbyexpo.reverse()
#survexlegs = SurvexLeg.objects.all()
renderDict = {**statsDict, **{ "nsurvexlegs": "{:,}".format(nimportlegs), "totalsurvexlength":totalsurvexlength/1000, "addupsurvexlength":addupsurvexlength/1000, "legsbyexpo":legsbyexpo }} # new syntax
return render(request,'statistics.html', renderDict)

@ -2,7 +2,6 @@ import sys
import os
import re
import time
import resource
from datetime import datetime, timedelta
from subprocess import call, Popen, PIPE
@ -17,35 +16,71 @@ import troggle.core.models_survex as models_survex
from troggle.parsers.people import GetPersonExpeditionNameLookup
from troggle.core.views_caves import MapLocations
"""A 'survex block' is a *begin...*end set of cave data.
A 'scansfolder' is what we today call a "survey scans folder" or a "wallet".
"""
rx_braskets= re.compile(r"[()]")
rx_line_length = re.compile(r"[\d\-+.]+$")
survexlegsalllength = 0.0
survexlegsnumber = 0
survexblockroot = None
ROOTBLOCK = "rootblock"
class SurvexLeg():
"""No longer a models.Model subclass, so no longer a database table
"""
tape = 0.0
compass = 0.0
clino = 0.0
def LoadSurvexLineLeg(survexblock, stardata, sline, comment, cave):
class LoadSurvex():
"""A 'survex block' is a *begin...*end set of cave data.
A survex file can contain many begin-end blocks, which can be nested, and which can *include
other survex files.
A 'scansfolder' is what we today call a "survey scans folder" or a "wallet".
"""
# This interprets the survex "*data normal" command which sets out the order of the fields in the data, e.g.
# *DATA normal from to length gradient bearing ignore ignore ignore ignore
stardatadefault = {"type":"normal", "t":"leg", "from":0, "to":1, "tape":2, "compass":3, "clino":4}
stardataparamconvert = {"length":"tape", "bearing":"compass", "gradient":"clino"}
rx_braskets= re.compile(r"[()]")
rx_linelen = re.compile(r"[\d\-+.]+$")
rx_team = re.compile(r"(?i)(Insts|Notes|Tape|Dog|Useless|Pics|Helper|Disto|Consultant)\s+(.*)$")
rx_person = re.compile(r"(?i) and | / |, | & | \+ |^both$|^none$")
rx_qm = re.compile(r'^\s*QM(\d)\s+?([a-dA-DxX])\s+([\w\-]+)\.(\d+)\s+(([\w\-]+)\.(\d+)|\-)\s+(.+)$')
# remember there is also QM_PATTERN used in views_other and set in settings.py
rx_comment = re.compile(r"([^;]*?)\s*(?:;\s*(.*))?\n?$")
rx_ref = re.compile(r'.*?ref.*?(\d+)\s*#\s*(X)?\s*(\d+)')
rx_star = re.compile(r'\s*\*[\s,]*(\w+)\s*(.*?)\s*(?:;.*)?$')
# years from 1960 to 2039
rx_starref = re.compile(r'(?i)^\s*\*ref[\s.:]*((?:19[6789]\d)|(?:20[0123]\d))\s*#?\s*(X)?\s*(.*?\d+.*?)$')
# rx_starref = re.compile("""?x # VERBOSE mode - can't get this to work
# ^\s*\*ref # look for *ref at start of line
# [\s.:]* # some spaces, stops or colons
# ((?:19[6789]\d)|(?:20[0123]\d)) # a date from 1960 to 2039 - captured as one field
# \s*# # spaces then hash separator
# ?\s*(X) # optional X - captured
# ?\s*(.*?\d+.*?) # maybe a space, then at least one digit in the string - captured
# $(?i)""", re.X) # the end (do the whole thing case insensitively)
survexlegsalllength = 0.0
survexlegsnumber = 0
insp = ""
callcount = 0
def __init__(self):
pass
def LoadSurvexLineLeg(self,survexblock, stardata, sline, comment):
"""This reads compass, clino and tape data but only keeps the tape lengths,
the rest is discarded after error-checking.
"""
global survexlegsalllength
global survexlegsnumber
ls = sline.lower().split()
survexleg = models_survex.SurvexLeg()
survexleg = SurvexLeg()
# this next fails for two surface survey svx files which use / for decimal point
# e.g. '29/09' in the tape measurement, or use decimals but in brackets, e.g. (06.05)
if stardata["type"] == "normal":
tape = rx_braskets.sub("",ls[stardata["tape"]])
tape = self.rx_braskets.sub("",ls[stardata["tape"]])
tape = tape.replace("/",".")
try:
survexleg.tape = float(tape)
survexlegsnumber += 1
self.survexlegsnumber += 1
except ValueError:
print(("! Tape misread in", survexblock.survexfile.path))
print((" Stardata:", stardata))
@ -89,14 +124,11 @@ def LoadSurvexLineLeg(survexblock, stardata, sline, comment, cave):
survexleg.compass = 1000
survexleg.clino = -90.0
else:
assert rx_line_length.match(lcompass), ls
assert rx_line_length.match(lclino) and lclino != "-", ls
assert self.rx_linelen.match(lcompass), ls
assert self.rx_linelen.match(lclino) and lclino != "-", ls
survexleg.compass = float(lcompass)
survexleg.clino = float(lclino)
if cave:
survexleg.cave = cave
# No need to save as we are measuring lengths only on parsing now.
# delete the object so that django autosaving doesn't save it.
survexleg = None
@ -105,50 +137,17 @@ def LoadSurvexLineLeg(survexblock, stardata, sline, comment, cave):
if itape:
try:
survexblock.totalleglength += float(ls[itape])
survexlegsalllength += float(ls[itape])
self.survexlegsalllength += float(ls[itape])
except ValueError:
print("! Length not added")
# No need to save as we are measuring lengths only on parsing now.
# def LoadSurvexEquate(survexblock, sline):
# #print sline #
# stations = sline.split()
# assert len(stations) > 1
# for station in stations:
# survexblock.MakeSurvexStation(station)
def LoadSurvexLinePassage(survexblock, stardata, sline, comment):
def LoadSurvexLinePassage(self,survexblock, stardata, sline, comment):
# do not import this: *data passage.. data which is LRUD not tape/compass/clino
pass
# This interprets the survex "*data normal" command which sets out the order of the fields in the data, e.g.
# *DATA normal from to length gradient bearing ignore ignore ignore ignore
stardatadefault = {"type":"normal", "t":"leg", "from":0, "to":1, "tape":2, "compass":3, "clino":4}
stardataparamconvert = {"length":"tape", "bearing":"compass", "gradient":"clino"}
rx_comment = re.compile(r"([^;]*?)\s*(?:;\s*(.*))?\n?$")
rx_ref = re.compile(r'.*?ref.*?(\d+)\s*#\s*(X)?\s*(\d+)')
rx_star = re.compile(r'\s*\*[\s,]*(\w+)\s*(.*?)\s*(?:;.*)?$')
# years from 1960 to 2039
rx_starref = re.compile(r'(?i)^\s*\*ref[\s.:]*((?:19[6789]\d)|(?:20[0123]\d))\s*#?\s*(X)?\s*(.*?\d+.*?)$')
# rx_starref = re.compile("""?x # VERBOSE mode - can't get this to work
# ^\s*\*ref # look for *ref at start of line
# [\s.:]* # some spaces, stops or colons
# ((?:19[6789]\d)|(?:20[0123]\d)) # a date from 1960 to 2039 - captured as one field
# \s*# # spaces then hash separator
# ?\s*(X) # optional X - captured
# ?\s*(.*?\d+.*?) # maybe a space, then at least one digit in the string - captured
# $(?i)""", re.X) # the end (do the whole thing case insensitively)
rx_team = re.compile(r"(?i)(Insts|Notes|Tape|Dog|Useless|Pics|Helper|Disto|Consultant)\s+(.*)$")
rx_person = re.compile(r"(?i) and | / |, | & | \+ |^both$|^none$")
rx_qm = re.compile(r'^\s*QM(\d)\s+?([a-dA-DxX])\s+([\w\-]+)\.(\d+)\s+(([\w\-]+)\.(\d+)|\-)\s+(.+)$')
insp = ""
callcount = 0
def RecursiveLoad(survexblock, survexfile, fin):
def RecursiveLoad(self,survexblock, survexfile, fin):
"""Follows the *include links in all the survex files from the root file 1623.svx
and reads in the survex blocks, other data and the wallet references (scansfolder) as it
goes. This part of the data import process is where the maximum memory is used and where it
@ -156,20 +155,19 @@ def RecursiveLoad(survexblock, survexfile, fin):
"""
iblankbegins = 0
text = [ ]
stardata = stardatadefault
stardata = self.stardatadefault
teammembers = [ ]
global insp
global callcount
global survexlegsnumber
insp =self.insp
blocklegs = self.survexlegsnumber
print(insp+" - MEM:{:.3f} Reading. parent:{} <> {} ".format(get_process_memory(),survexblock.survexfile.path,survexfile.path))
print(insp+" - MEM:{:.3f} Reading. parent:{} <> {} ".format(models.get_process_memory(),survexblock.survexfile.path,survexfile.path))
stamp = datetime.now()
lineno = 0
sys.stderr.flush();
callcount +=1
if callcount >=10:
callcount=0
self.callcount +=1
if self.callcount >=10:
self.callcount=0
print(".", file=sys.stderr,end='')
# Try to find the cave in the DB if not use the string as before
@ -186,9 +184,9 @@ def RecursiveLoad(survexblock, survexfile, fin):
for svxline in svxlines:
lineno += 1
# break the line at the comment
sline, comment = rx_comment.match(svxline.strip()).groups()
sline, comment = self.rx_comment.match(svxline.strip()).groups()
# detect ref line pointing to the scans directory
mref = comment and rx_ref.match(comment)
mref = comment and self.rx_ref.match(comment)
if mref:
yr, letterx, wallet = mref.groups()
if not letterx:
@ -209,11 +207,11 @@ def RecursiveLoad(survexblock, survexfile, fin):
# 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 rx_qm
# needs to match self.rx_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 rx_qm.match(comment)
qmline = comment and self.rx_qm.match(comment)
if qmline:
qm_no = qmline.group(1)
qm_grade = qmline.group(2)
@ -252,7 +250,7 @@ def RecursiveLoad(survexblock, survexfile, fin):
continue
# detect the star ref command
mstar = rx_starref.match(sline)
mstar = self.rx_starref.match(sline)
if mstar:
yr,letterx,wallet = mstar.groups()
if not letterx:
@ -275,13 +273,14 @@ def RecursiveLoad(survexblock, survexfile, fin):
continue
# detect the star command
mstar = rx_star.match(sline)
mstar = self.rx_star.match(sline)
if not mstar:
if "from" in stardata:
LoadSurvexLineLeg(survexblock, stardata, sline, comment, survexfile.cave)
self.LoadSurvexLineLeg(survexblock, stardata, sline, comment)
pass
elif stardata["type"] == "passage":
LoadSurvexLinePassage(survexblock, stardata, sline, comment)
pass
#self.LoadSurvexLinePassage(survexblock, stardata, sline, comment)
#Missing "station" in stardata.
continue
@ -308,11 +307,14 @@ def RecursiveLoad(survexblock, survexfile, fin):
if includesurvexfile.exists():
survexblock.save()
fininclude = includesurvexfile.OpenFile()
insp += "> "
RecursiveLoad(survexblock, includesurvexfile, fininclude)
self.survexlegsnumber = blocklegs
self.insp += "> "
self.RecursiveLoad(survexblock, includesurvexfile, fininclude)
#--------------------------------------------------------
fininclude.close()
insp = insp[2:]
self.insp = self.insp[2:]
insp = self.insp
blocklegs = self.survexlegsnumber
else:
print((insp+' ! ERROR *include file not found for %s' % includesurvexfile))
@ -333,7 +335,7 @@ def RecursiveLoad(survexblock, survexfile, fin):
else:
print((insp+' - No match (b) for %s' % newsvxpath))
previousnlegs = survexlegsnumber
previousnlegs = blocklegs
name = line.lower()
print(insp+' - Begin found for:{}, creating new SurvexBlock '.format(name))
survexblockdown = models_survex.SurvexBlock(name=name, parent=survexblock,
@ -344,12 +346,15 @@ def RecursiveLoad(survexblock, survexfile, fin):
survexblock.save()
survexblock = survexblockdown
print(insp+" - ENTERING nested *begin/*end block: {}".format(name))
insp += "> "
RecursiveLoad(survexblockdown, survexfile, fin)
self.survexlegsnumber = blocklegs
self.insp += "> "
self.RecursiveLoad(survexblockdown, survexfile, fin)
#--------------------------------------------------------
# do not close the file as there may be more blocks in this one
# and it is re-read afresh with every nested begin-end block.
insp = insp[2:]
self.insp = self.insp[2:]
insp = self.insp
blocklegs = self.survexlegsnumber
else:
iblankbegins += 1
@ -358,8 +363,8 @@ def RecursiveLoad(survexblock, survexfile, fin):
print(insp+" - RETURNING from nested *begin/*end block: {}".format(line))
iblankbegins -= 1
else:
legsinblock = survexlegsnumber - previousnlegs
print(insp+" - LEGS: {} (previous: {}, now:{})".format(legsinblock,previousnlegs,survexlegsnumber))
legsinblock = self.survexlegsnumber - previousnlegs
print(insp+" - LEGS: {} (previous: {}, now:{})".format(legsinblock,previousnlegs,self.survexlegsnumber))
survexblock.legsall = legsinblock
survexblock.save()
endstamp = datetime.now()
@ -379,9 +384,9 @@ def RecursiveLoad(survexblock, survexfile, fin):
elif re.match("team$(?i)", cmd):
pass
# print(insp+' - Team found: ')
mteammember = rx_team.match(line)
mteammember = self.rx_team.match(line)
if mteammember:
for tm in rx_person.split(mteammember.group(2)):
for tm in self.rx_person.split(mteammember.group(2)):
if tm:
personexpedition = survexblock.expedition and GetPersonExpeditionNameLookup(survexblock.expedition).get(tm.lower())
if (personexpedition, tm) not in teammembers:
@ -393,9 +398,7 @@ def RecursiveLoad(survexblock, survexfile, fin):
personrole.save()
elif cmd == "title":
# unused in troggle.
#survextitle = models_survex.SurvexTitle(survexblock=survexblock, title=line.strip('"'), cave=survexfile.cave)
#survextitle.save()
# unused in troggle today - but will become text list on SurvexBlock
pass
elif cmd == "require":
@ -406,11 +409,11 @@ def RecursiveLoad(survexblock, survexfile, fin):
ls = line.lower().split()
stardata = { "type":ls[0] }
for i in range(0, len(ls)):
stardata[stardataparamconvert.get(ls[i], ls[i])] = i - 1
stardata[self.stardataparamconvert.get(ls[i], ls[i])] = i - 1
if ls[0] in ["normal", "cartesian", "nosurvey"]:
assert (("from" in stardata and "to" in stardata) or "station" in stardata), line
elif ls[0] == "default":
stardata = stardatadefault
stardata = self.stardatadefault
else:
assert ls[0] == "passage", line
@ -425,8 +428,7 @@ def RecursiveLoad(survexblock, survexfile, fin):
# but this data is only used for sense-checking not to actually calculate anything important
pass
elif cmd == "fix":
# troggle does not use survex stations
#survexblock.MakeSurvexStation(line.split()[0])
# troggle does not use survex stations except for entrances which are loaded elsewhere
pass
elif cmd in ["alias", "calibrate", "cs","entrance", "export", "case",
"declination", "infer","instrument", "sd", "units"]:
@ -442,46 +444,39 @@ def RecursiveLoad(survexblock, survexfile, fin):
print((insp+message))
models.DataIssue.objects.create(parser='survex', message=message)
endstamp = datetime.now()
timetaken = endstamp - stamp
# print(insp+' - Time to process: ' + str(timetaken))
def get_process_memory():
usage=resource.getrusage(resource.RUSAGE_SELF)
return usage[2]/1024.0
def FindAndLoadAllSurvex(survexblockroot, survexfileroot):
"""Follows the *include links recursively to find files
"""
print(' - redirecting stdout to loadsurvexblks.log...')
print(' - redirecting stdout to svxblks.log...')
stdout_orig = sys.stdout
# Redirect sys.stdout to the file
sys.stdout = open('loadsurvexblks.log', 'w')
sys.stdout = open('svxblks.log', 'w')
svxl = LoadSurvex()
finroot = survexfileroot.OpenFile()
RecursiveLoad(survexblockroot, survexfileroot, finroot)
svxl.RecursiveLoad(survexblockroot, survexfileroot, finroot)
finroot.close()
# Close the logging file
survexlegsnumber = svxl.survexlegsnumber
survexlegsalllength = svxl.survexlegsalllength
# Close the logging file, Restore sys.stdout to our old saved file handle
sys.stdout.close()
print("+", file=sys.stderr)
sys.stderr.flush();
# Restore sys.stdout to our old saved file handler
sys.stdout = stdout_orig
return (survexlegsnumber, survexlegsalllength)
def LoadAllSurvexBlocks():
global survexlegsalllength
global survexlegsnumber
print(' - Flushing All Survex Blocks...')
models_survex.SurvexBlock.objects.all().delete()
models_survex.SurvexFile.objects.all().delete()
models_survex.SurvexDirectory.objects.all().delete()
# models_survex.SurvexEquate.objects.all().delete()
# models_survex.SurvexTitle.objects.all().delete()
models_survex.SurvexPersonRole.objects.all().delete()
models_survex.SurvexStation.objects.all().delete()
@ -497,7 +492,10 @@ def LoadAllSurvexBlocks():
survexblockroot.save()
print(' - Loading All Survex Blocks...')
FindAndLoadAllSurvex(survexblockroot, survexfileroot)
memstart = models.get_process_memory()
survexlegsnumber, survexlegsalllength = FindAndLoadAllSurvex(survexblockroot, survexfileroot)
memend = models.get_process_memory()
print(" - MEMORY start:{:.3f} MB end:{:.3f} MB increase={:.3f} MB",format(memstart,memend,memend-memstart))
survexblockroot.totalleglength = survexlegsalllength
survexblockroot.legsall = survexlegsnumber