troggle-unchained/parsers/surveys.py
2020-05-24 20:56:27 +01:00

335 lines
14 KiB
Python

from __future__ import (absolute_import, division,
print_function, unicode_literals)
import sys
import os
import types
import logging
import stat
import csv
import re
import datetime
from PIL import Image
from utils import save_carefully
from functools import reduce
import settings
from troggle.core.models import *
def get_or_create_placeholder(year):
""" All surveys must be related to a logbookentry. We don't have a way to
automatically figure out which survey went with which logbookentry,
so we create a survey placeholder logbook entry for each year. This
function always returns such a placeholder, and creates it if it doesn't
exist yet.
"""
lookupAttribs={'date__year':int(year), 'title':"placeholder for surveys",}
nonLookupAttribs={'text':"surveys temporarily attached to this should be re-attached to their actual trips", 'date':datetime.date(int(year),1,1)}
placeholder_logbook_entry, newly_created = save_carefully(LogbookEntry, lookupAttribs, nonLookupAttribs)
return placeholder_logbook_entry
# obsolete surveys.csv does not exist.
# def readSurveysFromCSV():
# try: # could probably combine these two
# surveytab = open(os.path.join(settings.SURVEY_SCANS, "Surveys.csv"))
# except IOError:
# import io, urllib.request, urllib.parse, urllib.error
# surveytab = io.StringIO(urllib.request.urlopen(settings.SURVEY_SCANS + "/Surveys.csv").read())
# dialect=csv.Sniffer().sniff(surveytab.read())
# surveytab.seek(0,0)
# surveyreader = csv.reader(surveytab,dialect=dialect)
# headers = next(surveyreader)
# header = dict(list(zip(headers, list(range(len(headers)))))) #set up a dictionary where the indexes are header names and the values are column numbers
# # test if the expeditions have been added yet
# if Expedition.objects.count()==0:
# print("There are no expeditions in the database. Please run the logbook parser.")
# sys.exit()
# logging.info("Deleting all scanned images")
# ScannedImage.objects.all().delete()
# logging.info("Deleting all survey objects")
# Survey.objects.all().delete()
# logging.info("Beginning to import surveys from "+str(os.path.join(settings.SURVEYS, "Surveys.csv"))+"\n"+"-"*60+"\n")
# for survey in surveyreader:
# #I hate this, but some surveys have a letter eg 2000#34a. The next line deals with that.
# walletNumberLetter = re.match(r'(?P<number>\d*)(?P<letter>[a-zA-Z]*)',survey[header['Survey Number']])
# # print(walletNumberLetter.groups())
# year=survey[header['Year']]
# surveyobj = Survey(
# expedition = Expedition.objects.filter(year=year)[0],
# wallet_number = walletNumberLetter.group('number'),
# logbook_entry = get_or_create_placeholder(year),
# comments = survey[header['Comments']],
# location = survey[header['Location']]
# )
# surveyobj.wallet_letter = walletNumberLetter.group('letter')
# if survey[header['Finished']]=='Yes':
# #try and find the sketch_scan
# pass
# surveyobj.save()
# logging.info("added survey " + survey[header['Year']] + "#" + surveyobj.wallet_number + "\r")
# dead
def listdir(*directories):
try:
return os.listdir(os.path.join(settings.SURVEYS, *directories))
except:
import urllib.request, urllib.parse, urllib.error
url = settings.SURVEYS + reduce(lambda x, y: x + "/" + y, ["listdir"] + list(directories))
folders = urllib.request.urlopen(url.replace("#", "%23")).readlines()
return [folder.rstrip(r"/") for folder in folders]
# add survey scans
# def parseSurveyScans(expedition, logfile=None):
# # yearFileList = listdir(expedition.year)
# try:
# yearPath=os.path.join(settings.SURVEY_SCANS, "surveyscans", expedition.year)
# yearFileList=os.listdir(yearPath)
# print(yearFileList)
# for surveyFolder in yearFileList:
# try:
# surveyNumber=re.match(rb'\d\d\d\d#(X?)0*(\d+)',surveyFolder).groups()
# #scanList = listdir(expedition.year, surveyFolder)
# scanList=os.listdir(os.path.join(yearPath,surveyFolder))
# except AttributeError:
# print(("Ignoring file in year folder: " + surveyFolder + "\r"))
# continue
# for scan in scanList:
# # Why does this insist on renaming all the scanned image files?
# # It produces duplicates names and all images have type .jpg in the scanObj.
# # It seems to rely on end users being particularly diligent in filenames which is NGtH
# try:
# #scanChopped=re.match(rb'(?i).*(notes|elev|plan|extend|elevation)-?(\d*)\.(png|jpg|jpeg|pdf)',scan).groups()
# scanChopped=re.match(rb'(?i)([a-z_-]*\d?[a-z_-]*)(\d*)\.(png|jpg|jpeg|pdf|top|dxf|svg|tdr|th2|xml|txt)',scan).groups()
# scanType,scanNumber,scanFormat=scanChopped
# except AttributeError:
# print(("Ignored (bad name format): " + surveyFolder + '/' + scan + "\r"))
# continue
# scanTest = scanType
# scanType = 'notes'
# match = re.search(rb'(?i)(elev|extend)',scanTest)
# if match:
# scanType = 'elevation'
# match = re.search(rb'(?i)(plan)',scanTest)
# if match:
# scanType = 'plan'
# if scanNumber=='':
# scanNumber=1
# if isinstance(surveyNumber, tuple):
# surveyLetter=surveyNumber[0]
# surveyNumber=surveyNumber[1]
# try:
# placeholder=get_or_create_placeholder(year=int(expedition.year))
# survey=Survey.objects.get_or_create(wallet_number=surveyNumber, wallet_letter=surveyLetter, expedition=expedition, defaults={'logbook_entry':placeholder})[0]
# except Survey.MultipleObjectsReturned:
# survey=Survey.objects.filter(wallet_number=surveyNumber, wallet_letter=surveyLetter, expedition=expedition)[0]
# file_=os.path.join(yearPath, surveyFolder, scan)
# scanObj = ScannedImage(
# file=file_,
# contents=scanType,
# number_in_wallet=scanNumber,
# survey=survey,
# new_since_parsing=False,
# )
# print(("Added scanned image at " + str(scanObj)))
# #if scanFormat=="png":
# #if isInterlacedPNG(os.path.join(settings.SURVEY_SCANS, "surveyscans", file_)):
# # print file_+ " is an interlaced PNG. No can do."
# #continue
# scanObj.save()
# except (IOError, OSError):
# yearPath=os.path.join(settings.SURVEY_SCANS, "surveyscans", expedition.year)
# print((" ! No folder found for " + expedition.year + " at:- " + yearPath))
# dead
# def parseSurveys(logfile=None):
# try:
# readSurveysFromCSV()
# except (IOError, OSError):
# print(" ! Survey CSV not found..")
# pass
# print(" - Loading scans by expedition year")
# for expedition in Expedition.objects.filter(year__gte=2000): #expos since 2000, because paths and filenames were nonstandard before then
# print("%s" % expedition, end=' ')
# parseSurveyScans(expedition)
# dead
# def isInterlacedPNG(filePath): #We need to check for interlaced PNGs because the thumbnail engine can't handle them (uses PIL)
# file=Image.open(filePath)
# print(filePath)
# if 'interlace' in file.info:
# return file.info['interlace']
# else:
# return False
# handles url or file, so we can refer to a set of scans on another server
def GetListDir(sdir):
res = [ ]
if sdir[:7] == "http://":
assert False, "Not written"
s = urllib.request.urlopen(sdir)
else:
for f in os.listdir(sdir):
if f[0] != ".":
ff = os.path.join(sdir, f)
res.append((f, ff, os.path.isdir(ff)))
return res
def LoadListScansFile(survexscansfolder):
gld = [ ]
# flatten out any directories in these wallet folders - should not be any
for (fyf, ffyf, fisdiryf) in GetListDir(survexscansfolder.fpath):
if fisdiryf:
gld.extend(GetListDir(ffyf))
else:
gld.append((fyf, ffyf, fisdiryf))
for (fyf, ffyf, fisdiryf) in gld:
#assert not fisdiryf, ffyf
if re.search(r"\.(?:png|jpg|jpeg|pdf|jpeg|svg)(?i)$", fyf):
survexscansingle = SurvexScanSingle(ffile=ffyf, name=fyf, survexscansfolder=survexscansfolder)
survexscansingle.save()
# this iterates through the scans directories (either here or on the remote server)
# and builds up the models we can access later
def LoadListScans():
print(' - Loading Survey Scans... (deleting all objects first)')
SurvexScanSingle.objects.all().delete()
SurvexScansFolder.objects.all().delete()
# first do the smkhs (large kh survey scans) directory
survexscansfoldersmkhs = SurvexScansFolder(fpath=os.path.join(settings.SURVEY_SCANS, "smkhs"), walletname="smkhs")
if os.path.isdir(survexscansfoldersmkhs.fpath):
survexscansfoldersmkhs.save()
LoadListScansFile(survexscansfoldersmkhs)
# iterate into the surveyscans directory
print(' - ', end=' ')
for f, ff, fisdir in GetListDir(os.path.join(settings.SURVEY_SCANS, "surveyscans")):
if not fisdir:
continue
# do the year folders
if re.match(r"\d\d\d\d$", f):
print("%s" % f, end=' ')
for fy, ffy, fisdiry in GetListDir(ff):
if fisdiry:
assert fisdiry, ffy
survexscansfolder = SurvexScansFolder(fpath=ffy, walletname=fy)
survexscansfolder.save()
LoadListScansFile(survexscansfolder)
# do the
elif f != "thumbs":
survexscansfolder = SurvexScansFolder(fpath=ff, walletname=f)
survexscansfolder.save()
LoadListScansFile(survexscansfolder)
def FindTunnelScan(tunnelfile, path):
scansfolder, scansfile = None, None
mscansdir = re.search(rb"(\d\d\d\d#X?\d+\w?|1995-96kh|92-94Surveybookkh|1991surveybook|smkhs)/(.*?(?:png|jpg|pdf|jpeg))$", path)
if mscansdir:
scansfolderl = SurvexScansFolder.objects.filter(walletname=mscansdir.group(1))
if len(scansfolderl):
assert len(scansfolderl) == 1
scansfolder = scansfolderl[0]
if scansfolder:
scansfilel = scansfolder.survexscansingle_set.filter(name=mscansdir.group(2))
if len(scansfilel):
if len(scansfilel) > 1:
print("BORK more than one image filename matches filter query. ", scansfilel[0])
print("BORK ", tunnelfile.tunnelpath, path)
print("BORK ", mscansdir.group(1), mscansdir.group(2), len(scansfilel))
#assert len(scansfilel) == 1
scansfile = scansfilel[0]
if scansfolder:
tunnelfile.survexscansfolders.add(scansfolder)
if scansfile:
tunnelfile.survexscans.add(scansfile)
elif path and not re.search(rb"\.(?:png|jpg|pdf|jpeg)$(?i)", path):
name = os.path.split(path)[1]
#print("debug-tunnelfileobjects ", tunnelfile.tunnelpath, path, name)
rtunnelfilel = TunnelFile.objects.filter(tunnelname=name)
if len(rtunnelfilel):
assert len(rtunnelfilel) == 1, ("two paths with name of", path, "need more discrimination coded")
rtunnelfile = rtunnelfilel[0]
#print "ttt", tunnelfile.tunnelpath, path, name, rtunnelfile.tunnelpath
tunnelfile.tunnelcontains.add(rtunnelfile)
tunnelfile.save()
def SetTunnelfileInfo(tunnelfile):
ff = os.path.join(settings.TUNNEL_DATA, tunnelfile.tunnelpath)
tunnelfile.filesize = os.stat(ff)[stat.ST_SIZE]
fin = open(ff,'rb')
ttext = fin.read()
fin.close()
if tunnelfile.filesize <= 0:
print("DEBUG - zero length xml file", ff)
return
mtype = re.search(rb"<(fontcolours|sketch)", ttext)
assert mtype, ff
tunnelfile.bfontcolours = (mtype.group(1)=="fontcolours")
tunnelfile.npaths = len(re.findall(rb"<skpath", ttext))
tunnelfile.save()
# <tunnelxml tunnelversion="version2009-06-21 Matienzo" tunnelproject="ireby" tunneluser="goatchurch" tunneldate="2009-06-29 23:22:17">
# <pcarea area_signal="frame" sfscaledown="12.282584" sfrotatedeg="-90.76982" sfxtrans="11.676667377221136" sfytrans="-15.677173422877454" sfsketch="204description/scans/plan(38).png" sfstyle="" nodeconnzsetrelative="0.0">
for path, style in re.findall(rb'<pcarea area_signal="frame".*?sfsketch="([^"]*)" sfstyle="([^"]*)"', ttext):
FindTunnelScan(tunnelfile, path)
# should also scan and look for survex blocks that might have been included
# and also survex titles as well.
tunnelfile.save()
def LoadTunnelFiles():
tunneldatadir = settings.TUNNEL_DATA
TunnelFile.objects.all().delete()
tunneldirs = [ "" ]
while tunneldirs:
tunneldir = tunneldirs.pop()
for f in os.listdir(os.path.join(tunneldatadir, tunneldir)):
if f[0] == "." or f[-1] == "~":
continue
lf = os.path.join(tunneldir, f)
ff = os.path.join(tunneldatadir, lf)
if os.path.isdir(ff):
tunneldirs.append(lf)
elif f[-4:] == ".xml":
tunnelfile = TunnelFile(tunnelpath=lf, tunnelname=os.path.split(f[:-4])[1])
tunnelfile.save()
for tunnelfile in TunnelFile.objects.all():
SetTunnelfileInfo(tunnelfile)