troggle/parsers/surveys.py

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 *
#from troggle.core.models_caves import *
from troggle.core.models_survex import SingleScan, ScansFolder, TunnelFile

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

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]


# 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(scansfolder):
    gld = [ ]
    # flatten out any directories in these wallet folders - should not be any
    for (fyf, ffyf, fisdiryf) in GetListDir(scansfolder.fpath):
        if fisdiryf:
            gld.extend(GetListDir(ffyf))
        else:
            gld.append((fyf, ffyf, fisdiryf))
    
    c=0
    for (fyf, ffyf, fisdiryf) in gld:
        #assert not fisdiryf, ffyf
        if re.search(r"\.(?:png|jpg|jpeg|pdf|svg|gif)(?i)$", fyf):
            singlescan = SingleScan(ffile=ffyf, name=fyf, scansfolder=scansfolder)
            singlescan.save()
            c+=1
            if c>=10:
                print(".", end='')
                c = 0

        
# 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')

    SingleScan.objects.all().delete()
    ScansFolder.objects.all().delete()
    print(' - deleting all scansFolder and scansSingle objects')

    # first do the smkhs (large kh survey scans) directory
    manyscansfoldersmkhs = ScansFolder(fpath=os.path.join(settings.SURVEY_SCANS, "../surveys/smkhs"), walletname="smkhs") 
    print("smkhs", end=' ')
    if os.path.isdir(manyscansfoldersmkhs.fpath):
        manyscansfoldersmkhs.save()
        LoadListScansFile(manyscansfoldersmkhs)
        
    
    # iterate into the surveyscans directory
    print(' - ', end=' ')
    for f, ff, fisdir in GetListDir(settings.SURVEY_SCANS):
        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
                    scansfolder = ScansFolder(fpath=ffy, walletname=fy)
                    scansfolder.save()
                    LoadListScansFile(scansfolder)
        
        # do the 
        elif f != "thumbs":
            scansfolder = ScansFolder(fpath=ff, walletname=f)
            scansfolder.save()
            LoadListScansFile(scansfolder)
            

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 = ScansFolder.objects.filter(walletname=mscansdir.group(1))
        if len(scansfolderl):
            assert len(scansfolderl) == 1
            scansfolder = scansfolderl[0]
        if scansfolder:
            scansfilel = scansfolder.singlescan_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.manyscansfolders.add(scansfolder)
        if scansfile:
            tunnelfile.scans.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)