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_survex import SingleScan, ScansFolder, TunnelFile
from troggle.core.models import DataIssue


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 find_tunnel_scan(tunnelfile, path):
    '''Is given a line of text 'path' which may or may not contain a recognisable name of a scanned file
    which we have already seen when we imported all the files we could find in teh surveyscans direstories
    '''
    scansfolder, scansfile = None, None
    mscansdir = re.search(r"(\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:
                    message = "! More than one image filename matches filter query. [{}]: {} {} {} {}".format(scansfilel[0], mscansdir.group(1), mscansdir.group(2), tunnelfile.tunnelpath, path)
                    print(message)
                    DataIssue.objects.create(parser='Tunnel', message=message)
                scansfile = scansfilel[0]

        if scansfolder:
            tunnelfile.manyscansfolders.add(scansfolder)
        if scansfile:
            tunnelfile.scans.add(scansfile)
    
    elif path and not re.search(r"\.(?:png|jpg|pdf|jpeg|gif|pdf)$(?i)", path):
        name = os.path.split(path)[1]
        rtunnelfilel = TunnelFile.objects.filter(tunnelname=name)
        if len(rtunnelfilel):
            message = "! Two paths with same name [{}]: {}".format(path, name)
            print(message)
            DataIssue.objects.create(parser='Tunnel', message=message)
            rtunnelfile = rtunnelfilel[0]
            tunnelfile.tunnelcontains.add(rtunnelfile)

    tunnelfile.save()

def findimageinsert(therionfile, xth_me):
    '''Tries to link the drawing file (Therion format) to the referenced image (scan) file
    '''
    pass

def findimportinsert(therionfile, imp):
    '''Tries to link the scrap (Therion format) to the referenced therion scrap
    '''
    pass

rx_xth_me = re.compile(r'xth_me_image_insert.*{.*}$', re.MULTILINE)
rx_scrap  = re.compile(r'^survey (\w*).*$', re.MULTILINE)
rx_input  = re.compile(r'^input (\w*).*$', re.MULTILINE)

def settherionfileinfo(filetuple):
    '''Read in the drawing file contents and sets values on the tunnelfile object
    '''
    thtype, therionfile = filetuple
    
    ff = os.path.join(settings.TUNNEL_DATA, therionfile.tunnelpath)
    therionfile.filesize = os.stat(ff)[stat.ST_SIZE]
    if therionfile.filesize <= 0:
        message = "! Zero length therion file {}".format(ff)
        print(message)
        DataIssue.objects.create(parser='Therion', message=message)
        return
    fin = open(ff,'r')
    ttext = fin.read()
    fin.close()
    
    # The equivalent for a tunnel 'path' would be a .th2 'line wall'  or 'scrap'
    # print(len(re.findall(r"line", ttext)))
    if thtype=='th':
        therionfile.npaths = len(re.findall(r"^input ", ttext, re.MULTILINE))
    elif thtype=='th2':
        therionfile.npaths = len(re.findall(r"^line ", ttext, re.MULTILINE))
    therionfile.save()
  
    # scan and look for survex blocks that might have been included, and image scans (as for tunnel drawings)
    # which would populate tunnelfile.survexfile
    
    # in .th2 files:
    # ##XTHERION## xth_me_image_insert {500 1 1.0} {1700 {}} ../../../expofiles/surveyscans/2014/01popped_elev1.jpeg 0 {}
    # scrap blownout -projection plan -scale [-81.0 -42.0 216.0 -42.0 0.0 0.0 7.5438 0.0 m]
    
    for xth_me in rx_xth_me.findall(ttext):
        message = f'! Un-parsed image filename: {therionfile.tunnelname} : {xth_me.split()[-3]} - {therionfile.tunnelpath}'
        #print(message)
        DataIssue.objects.create(parser='Therion', message=message)
        findimageinsert(therionfile, xth_me)
        
    for inp in rx_input.findall(ttext):
        # if this 'input' is a .th2 file we have already seen, then we can assign this as a sub-file
        # but we would need to disentangle to get the current path properly
        message = f'! Un-set Therion .th2 input: - {therionfile.tunnelname} : {inp} - {therionfile.tunnelpath}'
        #print(message)
        DataIssue.objects.create(parser='Therion', message=message)
        findimportinsert(therionfile, inp)
    
    therionfile.save()
    
rx_skpath = re.compile(rb'<skpath')
rx_pcpath = re.compile(rb'<pcarea area_signal="frame".*?sfsketch="([^"]*)" sfstyle="([^"]*)"')

def settunnelfileinfo(tunnelfile):
    '''Read in the drawing file contents and sets values on the tunnelfile object
    Should try to read the date too e.g. tunneldate="2010-08-16 22:51:57
    then we could display on the master calendar per expo.
    '''
    ff = os.path.join(settings.TUNNEL_DATA, tunnelfile.tunnelpath)
    tunnelfile.filesize = os.stat(ff)[stat.ST_SIZE]
    if tunnelfile.filesize <= 0:
        message = "! Zero length xml file {}".format(ff)
        print(message)
        DataIssue.objects.create(parser='Drawings', message=message)
        return
    fin = open(ff,'rb')
    ttext = fin.read()
    fin.close()
   
    tunnelfile.npaths = len(rx_skpath.findall(ttext))
    tunnelfile.save()
    
    # example drawing file in Tunnel format.
    # <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 rx_pcpath.findall(ttext):
        find_tunnel_scan(tunnelfile, path.decode())
    
    # should also scan and look for survex blocks that might have been included, and image scans
    # which would populate tunnelfile.survexfile

    tunnelfile.save()


def load_drawings_files():
    '''Breadth first search of drawings directory looking for sub-directories and *.xml filesize
    '''
    all_xml = []
    drawdatadir = settings.TUNNEL_DATA
    TunnelFile.objects.all().delete()
    DataIssue.objects.filter(parser='Drawings').delete()
    DataIssue.objects.filter(parser='Therion').delete()
    DataIssue.objects.filter(parser='Tunnel').delete()

    drawingsdirs = [ "" ]
    while drawingsdirs:
        drawdir = drawingsdirs.pop()
        for f in os.listdir(os.path.join(drawdatadir, drawdir)):
            if f[0] == "." or f[-1] == "~":
                continue
            lf = os.path.join(drawdir, f)
            ff = os.path.join(drawdatadir, lf)
            if os.path.isdir(ff):
                drawingsdirs.append(lf) # lunatic! adding to list in middle of list while loop!
            elif f[-4:] == ".xml":
                # Always creates new
                tunnelfile = TunnelFile(tunnelpath=lf, tunnelname=os.path.split(f[:-4])[1])
                tunnelfile.save()
                all_xml.append(('xml',tunnelfile))
            elif f[-3:] == ".th":
                # Always creates new
                tunnelfile = TunnelFile(tunnelpath=lf, tunnelname=os.path.split(f[:-4])[1])
                tunnelfile.save()
                all_xml.append(('th',tunnelfile))
            elif f[-4:] == ".th2":
                # Always creates new
                tunnelfile = TunnelFile(tunnelpath=lf, tunnelname=os.path.split(f[:-4])[1])
                tunnelfile.save()
                all_xml.append(('th2',tunnelfile))

    print(f' - {len(all_xml)} Drawings files found')

    for d in all_xml:
        if d[0] == 'xml':
            settunnelfileinfo(d[1])
        # important to import .th2 files before .th so that we can assign them when found in .th files
        if d[0] == 'th2':
            settherionfileinfo(d)
        if d[0] == 'th':
            settherionfileinfo(d)
           
    # for drawfile in TunnelFile.objects.all():
        # SetTunnelfileInfo(drawfile)