troggle-unchained/parsers/scans.py

186 lines
7.3 KiB
Python
Raw Normal View History

2021-05-03 20:36:29 +01:00
import sys
import os
import subprocess
2021-05-03 20:36:29 +01:00
import types
import stat
import csv
import re
import datetime
import shutil, filecmp
2021-05-03 20:36:29 +01:00
from functools import reduce
from pathlib import Path
2021-05-03 20:36:29 +01:00
import settings
from troggle.core.models.survex import SingleScan, Wallet, DrawingFile
from troggle.core.models.troggle import DataIssue
from troggle.core.utils import save_carefully
2022-08-01 15:32:35 +01:00
from troggle.core.views.scans import datewallet
2021-05-03 20:36:29 +01:00
'''Searches through all the survey scans directories (wallets) in expofiles, looking for images to be referenced.
'''
contentsjson = "contents.json"
git = settings.GIT
# to do: Actually read all the JSON files and set the survex file field appropriately!
# def GetListDir(sdir):
# '''handles url or file, so we can refer to a set of scans (not drawings) on another server
# returns a list of f (file), ff (file full path), is_dir (bool)
# REPLACE all use of this with Path.rglob() !
# '''
# res = [ ]
# if type(sdir) is str and sdir[:7] == "http://":
# # s = urllib.request.urlopen(sdir)
# message = f"! Requesting loading from http:// NOT IMPLEMENTED. [{sdir}]"
# print(message)
# DataIssue.objects.create(parser='Drawings', message=message)
# sdir[:7] = ""
2021-05-03 20:36:29 +01:00
# 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(wallet):
# # formerly a generic troggle utility, written by who ? Being gradually expunged and replaced by python standard library functions
# gld = [ ]
# # flatten out any directories in these wallet folders - should not be any
# for (fyf, ffyf, fisdiryf) in GetListDir(wallet.fpath):
# if fisdiryf:
# gld.extend(GetListDir(ffyf))
# else:
# gld.append((fyf, ffyf, fisdiryf))
2021-05-03 20:36:29 +01:00
# c=0
# for (fyf, ffyf, fisdiryf) in gld:
# if re.search(r"\.(?:png|jpg|jpeg|pdf|svg|gif|xvi)(?i)$", fyf):
# singlescan = SingleScan(ffile=ffyf, name=fyf, wallet=wallet)
# singlescan.save()
# c+=1
# if c>=10:
# print(".", end='')
# c = 0
2021-05-03 20:36:29 +01:00
def load_all_scans():
'''This iterates through the scans directories (either here or on the remote server)
and builds up the models we can access later.
It does NOT read or validate anything in the JSON data attached to each wallet. Those checks
are done at runtime, when a wallet is accessed, not at import time.
'''
2021-05-03 20:36:29 +01:00
print(' - Loading Survey Scans')
SingleScan.objects.all().delete()
Wallet.objects.all().delete()
print(' - deleting all Wallet and SingleScan objects')
2022-03-15 20:53:55 +00:00
DataIssue.objects.filter(parser='scans').delete()
# These are valid old file types to be visible, they are not necessarily allowed to be uploaded to a new wallet.
valids = [".top",".txt",".tif",".png",".jpg",".jpeg",".pdf",".svg",".gif",".xvi",
".json",".autosave",".sxd",".svx",".th",".th2",".tdr",".sql",".zip",".dxf",".3d",
".ods",".csv",".xcf",".xml"]
validnames = ["thconfig","manifest"]
2022-07-22 09:23:00 +01:00
2021-05-03 20:36:29 +01:00
# iterate into the surveyscans directory
# Not all folders with files in them are wallets.
# they are if they are /2010/2010#33
# or /1996-1999NotKHbook/
# but not if they are /2010/2010#33/therion/ : the wallet is /2010#33/ not /therion/
print(' - ', end='')
scans_path = Path(settings.SCANS_ROOT)
seen = []
c=0
wallets = {}
for p in scans_path.rglob('*'):
if p.is_file():
if p.suffix.lower() not in valids and p.name.lower() not in validnames:
# print(f"'{p}'", end='\n')
pass
elif p.parent == scans_path: # skip files directly in /surveyscans/
pass
else:
c+=1
if c % 15 == 0 :
print(".", end='')
if c % 500 == 0 :
print("\n -", end='')
if p.parent.parent.parent.parent == scans_path:
# print(f"too deep {p}", end='\n')
fpath = p.parent.parent
walletname = p.parent.parent.name # wallet is one level higher
else:
fpath = p.parent
walletname = p.parent.name
if walletname in wallets:
wallet = wallets[walletname]
else:
print("", flush=True, end='')
wallet = Wallet(fpath=fpath, walletname=walletname)
wallet.save()
wallets[walletname] = wallet
singlescan = SingleScan(ffile=fpath, name=p.name, wallet=wallet)
singlescan.save()
# only printing progress:
tag = p.parent
if len(walletname)>4:
if walletname[4] == "#":
tag = p.parent.parent
if tag not in seen:
print(f" {tag.name} ", end='')
seen.append(tag)
print(f'\n - found and loaded {c:,} acceptable scan files in {len(wallets):,} wallets')
# if False:
# n=0
# for topfolder, fpath, fisdir in GetListDir(settings.SCANS_ROOT):
# if not fisdir:
# continue
# # do the year folders
# # if re.match(r"\d\d\d\d$", topfolder):
# print(f"{topfolder}", end=' ')
# for walletname, fpath, fisdir in GetListDir(fpath):
# if fisdir:
# wallet = Wallet(fpath=fpath, walletname=walletname)
# # this is where we should record the year explicitly
# # line 347 of view/uploads.py and needs refactoring for loading contentsjson
# wallet.save()
# LoadListScansFile(wallet)
# # else:
# # # but We *should* load all the scans, even for nonstandard names.
# # print(f'\n - IGNORE {topfolder} - {fpath}')
# print("", flush=True)
# but we also need to check if JSON exists, even if there are no uploaded scan files.
# Here we know there is a rigid folder structure, so no need to look for sub folders
contents_path = Path(settings.DRAWINGS_DATA, "walletjson")
for yeardir in contents_path.iterdir():
if yeardir.is_dir():
for walletpath in yeardir.iterdir():
if Path(walletpath, contentsjson).is_file():
walletname = walletpath.name
if walletname not in wallets:
print(f" - {walletname} creation attempting: only JSON, no actual uploaded scan files.", end=' ')
wallet, created = Wallet.objects.update_or_create(walletname=walletname)
# should now also load the json and use it ! check &ref is correct or missing too
if created:
print(f" - {walletname} created: only JSON, no actual uploaded scan files.", end=' ')
wallet.save()