path: root/core/utils.py

import hashlib
import json
import logging
import math
import os
import random
import re
import resource
import socket
import string
import subprocess
from datetime import datetime, timezone
from decimal import getcontext
from pathlib import Path

from django.contrib.auth.models import User

from troggle.core.models.troggle import Expedition
from troggle.core.models.troggle import Person


getcontext().prec = 2  # use 2 significant figures for decimal calculations

import settings

"""This file declares TROG a globally visible object for caches.

TROG is a dictionary holding globally visible indexes and cache functions.
It is a Global Object, see https://python-patterns.guide/python/module-globals/
troggle.utils.TROG

chaosmonkey(n) - used by survex import to regenerate some .3d files

various git add/commit functions that need refactoring together

NOTE that TROG is not serialized! Two users can update it and conflict !! 
This needs to be in a multi-user database with transactions. However it may be
useful when doing a data import with databaseReset.py as that has a single
thread.

TIMEZONES
Read this now: https://nerderati.com/a-python-epoch-timestamp-timezone-trap/

"""

TROG = {"pagecache": {"expedition": {}}, "caves": {"gcavelookup": {}, "gcavecount": {}}}
alphabet = []
sha = hashlib.new('sha256')
COOKIE_MAX_AGE = 2*365*24*60*60 # seconds
COOKIE_SHORT_TIMEOUT = 60*60 # seconds
PUBLIC_LAPTOP_COOKIE_NAME = "public_laptop"
PUBLIC_LAPTOP_COOKIE_TEXT = "this is a public laptop"
throw = 35.0

EXPOSERVER = "expo" # hostname of the server at expo.survex.com

DEV_OK = """On branch master
Your branch is ahead of 'origin/master' by 1 commit.
  (use "git push" to publish your local commits)

nothing to commit, working tree clean
"""

class DatabaseResetOngoing(Exception):
    """Exception class for errors while the server is reimporting everything"""

    def __init__(self, message):
        if message:
            self.message = message
        else:
            self.message = f"Expo Database re-import ongoing on server.\nPlease wait 7 minutes.\n If still not working, contact a nerd."

    def __str__(self):
        return f"DatabaseResetOngoing: {self.message}"

# This is module-level executable. This is a Bad Thing. Especially when it touches the file system.
try:
    logging.basicConfig(level=logging.DEBUG, filename=settings.LOGFILE, filemode="w")
except:
    # Opening of file for writing is going to fail currently, so decide it doesn't matter for now
    pass

def get_cookie_max_age(request=None):
    """This is where we detect whether the machine the user is using is a shared-use device or a personal device.
    If it is shared-use, then we set a much shorter cookie timout period.
    """
    if shared_use_machine(request):
        return COOKIE_SHORT_TIMEOUT
    else:        
        return COOKIE_MAX_AGE


def sanitize_name(name):
    """Filenames sould not contain these characters as then the system barf when it tries to use them in URLs
    """
    return name.replace("#","-").replace("?","=").replace("&","+").replace(":","^")

def get_process_memory():
    usage = resource.getrusage(resource.RUSAGE_SELF)
    return usage[2] / 1024.0


def chaosmonkey(n):
    """returns True once every n calls - randomly"""
    if random.randrange(0, n) != 0:
        return False
    # print("CHAOS strikes !", file=sys.stderr)
    return True
    
def unique_slug(text, n):
    """This gives an almost-unique id based on the text, 
    2 hex digits would seem adequate, but we might get a collision.
    Deterministic
    """
    sha.update(text.encode('utf-8'))
    return sha.hexdigest()[0:n]

def random_slug(text, n):
    """This gives an almost-unique id based on the text, 
    2 hex digits would seem adequate, but we might get a collision.
    Random
    """
    text = text + alphabet_suffix(3)
    sha.update(text.encode('utf-8'))
    return sha.hexdigest()[0:n]
    
def alphabet_suffix(n):
    """This is called repeatedly during initial parsing import, hence the cached list
    """
    global alphabet
    if not alphabet:
        alphabet = list(string.ascii_lowercase)
        
    if n < len(alphabet) and n > 0:
        suffix = alphabet[n-1]
    else:
        suffix = "_X_" + random.choice(string.ascii_lowercase) + random.choice(string.ascii_lowercase)
    return suffix   

def wrap_text(text):
    """
    Splits a long string into multiple lines, ensuring that each line is no more than
    70 characters long. Newline characters are inserted immediately before spaces to
    prevent breaking words.

    Parameters:
    text (str): The input string that needs to be wrapped.

    Returns:
    str: The input string formatted with newline characters such that each line
         does not exceed 70 characters in length.

    Functionality:
    1. The function first splits the input string into individual words.
    2. It iterates through the words and adds them to the current line, ensuring the
       line's length does not exceed 70 characters.
    3. If adding a word would cause the line to exceed the limit, the current line is
       added to a list of lines, and a new line is started with the word.
    4. The process repeats until all words are processed, and the final line is added.
    5. The list of lines is then joined with newline characters and returned as a
       single formatted string.
       
    This function written by CoPilot.
    BUT interactions between existing \n characters makes this much less simple than it appears.
    """

    return text # abort all wrap processing pending proper redesign
    
    words = text.split(' ')
    lines = []
    current_line = ""

    for word in words:
        # Check if adding the word exceeds 70 characters
        if len(current_line) + len(word) + 1 <= 70:
            if current_line:
                current_line += ' '
            current_line += word
        else:
            lines.append(current_line.strip(' '))
            current_line = word

    # Add the last line
    if current_line:
        lines.append(current_line.strip(' '))

    # Join the lines with newline characters
    return '\n'.join(lines)
    
def make_new_expo(year):
        coUniqueAttribs = {"year": year}
        otherAttribs = {"name": f"CUCC expo {year}", "logbookfile": "logbook.html"}
        e = Expedition.objects.create(**otherAttribs, **coUniqueAttribs)   
        u = User.objects.get(username='expo')
        u.current_year = year
        u.save()    

def make_new_expo_dir(year):
    pages = ["index", "logbook", "mission", "travel"]
    t = "<hr />"
    for ff in pages:
        t += f"<a href='{ff}.html'>{ff}</a><br>\n"
    year_dir = Path(settings.EXPOWEB, 'years', year)
    if not year_dir.is_dir():
        year_dir.mkdir(parents=True, exist_ok=True)
        for ff in pages:
            content = f"<html><head><title>{year} {ff}</title></head><body><h1>{ff}</h1>Add content here.{t}</body></html>"
            p = Path(year_dir, ff+".html") 
            if not p.is_file():
                write_and_commit( [(p, content, "utf8")], f"Auto new year {ff} file creation", "Auto New Year <make_new_expo_dir@troggle.expo>")
    
def current_expo():
    """Returns the current expo year, but also checks if the most recent expo year is the same 
    as this year. If it is not, then it creates an empty Expedition and fixes some files and 
    folders. If we are more than one year out of date, it creates all intervening Expo objects
    and folders. You will need to tidy this up manually.
    """
    expos = Expedition.objects.all().order_by('-year')
    if expos:
        year = str(datetime.now(timezone.utc).year)
        last_expo = expos[0].year
        if int(last_expo) < int(year): # coming year, after Dec.31st
            for y in range(int(last_expo)+1, int(year)+1):
                #print(f"--!{year}---")

                make_new_expo(str(y))
                make_new_expo_dir(str(y))

        # print(f"---{year}---")
        return year
    else:
        return settings.EPOCH.year # this is 1970

def parse_aliases(aliasfile):
    """Reads a long text string containing pairs of strings: 
        (alias, target)
     where the alias is an old name for a cave and the target is the current, valid name for the cave, e.g.
     ("2015-mf-06", "1623-288"),
     returns a list of tuple pairs
     
     May fail if there is more than one space after the comma separating strings
     """

    report = [] # Messages about failed lines
    aliases = []
    
    filepath = Path(settings.EXPOWEB) / "cave_data" / aliasfile
    if not filepath.is_file():
        message = f' ** {filepath} is not a file.'
        print(message)
        return [(None, None)]
    try:
        with open(filepath, "r") as aliasfile:
            for line in aliasfile:
                l, sep, tail = line.partition('#')
                l, sep, tail = l.partition(';')
                l = l.strip().strip(',') # remove terminal comma if present
                l = l.strip().strip('()')
                l = l.replace("\"","")
                l = l.replace("\'","")
                l = l.replace(" ","") # removes all spaces
                l = " ".join(l.split(',')) # subtle, splits on comma, joins with one space
                if len(l) == 0:
                    # print(f"no parseable content: {line}")
                    continue
                key, sep, target = l.partition(' ')
                if len(key) == 0 or len(target) == 0:
                    message = f' ** Parsing failure for {line}'
                    print(message)
                    continue
                
                # print(f"{key} => {target}")
                aliases.append((key,target))
    except:
        message = f' ** Cannot open {filepath} for text file reading even though it is a file.'
        print(message)
        return [(None, None)], "Fail on file reading"
    return aliases, report

def get_editor(request):
    current_user = request.user
    if is_identified_user(current_user):
        return get_git_string(current_user)    
    else:
        return get_cookie(request)  
        
def is_identified_user(user):        
    if user.is_anonymous:
        return False
    if user.username in ["expo", "expoadmin"]:
        return False  
    return True
        
def is_admin_user(user):        
    if user.is_anonymous:
        return False
    if user.username in ["expoadmin"]:
        return True  
    if user.is_superuser: # set in parsers/users.py i.e. Wookey, Philip S.
        return True      
    return False

def get_git_string(user):
    if not is_identified_user(user):
        return None
    else:
        people = Person.objects.filter(user=user)
        if len(people) != 1:
            # someone like "fluffy-bunny" not associated with a Person
            return None
        person = people[0]
        return f"{person.fullname} <{user.email}>"    

def shared_use_machine(request):
    """Looks for a cookie which only exists on shared use machines
    """
    print(f" -  shared use cookie check {request}")
    
    if not request: # temporary while rolling out implementation to all calling functions
        return False      

    if not (cookie_txt := request.COOKIES.get(PUBLIC_LAPTOP_COOKIE_NAME, "")):
        return False
    elif cookie_txt == PUBLIC_LAPTOP_COOKIE_TEXT:        
        print(f" -  shared use cookie exists, and has expected value: '{cookie_txt}'")
        return True
    else:
        print(f" -  shared use cookie exists, but has wrong value: '{cookie_txt}' not '{PUBLIC_LAPTOP_COOKIE_TEXT}'")
        return True



    
def get_cookie(request):
    """The initial idea of having a default turned out to be a bad idea as people just ignore the field.
    If the default value is blank, then the form validation code makes the user type something in. 
    So having a blank is best if the cookie is to be displayed as the first value seen on a form.
    But if the cookie is to be stored, then "Unset" may be better.
    """
    # NO_COOKIE_DEFAULT = 'Unset Cookie <hohlenforscher@potatohut.expo>'
    # print(f"-- Getting cookie...")
    editor_id = request.COOKIES.get('editor_id', "") 
    editor =  git_string(editor_id) # belt and braces, should have been validity checked on saving already
    # print(f"-- Cookie to be used: {editor=}") 
    return editor

def git_string(author_string):
    """Rewrites the supplied editor string into a git-complient author string
    Uses a regular expression for a git-compatible author string written mostly by Copilot
    valid example "John Doe <john.doe@example.com>"
    """
    author_regex = re.compile(r'^[a-zA-Z][\w\s\_\.\-]* <[a-zA-Z0-9._%+-]+@[a-zA-Z0-9.-_]+\.[a-zA-Z]{2,}>$')

    if author_string.startswith("Unset"):
        # clean out laziness in users' PCs
        return "" 

    if author_regex.match(author_string):
        # print(f"++ Is valid git-compatible author string: '{author_string}'")
        return author_string
    else:
        editor = author_string.replace("@","_at_")
        editor = re.sub('[^0-9a-zA-Z_]+', '_', editor)
        if editor.startswith("_"):
            editor = "X" + editor
        editor += f" <{editor}@potatohut.expo>"
        print(f"++ Not git-compatible author string '{author_string}', replacing as '{editor}'")
        return editor
        
def git_add(filename, cwd, commands=[]):
    """Add a file to the list of Staged files ready for a later git commit
    """ 
    git = settings.GIT
    
    # what is the purpose of this 'git diff' ? To prevent merge conflicts happening I guess, 
    # so we do not have to reverse a 'git add'
    # print(f"git diff {filename} in {cwd}")
    cmd_diff = [git, "diff", filename]
    commands.append(cmd_diff)
    cp_diff = subprocess.run(cmd_diff, cwd=cwd, capture_output=True, text=True)
    if cp_diff.returncode != 0:
        msgdata = f"Ask a nerd to fix this DIFF problem in git_add().\n--{cp_diff.stderr}\n--{cp_diff.stdout}\n--return code:{str(cp_diff.returncode)}"
        raise WriteAndCommitError(
            f"CANNOT git ADD on server for this file {filename}.\n\n" + msgdata
        )
    
    print(f"git add {filename} in {cwd}")
    cmd_add = [git, "add", filename]
    commands.append(cmd_add)
    cp_add = subprocess.run(cmd_add, cwd=cwd, capture_output=True, text=True)
    if cp_add.returncode != 0:
        msgdata = f"Ask a nerd to fix this ADD problem in git_add().\n--{cp_add.stderr}\n--{cp_add.stdout}\n--return code:{str(cp_add.returncode)}"
        raise WriteAndCommitError(
            f"CANNOT git ADD on server for this file {filename}.\n\n" + msgdata
        )     
    return commands


def git_commit(cwd, message, editor, commands=[]):
    """Commits whatever has been Staged by git in this directory 'cwd'
    """
    git = settings.GIT    
    print(f"git commit in {cwd}")
    
    if socket.gethostname() != EXPOSERVER:
        message += f" - on dev machine '{socket.gethostname()}'"          
    elif settings.DEVSERVER:
        message += " - on a dev machine using 'runserver'" 
    print(f"..{message=}\n..{editor=}")
    cmd_commit = [git, "commit", "-m", message, "--author", f"{editor}"]
    commands.append(cmd_commit)

    cp_commit = subprocess.run(cmd_commit, cwd=cwd, capture_output=True, text=True)
    # This produces return code = 1 if it commits OK, but when the local repo still needs to be pushed to origin/repo
    # which will be the case when running a test troggle system on a development machine
    
    # Several ways of testing if the commit failed
    # if  cp_commit.stdout.split("\n")[-2] != "nothing to commit, working tree clean":
    # if cp_commit.returncode == 1 and cp_commit.stdout == DEV_OK: # only good for 1 commit ahead of origin/repo

    if cp_commit.returncode != 0 and not cp_commit.stdout.strip().endswith(
        "nothing to commit, working tree clean"
    ):
        msgdata = f'--Ask a nerd to fix this problem in git_commit().\n--{cp_commit.stderr}\n--"{cp_commit.stdout}"\n--return code:{str(cp_commit.returncode)}'
        print(msgdata)
        raise WriteAndCommitError(
            f"Error code with git on server in this directory: {cwd}. Edits saved, added to git, but NOT committed.\n\n"
            + msgdata
        )
    return commands
    
def add_commit(fname, message, editor):
    """Used to commit a survex file edited and saved in view/survex.py,
    also contents.json for an edited wallet,
    also encrypted.json for an edited user registration
    """
    cwd = fname.parent
    filename = fname.name
    commands = []
 
    if editor:
        editor =  git_string(editor)
    else:
        # 'cannot happen' as form verification has this is an obligatory field
        editor = "Anathema Device <a.device@potatohut.expo>"
        
    try:
        # print(f"add_commit: {editor=} {filename=} {cwd=}")
        commands = git_add(filename, cwd, commands)
        commands = git_commit(cwd, message, editor, commands)
 
    except subprocess.SubprocessError:
        msg =  f"CANNOT git ADD or COMMIT on server for this file {filename}.\nSubprocess error: {commands}\nEdits probably not saved.\nAsk a nerd to fix this."
        print(msg)
        raise WriteAndCommitError(msg)

def write_binary_file(filepath, content):
    print(f"write_binary_file: {filepath}")
    write_files([(filepath, content, "")]) # null encoding does "wb"

def ensure_dir_exists(filepath):
    """Takes a filepath for a file and all the parent directiories,
    makes any directories necessary to make the filepath valid
    """
    if filepath.is_dir():
        raise OSError(
            f"CANNOT  write this file {filepath} as this is an existing DIRECTORY."
        )
    try:
        filepath.parent.mkdir(parents=True, exist_ok=True)
        # os.makedirs(os.path.dirname(filepath), exist_ok = True)
    except PermissionError as e:
        raise PermissionError(
            f"CANNOT  make the directory.\nPERMISSIONS incorrectly set on server for this file {filepath}. Ask a nerd to fix this: {e}"
        )
    except Exception as e:
        raise OSError(
            f"CANNOT  make the directory for {filepath}. Ask a nerd to fix this: {e}"
        )      
     
def write_files(files):
    for filepath, content, encoding in files:
        filename = filepath.name
        ensure_dir_exists(filepath)
        if encoding:
            mode = "w"
            kwargs = {"encoding": encoding}
        else:
            mode = "wb"
            kwargs = {}
        try:
            with open(filepath, mode, **kwargs) as f:
                # print(f"WRITING {filepath} ")
                f.write(content)
        except PermissionError as e:
            raise PermissionError(
                f"CANNOT  save this file.\nPERMISSIONS incorrectly set on server for this file {filepath}. Ask a nerd to fix this: {e}"
            )
        except Exception as e:
            raise OSError(
                f"CANNOT  write this file {filepath}. Ask a nerd to fix this: {e}"
            )      
    return

def write_and_commit(files, message, editor):
    """For each file in files, it writes the content to the filepath 
    and adds and commits the file to git.  
    filepath, content, encoding = file
    If this fails, a WriteAndCommitError is raised.
   
    message - the "-m" comment field for the git commit
    editor  - the "--author" field for the git commit
    """
    # GIT see also core/views/uploads.py dwgupload()
    # GIT see also core/views/expo.py editexpopage()
    git = settings.GIT
    commands = []
    editor =  git_string(editor)
    write_files(files)
    try:
        for filepath, content, encoding in files:
            cwd = filepath.parent
            filename = filepath.name
            commands = git_add(filename, cwd, commands)               
        commands = git_commit(cwd, message, editor, commands)
        
    except subprocess.SubprocessError:
        raise WriteAndCommitError(
            f"CANNOT git on server for this file {filename}. Subprocess error. Edits not saved.\nAsk a nerd to fix this."
        )
    return True


class WriteAndCommitError(Exception):
    """Exception class for errors writing files and comitting them to git"""

    def __init__(self, message):
        self.message = message

    def __str__(self):
        return f"WriteAndCommitError: {self.message}"


"""The following is a Bard converted version of Radosts's MIT copyrighted Javascript on 2023-10-27
    with hand-editing.

    >>> height_from_utm(326123.456, 4896789.012)
    1234.56 (reference: 100m away)
"""
def height_from_utm(easting, northing):
    """Returns the height from UTM coordinates.
    This does NO interpolation. It just return the height of the nearest point in the height data list"""
    global rawheights
    
    res = find_nearest_point(rawheights, (easting, northing))
    winner, distance = res
    #return f"{winner} (reference: {(distance).toFixed(0)}m away)"
    return winner, distance


def find_nearest_point(points, target_point):
    """Returns the nearest point to a target point from a list of points.
    
    TODO FIND OUT 
    1. is this SRTM data ? TICK. Yes.
    2. what is the zero altitude datum? Geoid or ellisoid ? Do we need to subtract 47m ??
    3. remove all these numbers from the .py file as it is confusing the code length calcs
    
    In our dataset, the survey stations are all within 30m of an srtm reference point.
    So we can safely ignore points more than 100m away in either x or y directions.
    
    This is not true for the 1624 and 1627 areas though.
    
    TO DO: store this list twice, once sorted by x and once sorted by y,
    do a bounding box search for a set of nearby points before doing the exhaustive pythagorean distance search."""
    nearest_distance_squared = float("inf")
    nearest_point = None
    x_target, y_target = target_point
    
    max_ds = throw * throw # 1089 = 33 x 33

    for point in points:
        x, y, z = point
        if z < 0:
            distance_squared = 1000000.0 # ie ignore it
        else:
            if x - x_target > throw:
                distance_squared = max_ds
            elif y - y_target > throw:
                distance_squared = max_ds
            else:
                distance_squared = math.pow(x - x_target, 2) + math.pow(y - y_target, 2)

        if distance_squared < nearest_distance_squared:
            nearest_distance_squared = distance_squared
            nearest_point_alt = z
    nearest_distance = math.sqrt(nearest_distance_squared)
    return nearest_point_alt, nearest_distance

# This data was extarcted by Radost:
# The original source of heights (and positions is the file) 
# loser/surface/terrain/SRTM-7x9-M31-482-279-offset.svx
"""
; Survex terrain data created by TerrainTool version 1.12a from Shuttle Radar Topography Mission data
; SRTM DEM data is public-domain.
; Used coordinate system AustrianM31 with ellipsoid "Bessel 1841" and datum "MGI Datum (Austria)"
; Grid centred at XXX   488000   283750
; Lat/Long of centre =  47.691318   13.838777  ( 13 E 50' 19.6"  47 N 41' 28.7") relative to WGS84/GRS80 datum
; Grid convergence at centre   =  -0.374 degrees (  0 W 22' 27.9")
; Point scale factor at centre =  1.0000177
; Output offset by adding (-450000.00,-200000.00,    0.00) to calculated results;
; TerrainTool (c) 2008 - 2012 Mike McCombe 
"""


# Load the data from the JSON file
hpath = settings.TROGGLE_PATH / "core" / "heights.json"
with open(hpath, 'r') as file:
    rawheights = json.load(file)

# Now you can use rawheights as before
# print(rawheights)