Move slow code to async QThread workers (WIP)

requirements.d/Brewfile

@@ -4,6 +4,7 @@
 brew 'create-dmg'
 brew 'qt'
 brew 'hub'
+brew 'pre-commit'
 brew 'xmlstarlet'
 cask 'qt-creator'
 cask 'sparkle'

setup.cfg

@@ -105,7 +105,7 @@ extension-pkg-whitelist=PyQt6
 load-plugins=
 
 [pylint.messages control]
-disable= W0503,W0511,C0301,R0903,R0201,W0212,C0114,C0115,C0116,C0103,E0611,E1120,C0415,R0914,R0912,R0915
+disable= W0511,C0301,R0903,W0212,C0114,C0115,C0116,C0103,E0611,E1120,C0415,R0914,R0912,R0915
 
 [pylint.format]
 max-line-length=120

src/vorta/store/models.py

@@ -12,7 +12,7 @@
 import peewee as pw
 from playhouse import signals
 
-from vorta.utils import slugify
+from vorta.store.utils import slugify
 from vorta.views.utils import get_exclusion_presets
 
 DB = pw.Proxy()

src/vorta/store/utils.py

@@ -0,0 +1,15 @@
+import re
+import unicodedata
+
+
+def slugify(value):
+    """
+    Converts to lowercase, removes non-word characters (alphanumerics and
+    underscores) and converts spaces to hyphens. Also strips leading and
+    trailing whitespace.
+
+    Copied from Django.
+    """
+    value = unicodedata.normalize('NFKD', value).encode('ascii', 'ignore').decode('ascii')
+    value = re.sub(r'[^\w\s-]', '', value).strip().lower()
+    return re.sub(r'[-\s]+', '-', value)

src/vorta/utils.py

@@ -1,21 +1,21 @@
 import argparse
 import errno
-import fnmatch
 import getpass
 import math
 import os
 import re
 import socket
 import sys
-import unicodedata
 from datetime import datetime as dt
 from functools import reduce
 from typing import Any, Callable, Iterable, List, Optional, Tuple, TypeVar
 
-import psutil
 from PyQt6 import QtCore
-from PyQt6.QtCore import QFileInfo, QThread, pyqtSignal
-from PyQt6.QtWidgets import QApplication, QFileDialog, QSystemTrayIcon
+from PyQt6.QtWidgets import (
+    QApplication,
+    QFileDialog,
+    QSystemTrayIcon,
+)
 
 from vorta.borg._compatibility import BorgCompatibility
 from vorta.log import logger
@@ -26,130 +26,9 @@
 DEFAULT_DIR_FLAG = object()
 METRIC_UNITS = ['', 'K', 'M', 'G', 'T', 'P', 'E', 'Z', 'Y']
 NONMETRIC_UNITS = ['', 'Ki', 'Mi', 'Gi', 'Ti', 'Pi', 'Ei', 'Zi', 'Yi']
-
-borg_compat = BorgCompatibility()
 _network_status_monitor = None
 
-
-class FilePathInfoAsync(QThread):
-    signal = pyqtSignal(str, str, str)
-
-    def __init__(self, path, exclude_patterns_str):
-        self.path = path
-        QThread.__init__(self)
-        self.exiting = False
-        self.exclude_patterns = []
-        for _line in (exclude_patterns_str or '').splitlines():
-            line = _line.strip()
-            if line != '':
-                self.exclude_patterns.append(line)
-
-    def run(self):
-        # logger.info("running thread to get path=%s...", self.path)
-        self.size, self.files_count = get_path_datasize(self.path, self.exclude_patterns)
-        self.signal.emit(self.path, str(self.size), str(self.files_count))
-
-
-def normalize_path(path):
-    """normalize paths for MacOS (but do nothing on other platforms)"""
-    # HFS+ converts paths to a canonical form, so users shouldn't be required to enter an exact match.
-    # Windows and Unix filesystems allow different forms, so users always have to enter an exact match.
-    return unicodedata.normalize('NFD', path) if sys.platform == 'darwin' else path
-
-
-# prepare patterns as borg does
-# see `FnmatchPattern._prepare` at
-# https://github.com/borgbackup/borg/blob/master//src/borg/patterns.py
-def prepare_pattern(pattern):
-    """Prepare and process fnmatch patterns as borg does"""
-    if pattern.endswith(os.path.sep):
-        # trailing sep indicates that the contents should be excluded
-        # but not the directory it self.
-        pattern = os.path.normpath(pattern).rstrip(os.path.sep)
-        pattern += os.path.sep + '*' + os.path.sep
-    else:
-        pattern = os.path.normpath(pattern) + os.path.sep + '*'
-
-    pattern = pattern.lstrip(os.path.sep)  # sep at beginning is removed
-    return re.compile(fnmatch.translate(pattern))
-
-
-def match(pattern: re.Pattern, path: str):
-    """Check whether a path matches the given pattern."""
-    path = path.lstrip(os.path.sep) + os.path.sep
-    return pattern.match(path) is not None
-
-
-def get_directory_size(dir_path, exclude_patterns):
-    '''Get number of files only and total size in bytes from a path.
-    Based off https://stackoverflow.com/a/17936789'''
-    exclude_patterns = [prepare_pattern(p) for p in exclude_patterns]
-
-    data_size_filtered = 0
-    seen = set()
-    seen_filtered = set()
-
-    for dir_path, subdirectories, file_names in os.walk(dir_path, topdown=True):
-        is_excluded = False
-        for pattern in exclude_patterns:
-            if match(pattern, dir_path):
-                is_excluded = True
-                break
-
-        if is_excluded:
-            subdirectories.clear()  # so that os.walk won't walk them
-            continue
-
-        for file_name in file_names:
-            file_path = os.path.join(dir_path, file_name)
-
-            # Ignore symbolic links, since borg doesn't follow them
-            if os.path.islink(file_path):
-                continue
-
-            is_excluded = False
-            for pattern in exclude_patterns:
-                if match(pattern, file_path):
-                    is_excluded = True
-                    break
-
-            try:
-                stat = os.stat(file_path)
-                if stat.st_ino not in seen:  # Visit each file only once
-                    # this won't add the size of a hardlinked file
-                    seen.add(stat.st_ino)
-                    if not is_excluded:
-                        data_size_filtered += stat.st_size
-                        seen_filtered.add(stat.st_ino)
-            except (FileNotFoundError, PermissionError):
-                continue
-
-    files_count_filtered = len(seen_filtered)
-
-    return data_size_filtered, files_count_filtered
-
-
-def get_network_status_monitor():
-    global _network_status_monitor
-    if _network_status_monitor is None:
-        _network_status_monitor = NetworkStatusMonitor.get_network_status_monitor()
-        logger.info(
-            'Using %s NetworkStatusMonitor implementation.',
-            _network_status_monitor.__class__.__name__,
-        )
-    return _network_status_monitor
-
-
-def get_path_datasize(path, exclude_patterns):
-    file_info = QFileInfo(path)
-
-    if file_info.isDir():
-        data_size, files_count = get_directory_size(file_info.absoluteFilePath(), exclude_patterns)
-    else:
-        data_size = file_info.size()
-        files_count = 1
-
-    return data_size, files_count
+borg_compat = BorgCompatibility()
 
 
 def nested_dict():
@@ -220,23 +99,6 @@ def get_private_keys() -> List[str]:
     return available_private_keys
 
 
-def sort_sizes(size_list):
-    """Sorts sizes with extensions. Assumes that size is already in largest unit possible"""
-    final_list = []
-    for suffix in [" B", " KB", " MB", " GB", " TB", " PB", " EB", " ZB", " YB"]:
-        sub_list = [
-            float(size[: -len(suffix)])
-            for size in size_list
-            if size.endswith(suffix) and size[: -len(suffix)][-1].isnumeric()
-        ]
-        sub_list.sort()
-        final_list += [(str(size) + suffix) for size in sub_list]
-        # Skip additional loops
-        if len(final_list) == len(size_list):
-            break
-    return final_list
-
-
 Number = TypeVar("Number", int, float)
 
 
@@ -245,6 +107,17 @@ def clamp(n: Number, min_: Number, max_: Number) -> Number:
     return min(max_, max(n, min_))
 
 
+def get_network_status_monitor():
+    global _network_status_monitor
+    if _network_status_monitor is None:
+        _network_status_monitor = NetworkStatusMonitor.get_network_status_monitor()
+        logger.info(
+            'Using %s NetworkStatusMonitor implementation.',
+            _network_status_monitor.__class__.__name__,
+        )
+    return _network_status_monitor
+
+
 def find_best_unit_for_sizes(sizes: Iterable[int], metric: bool = True, precision: int = 1) -> int:
     """
     Selects the index of the biggest unit (see the lists in the pretty_bytes function) capable of
@@ -303,38 +176,6 @@ def get_asset(path):
     return os.path.join(bundle_dir, path)
 
 
-def get_sorted_wifis(profile):
-    """
-    Get Wifi networks known to the OS (only current one on macOS) and
-    merge with networks from other profiles. Update last connected time.
-    """
-
-    from vorta.store.models import WifiSettingModel
-
-    # Pull networks known to OS and all other backup profiles
-    system_wifis = get_network_status_monitor().get_known_wifis()
-    from_other_profiles = WifiSettingModel.select().where(WifiSettingModel.profile != profile.id).execute()
-
-    for wifi in list(from_other_profiles) + system_wifis:
-        db_wifi, created = WifiSettingModel.get_or_create(
-            ssid=wifi.ssid,
-            profile=profile.id,
-            defaults={'last_connected': wifi.last_connected, 'allowed': True},
-        )
-
-        # Update last connected time
-        if not created and db_wifi.last_connected != wifi.last_connected:
-            db_wifi.last_connected = wifi.last_connected
-            db_wifi.save()
-
-    # Finally return list of networks and settings for that profile
-    return (
-        WifiSettingModel.select()
-        .where(WifiSettingModel.profile == profile.id)
-        .order_by(-WifiSettingModel.last_connected)
-    )
-
-
 def parse_args():
     parser = argparse.ArgumentParser(description='Vorta Backup GUI for Borg.')
     parser.add_argument('--version', '-V', action='store_true', help="Show version and exit.")
@@ -368,19 +209,6 @@ def parse_args():
     return parser.parse_known_args()[0]
 
 
-def slugify(value):
-    """
-    Converts to lowercase, removes non-word characters (alphanumerics and
-    underscores) and converts spaces to hyphens. Also strips leading and
-    trailing whitespace.
-
-    Copied from Django.
-    """
-    value = unicodedata.normalize('NFKD', value).encode('ascii', 'ignore').decode('ascii')
-    value = re.sub(r'[^\w\s-]', '', value).strip().lower()
-    return re.sub(r'[-\s]+', '-', value)
-
-
 def uses_dark_mode():
     """
     This function detects if we are running in dark mode (e.g. macOS dark mode).
@@ -431,61 +259,6 @@ def format_archive_name(profile, archive_name_tpl):
 SHELL_PATTERN_ELEMENT = re.compile(r'([?\[\]*])')
 
 
-def get_mount_points(repo_url):
-    mount_points = {}
-    repo_mounts = []
-    for proc in psutil.process_iter():
-        try:
-            name = proc.name()
-            if name == 'borg' or name.startswith('python'):
-                if 'mount' not in proc.cmdline():
-                    continue
-
-                if borg_compat.check('V2'):
-                    # command line syntax:
-                    # `borg mount -r <repo> <mountpoint> <path> (-a <archive_pattern>)`
-                    cmd = proc.cmdline()
-                    if repo_url in cmd:
-                        i = cmd.index(repo_url)
-                        if len(cmd) > i + 1:
-                            mount_point = cmd[i + 1]
-
-                            # Archive mount?
-                            ao = '-a' in cmd
-                            if ao or '--match-archives' in cmd:
-                                i = cmd.index('-a' if ao else '--match-archives')
-                                if len(cmd) >= i + 1 and not SHELL_PATTERN_ELEMENT.search(cmd[i + 1]):
-                                    mount_points[mount_point] = cmd[i + 1]
-                            else:
-                                repo_mounts.append(mount_point)
-                else:
-                    for idx, parameter in enumerate(proc.cmdline()):
-                        if parameter.startswith(repo_url):
-                            # mount from this repo
-
-                            # The borg mount command specifies that the mount_point
-                            # parameter comes after the archive name
-                            if len(proc.cmdline()) > idx + 1:
-                                mount_point = proc.cmdline()[idx + 1]
-
-                                # archive or full mount?
-                                if parameter[len(repo_url) :].startswith('::'):
-                                    archive_name = parameter[len(repo_url) + 2 :]
-                                    mount_points[archive_name] = mount_point
-                                    break
-                                else:
-                                    # repo mount point
-                                    repo_mounts.append(mount_point)
-
-        except (psutil.ZombieProcess, psutil.AccessDenied, psutil.NoSuchProcess):
-            # Getting process details may fail (e.g. zombie process on macOS)
-            # or because the process is owned by another user.
-            # Also see https://github.com/giampaolo/psutil/issues/783
-            continue
-
-    return mount_points, repo_mounts
-
-
 def is_system_tray_available():
     app = QApplication.instance()
     if app is None: