ci: Merge artifacts

- Merge the resulting artifacts
- Use only the .so files from build

.github/workflows/default.yml

@@ -4,71 +4,42 @@ name: Default CI/CD
 
 on:
   push:
-    branches: [master]
   pull_request:
     branches: [master]
 
 jobs:
-  lint:
+  pre-commit:
     runs-on: ubuntu-latest
     steps:
-      - uses: actions/checkout@v2
-      - name: Set up Python 3.10
-        uses: actions/setup-python@v2
+      - uses: actions/checkout@v4
+      - name: Set up Python 3.12
+        uses: actions/setup-python@v5
         with:
-          python-version: "3.10"
-      - name: Install dependencies
-        run: |
-          python -m pip install --upgrade pip
-          pip install -r requirements.txt -r requirements-extra.txt
-      - name: Lint with flake8
-        run: |
-          flake8 .
-  format:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v2
-      - name: Set up Python 3.10
-        uses: actions/setup-python@v2
-        with:
-          python-version: "3.10"
-      - name: Install dependencies
-        run: |
-          python -m pip install --upgrade pip
-          pip install -r requirements.txt -r requirements-extra.txt
-      - name: Check format with black
-        run: |
-          black .
+          python-version: "3.12"
+      - uses: pre-commit/action@v3.0.1
   test:
-    needs: [lint, format]
+    needs: [pre-commit]
     runs-on: ${{ matrix.os }}
     strategy:
       matrix:
-        os: [ubuntu-latest, macos-latest, windows-latest]
-        python-version: [3.7, 3.8, 3.9, "3.10"]
-        exclude:
-          - os: macos-latest
-            python-version: 3.7
-          - os: macos-latest
-            python-version: 3.8
-          - os: macos-latest
-            python-version: 3.9
+        os: [ubuntu-latest]
+        python-version: [3.7, 3.8, 3.9, "3.10", "3.11", "3.12"]
+        include:
           - os: windows-latest
-            python-version: 3.7
-          - os: windows-latest
-            python-version: 3.8
-          - os: windows-latest
-            python-version: 3.9
+            python-version: "3.12"
+          - os: macos-latest
+            python-version: "3.12"
 
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v4
       - name: Set up Python ${{ matrix.python-version }}
-        uses: actions/setup-python@v2
+        uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python-version }}
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
+          pip install setuptools
           pip install -r requirements.txt -r requirements-extra.txt
       - name: Build python modules
         run: |
@@ -78,7 +49,17 @@ jobs:
           pytest core hscommon
       - name: Upload Artifacts
         if: matrix.os == 'ubuntu-latest'
-        uses: actions/upload-artifact@v3
+        uses: actions/upload-artifact@v4
         with:
           name: modules ${{ matrix.python-version }}
-          path: ${{ github.workspace }}/**/*.so
+          path: build/**/*.so
+  merge-artifacts:
+    needs: [test]
+    runs-on: ubuntu-latest
+    steps:
+      - name: Merge Artifacts
+        uses: actions/upload-artifact/merge@v4
+        with:
+          name: modules
+          pattern: modules*
+          delete-merged: true

.github/workflows/tx-push.yml

@@ -0,0 +1,26 @@
+# Push translation source to Transifex
+name: Transifex Sync
+
+on:
+  push:
+    branches:
+      - master
+    paths:
+      - locale/*.pot
+
+env:
+  TX_VERSION: "v1.6.10"
+
+jobs:
+  push-source:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Get Transifex Client
+        run: |
+          curl -o- https://raw.githubusercontent.com/transifex/cli/master/install.sh | bash -s -- $TX_VERSION
+      - name: Update & Push Translation Sources
+        env:
+          TX_TOKEN: ${{ secrets.TX_TOKEN }}
+        run: |
+          ./tx push -s --use-git-timestamps

.gitignore

@@ -87,8 +87,8 @@ cython_debug/
 # Visual Studio Code
 .vscode/*
 !.vscode/settings.json
-#!.vscode/tasks.json
-#!.vscode/launch.json
+!.vscode/tasks.json
+!.vscode/launch.json
 !.vscode/extensions.json
 !.vscode/*.code-snippets
 
@@ -108,4 +108,4 @@ cocoa/autogen
 
 *.waf*
 .lock-waf*
-/tags
+/tags

.pre-commit-config.yaml

@@ -0,0 +1,24 @@
+repos:
+  - repo: https://github.com/pre-commit/pre-commit-hooks
+    rev: v4.5.0
+    hooks:
+      - id: check-yaml
+      - id: check-toml
+      - id: end-of-file-fixer
+        exclude: ".*.json"
+      - id: trailing-whitespace
+  - repo: https://github.com/psf/black
+    rev: 24.2.0
+    hooks:
+      - id: black
+  - repo: https://github.com/PyCQA/flake8
+    rev: 7.0.0
+    hooks:
+      - id: flake8
+        exclude: ^(.tox|env|build|dist|help|qt/dg_rc.py|pkg).*
+  - repo: https://github.com/alessandrojcm/commitlint-pre-commit-hook
+    rev: v9.11.0
+    hooks:
+      - id: commitlint
+        stages: [commit-msg]
+        additional_dependencies: ["@commitlint/config-conventional"]

.sonarcloud.properties

@@ -1 +1 @@
-sonar.python.version=3.7, 3.8, 3.9, 3.10
+sonar.python.version=3.7, 3.8, 3.9, 3.10, 3.11

.tx/config

@@ -13,15 +13,8 @@ source_file = locale/core.pot
 source_lang = en
 type        = PO
 
-[o:voltaicideas:p:dupeguru-1:r:qtlib]
-file_filter = qtlib/locale/<lang>/LC_MESSAGES/qtlib.po
-source_file = qtlib/locale/qtlib.pot
-source_lang = en
-type        = PO
-
 [o:voltaicideas:p:dupeguru-1:r:ui]
 file_filter = locale/<lang>/LC_MESSAGES/ui.po
 source_file = locale/ui.pot
 source_lang = en
 type        = PO
-

.vscode/extensions.json

@@ -3,8 +3,10 @@
     "recommendations": [
         "redhat.vscode-yaml",
         "ms-python.vscode-pylance",
-        "ms-python.python"
+        "ms-python.python",
+        "ms-python.black-formatter",
     ],
-    // List of extensions recommended by VS Code that should not be recommended for users of this workspace.
+    // List of extensions recommended by VS Code that should not be recommended for
+    // users of this workspace.
     "unwantedRecommendations": []
 }

.vscode/launch.json

@@ -0,0 +1,17 @@
+{
+    // Use IntelliSense to learn about possible attributes.
+    // Hover to view descriptions of existing attributes.
+    // For more information, visit: https://go.microsoft.com/fwlink/?linkid=830387
+    "version": "0.2.0",
+    "configurations": [
+        {
+            "name": "DupuGuru",
+            "type": "debugpy",
+            "request": "launch",
+            "program": "run.py",
+            "console": "integratedTerminal",
+            "subProcess": true,
+            "justMyCode": false
+        },
+    ]
+}

.vscode/settings.json

@@ -1,12 +1,17 @@
 {
-    "python.formatting.provider": "black",
     "cSpell.words": [
         "Dupras",
         "hscommon"
     ],
+    "editor.rulers": [
+        88,
+        120
+    ],
     "python.languageServer": "Pylance",
     "yaml.schemaStore.enable": true,
-    "yaml.schemas": {
-        "https://json.schemastore.org/github-workflow.json": ".github/workflows/*.yml"
-    }
+    "[python]": {
+        "editor.formatOnSave": true,
+        "editor.defaultFormatter": "ms-python.black-formatter"
+    },
+    "python.testing.pytestEnabled": true
 }

LICENSE

@@ -619,4 +619,3 @@ Program, unless a warranty or assumption of liability accompanies a
 copy of the Program in return for a fee.
 
                      END OF TERMS AND CONDITIONS
-

MANIFEST.in

@@ -3,4 +3,3 @@ recursive-include core *.m
 include run.py
 graft locale
 graft help
-graft qtlib/locale

Makefile

@@ -4,7 +4,7 @@ PYRCC5 ?= pyrcc5
 REQ_MINOR_VERSION = 7
 PREFIX ?= /usr/local
 
-# Window compatability via Msys2 
+# Window compatability via Msys2
 # - venv creates Scripts instead of bin
 # - compile generates .pyd instead of .so
 # - venv with --sytem-site-packages has issues on windows as well...
@@ -12,7 +12,7 @@ PREFIX ?= /usr/local
 ifeq ($(shell ${PYTHON} -c "import platform; print(platform.system())"), Windows)
 	BIN = Scripts
 	SO = *.pyd
-	VENV_OPTIONS = 
+	VENV_OPTIONS =
 else
 	BIN = bin
 	SO = *.so
@@ -35,7 +35,7 @@ endif
 # Our build scripts are not very "make like" yet and perform their task in a bundle. For now, we
 # use one of each file to act as a representative, a target, of these groups.
 
-packages = hscommon qtlib core qt
+packages = hscommon core qt
 localedirs = $(wildcard locale/*/LC_MESSAGES)
 pofiles = $(wildcard locale/*/LC_MESSAGES/*.po)
 mofiles = $(patsubst %.po,%.mo,$(pofiles))
@@ -43,7 +43,7 @@ mofiles = $(patsubst %.po,%.mo,$(pofiles))
 vpath %.po $(localedirs)
 vpath %.mo $(localedirs)
 
-all: | env i18n modules qt/dg_rc.py 
+all: | env i18n modules qt/dg_rc.py
 	@echo "Build complete! You can run dupeGuru with 'make run'"
 
 run:
@@ -82,7 +82,7 @@ qt/dg_rc.py: qt/dg.qrc
 i18n: $(mofiles)
 
 %.mo: %.po
-	msgfmt -o $@ $<	
+	msgfmt -o $@ $<
 
 modules: | env
 	$(VENV_PYTHON) build.py --modules

README.md

@@ -1,16 +1,12 @@
 # dupeGuru
 
 [dupeGuru][dupeguru] is a cross-platform (Linux, OS X, Windows) GUI tool to find duplicate files in
-a system. It is written mostly in Python 3 and has the peculiarity of using
-[multiple GUI toolkits][cross-toolkit], all using the same core Python code. On OS X, the UI layer
-is written in Objective-C and uses Cocoa. On Linux, it is written in Python and uses Qt5.
-
-The Cocoa UI of dupeGuru is hosted in a separate repo: https://github.com/arsenetar/dupeguru-cocoa
+a system. It is written mostly in Python 3 and uses [qt](https://www.qt.io/) for the UI.
 
 ## Current status
 Still looking for additional help especially with regards to:
-* OSX maintenance: reproducing bugs & cocoa version, building package with Cocoa UI.
-* Linux maintenance: reproducing bugs, maintaining PPA repository, Debian package.
+* OSX maintenance: reproducing bugs, packaging verification.
+* Linux maintenance: reproducing bugs, maintaining PPA repository, Debian package, rpm package.
 * Translations: updating missing strings, transifex project at https://www.transifex.com/voltaicideas/dupeguru-1
 * Documentation: keeping it up-to-date.
 
@@ -26,7 +22,6 @@ This folder contains the source for dupeGuru. Its documentation is in `help`, bu
 * help: Help document, written for Sphinx.
 * locale: .po files for localization.
 * hscommon: A collection of helpers used across HS applications.
-* qtlib: A collection of helpers used across Qt UI codebases of HS applications.
 
 ## How to build dupeGuru from source
 
@@ -43,12 +38,10 @@ For macos instructions (qt version) see the [macOS Instructions](macos.md).
 When running in a linux based environment the following system packages or equivalents are needed to build:
 * python3-pyqt5
 * pyqt5-dev-tools (on some systems, see note)
-* python3-wheel (for hsaudiotag3k)
 * python3-venv (only if using a virtual environment)
 * python3-dev
 * build-essential
 
-
 Note: On some linux systems pyrcc5 is not put on the path when installing python3-pyqt5, this will cause some issues with the resource files (and icons). These systems should have a respective pyqt5-dev-tools package, which should also be installed. The presence of pyrcc5 can be checked with `which pyrcc5`.  Debian based systems need the extra package, and Arch does not.
 
 To create packages the following are also needed:
@@ -70,7 +63,7 @@ dupeGuru comes with a makefile that can be used to build and run:
     $ python run.py
 
 ### Generating Debian/Ubuntu package
-To generate packages the extra requirements in requirements-extra.txt must be installed, the 
+To generate packages the extra requirements in requirements-extra.txt must be installed, the
 steps are as follows:
 
     $ cd <dupeGuru directory>

Windows.md

@@ -29,7 +29,7 @@ To build with a different python version 3.7 vs 3.8 or 32 bit vs 64 bit specify
 It is possible to build dupeGuru with the makefile on windows using a compatable POSIX environment.  The following steps have been tested using [msys2][msys2]. Before running make:
 1. Install msys2 or other POSIX environment
 2. Install PyQt5 globally via pip
-3. Use the respective console for msys2 it is `msys2 msys` 
+3. Use the respective console for msys2 it is `msys2 msys`
 
 Then the following execution of the makefile should work.  Pass the correct value for PYTHON to the makefile if not on the path as python3.
 

build.py

@@ -61,7 +61,7 @@ def parse_args():
 
 
 def build_one_help(language):
-    print("Generating Help in {}".format(language))
+    print(f"Generating Help in {language}")
     current_path = Path(".").absolute()
     changelog_path = current_path.joinpath("help", "changelog")
     tixurl = "https://github.com/arsenetar/dupeguru/issues/{}"
@@ -88,14 +88,8 @@ def build_help():
         p.map(build_one_help, languages)
 
 
-def build_qt_localizations():
-    loc.compile_all_po(Path("qtlib", "locale"))
-    loc.merge_locale_dir(Path("qtlib", "locale"), "locale")
-
-
 def build_localizations():
     loc.compile_all_po("locale")
-    build_qt_localizations()
     locale_dest = Path("build", "locale")
     if locale_dest.exists():
         shutil.rmtree(locale_dest)
@@ -109,25 +103,16 @@ def build_updatepot():
     print("Building columns.pot")
     loc.generate_pot(["core"], Path("locale", "columns.pot"), ["coltr"])
     print("Building ui.pot")
-    # When we're not under OS X, we don't want to overwrite ui.pot because it contains Cocoa locs
-    # We want to merge the generated pot with the old pot in the most preserving way possible.
-    ui_packages = ["qt", Path("cocoa", "inter")]
-    loc.generate_pot(ui_packages, Path("locale", "ui.pot"), ["tr"], merge=True)
-    print("Building qtlib.pot")
-    loc.generate_pot(["qtlib"], Path("qtlib", "locale", "qtlib.pot"), ["tr"])
+    loc.generate_pot(["qt"], Path("locale", "ui.pot"), ["tr"], merge=True)
 
 
 def build_mergepot():
     print("Updating .po files using .pot files")
     loc.merge_pots_into_pos("locale")
-    loc.merge_pots_into_pos(Path("qtlib", "locale"))
-    # loc.merge_pots_into_pos(Path("cocoalib", "locale"))
 
 
 def build_normpo():
     loc.normalize_all_pos("locale")
-    loc.normalize_all_pos(Path("qtlib", "locale"))
-    # loc.normalize_all_pos(Path("cocoalib", "locale"))
 
 
 def build_pe_modules():
@@ -144,7 +129,8 @@ def build_normal():
     print("Building localizations")
     build_localizations()
     print("Building Qt stuff")
-    print_and_do("pyrcc5 {0} > {1}".format(Path("qt", "dg.qrc"), Path("qt", "dg_rc.py")))
+    Path("qt", "dg_rc.py").unlink(missing_ok=True)
+    print_and_do("pyrcc5 {} > {}".format(Path("qt", "dg.qrc"), Path("qt", "dg_rc.py")))
     fix_qt_resource_file(Path("qt", "dg_rc.py"))
     build_help()
 

commitlint.config.js

@@ -0,0 +1,17 @@
+const Configuration = {
+    /*
+     * Resolve and load @commitlint/config-conventional from node_modules.
+     * Referenced packages must be installed
+     */
+    extends: ['@commitlint/config-conventional'],
+    /*
+     * Any rules defined here will override rules from @commitlint/config-conventional
+     */
+    rules: {
+        'header-max-length': [2, 'always', 72],
+        'subject-case': [2, 'always', 'sentence-case'],
+        'scope-enum': [2, 'always'],
+    },
+};
+
+module.exports = Configuration;

core/__init__.py

@@ -1,2 +1,2 @@
-__version__ = "4.2.1"
+__version__ = "4.3.1"
 __appname__ = "dupeGuru"

core/app.py

@@ -4,37 +4,39 @@
 # which should be included with this package. The terms are also available at
 # http://www.gnu.org/licenses/gpl-3.0.html
 
+import cProfile
+import datetime
 import os
 import os.path as op
 import logging
 import subprocess
 import re
 import shutil
+from pathlib import Path
 
 from send2trash import send2trash
 from hscommon.jobprogress import job
 from hscommon.notify import Broadcaster
-from hscommon.path import Path
 from hscommon.conflict import smart_move, smart_copy
 from hscommon.gui.progress_window import ProgressWindow
 from hscommon.util import delete_if_empty, first, escape, nonone, allsame
 from hscommon.trans import tr
 from hscommon import desktop
 
-from . import se, me, pe
-from .pe.photo import get_delta_dimensions
-from .util import cmp_value, fix_surrogate_encoding
-from . import directories, results, export, fs, prioritize
-from .ignore import IgnoreList
-from .exclude import ExcludeDict as ExcludeList
-from .scanner import ScanType
-from .gui.deletion_options import DeletionOptions
-from .gui.details_panel import DetailsPanel
-from .gui.directory_tree import DirectoryTree
-from .gui.ignore_list_dialog import IgnoreListDialog
-from .gui.exclude_list_dialog import ExcludeListDialogCore
-from .gui.problem_dialog import ProblemDialog
-from .gui.stats_label import StatsLabel
+from core import se, me, pe
+from core.pe.photo import get_delta_dimensions
+from core.util import cmp_value, fix_surrogate_encoding
+from core import directories, results, export, fs, prioritize
+from core.ignore import IgnoreList
+from core.exclude import ExcludeDict as ExcludeList
+from core.scanner import ScanType
+from core.gui.deletion_options import DeletionOptions
+from core.gui.details_panel import DetailsPanel
+from core.gui.directory_tree import DirectoryTree
+from core.gui.ignore_list_dialog import IgnoreListDialog
+from core.gui.exclude_list_dialog import ExcludeListDialogCore
+from core.gui.problem_dialog import ProblemDialog
+from core.gui.stats_label import StatsLabel
 
 HAD_FIRST_LAUNCH_PREFERENCE = "HadFirstLaunch"
 DEBUG_MODE_PREFERENCE = "DebugMode"
@@ -124,15 +126,13 @@ class DupeGuru(Broadcaster):
 
     NAME = PROMPT_NAME = "dupeGuru"
 
-    PICTURE_CACHE_TYPE = "sqlite"  # set to 'shelve' for a ShelveCache
-
     def __init__(self, view, portable=False):
         if view.get_default(DEBUG_MODE_PREFERENCE):
             logging.getLogger().setLevel(logging.DEBUG)
             logging.debug("Debug mode enabled")
         Broadcaster.__init__(self)
         self.view = view
-        self.appdata = desktop.special_folder_path(desktop.SpecialFolder.APPDATA, appname=self.NAME, portable=portable)
+        self.appdata = desktop.special_folder_path(desktop.SpecialFolder.APPDATA, portable=portable)
         if not op.exists(self.appdata):
             os.makedirs(self.appdata)
         self.app_mode = AppMode.STANDARD
@@ -151,7 +151,8 @@ class DupeGuru(Broadcaster):
             "clean_empty_dirs": False,
             "ignore_hardlink_matches": False,
             "copymove_dest_type": DestType.RELATIVE,
-            "picture_cache_type": self.PICTURE_CACHE_TYPE,
+            "include_exists_check": True,
+            "rehash_ignore_mtime": False,
         }
         self.selected_dupes = []
         self.details_panel = DetailsPanel(self)
@@ -181,8 +182,7 @@ class DupeGuru(Broadcaster):
         self.view.create_results_window()
 
     def _get_picture_cache_path(self):
-        cache_type = self.options["picture_cache_type"]
-        cache_name = "cached_pictures.shelve" if cache_type == "shelve" else "cached_pictures.db"
+        cache_name = "cached_pictures.db"
         return op.join(self.appdata, cache_name)
 
     def _get_dupe_sort_key(self, dupe, get_group, key, delta):
@@ -248,7 +248,7 @@ class DupeGuru(Broadcaster):
             ref = group.ref
             linkfunc = os.link if use_hardlinks else os.symlink
             linkfunc(str(ref.path), str_path)
-        self.clean_empty_dirs(dupe.path.parent())
+        self.clean_empty_dirs(dupe.path.parent)
 
     def _create_file(self, path):
         # We add fs.Folder to fileclasses in case the file we're loading contains folder paths.
@@ -262,7 +262,7 @@ class DupeGuru(Broadcaster):
         try:
             f._read_all_info(attrnames=self.METADATA_TO_READ)
             return f
-        except EnvironmentError:
+        except OSError:
             return None
 
     def _get_export_data(self):
@@ -415,7 +415,7 @@ class DupeGuru(Broadcaster):
     def clean_empty_dirs(self, path):
         if self.options["clean_empty_dirs"]:
             while delete_if_empty(path, [".DS_Store"]):
-                path = path.parent()
+                path = path.parent
 
     def clear_picture_cache(self):
         try:
@@ -428,25 +428,25 @@ class DupeGuru(Broadcaster):
 
     def copy_or_move(self, dupe, copy: bool, destination: str, dest_type: DestType):
         source_path = dupe.path
-        location_path = first(p for p in self.directories if dupe.path in p)
+        location_path = first(p for p in self.directories if p in dupe.path.parents)
         dest_path = Path(destination)
         if dest_type in {DestType.RELATIVE, DestType.ABSOLUTE}:
             # no filename, no windows drive letter
-            source_base = source_path.remove_drive_letter().parent()
+            source_base = source_path.relative_to(source_path.anchor).parent
             if dest_type == DestType.RELATIVE:
-                source_base = source_base[location_path:]
-            dest_path = dest_path[source_base]
+                source_base = source_base.relative_to(location_path.relative_to(location_path.anchor))
+            dest_path = dest_path.joinpath(source_base)
         if not dest_path.exists():
-            dest_path.makedirs()
+            dest_path.mkdir(parents=True)
         # Add filename to dest_path. For file move/copy, it's not required, but for folders, yes.
-        dest_path = dest_path[source_path.name]
+        dest_path = dest_path.joinpath(source_path.name)
         logging.debug("Copy/Move operation from '%s' to '%s'", source_path, dest_path)
         # Raises an EnvironmentError if there's a problem
         if copy:
             smart_copy(source_path, dest_path)
         else:
             smart_move(source_path, dest_path)
-            self.clean_empty_dirs(source_path.parent())
+            self.clean_empty_dirs(source_path.parent)
 
     def copy_or_move_marked(self, copy):
         """Start an async move (or copy) job on marked duplicates.
@@ -553,9 +553,15 @@ class DupeGuru(Broadcaster):
                 # a workaround to make the damn thing work.
                 exepath, args = match.groups()
                 path, exename = op.split(exepath)
-                subprocess.Popen(exename + args, shell=True, cwd=path)
+                p = subprocess.Popen(
+                    exename + args, shell=True, cwd=path, stdout=subprocess.PIPE, stderr=subprocess.STDOUT
+                )
+                output = p.stdout.read()
+                logging.info("Custom command %s %s: %s", exename, args, output)
             else:
-                subprocess.Popen(dupe_cmd, shell=True)
+                p = subprocess.Popen(dupe_cmd, shell=True, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
+                output = p.stdout.read()
+                logging.info("Custom command %s: %s", dupe_cmd, output)
 
     def load(self):
         """Load directory selection and ignore list from files in appdata.
@@ -780,12 +786,13 @@ class DupeGuru(Broadcaster):
         except OSError as e:
             self.view.show_message(tr("Couldn't write to file: {}").format(str(e)))
 
-    def start_scanning(self):
+    def start_scanning(self, profile_scan=False):
         """Starts an async job to scan for duplicates.
 
         Scans folders selected in :attr:`directories` and put the results in :attr:`results`
         """
         scanner = self.SCANNER_CLASS()
+        fs.filesdb.ignore_mtime = self.options["rehash_ignore_mtime"] is True
         if not self.directories.has_any_file():
             self.view.show_message(tr("The selected directories contain no scannable file."))
             return
@@ -800,6 +807,9 @@ class DupeGuru(Broadcaster):
         self._results_changed()
 
         def do(j):
+            if profile_scan:
+                pr = cProfile.Profile()
+                pr.enable()
             j.set_progress(0, tr("Collecting files to scan"))
             if scanner.scan_type == ScanType.FOLDERS:
                 files = list(self.directories.get_folders(folderclass=se.fs.Folder, j=j))
@@ -810,6 +820,9 @@ class DupeGuru(Broadcaster):
             logging.info("Scanning %d files" % len(files))
             self.results.groups = scanner.get_dupe_groups(files, self.ignore_list, j)
             self.discarded_file_count = scanner.discarded_file_count
+            if profile_scan:
+                pr.disable()
+                pr.dump_stats(op.join(self.appdata, f"{datetime.datetime.now():%Y-%m-%d_%H-%M-%S}.profile"))
 
         self._start_job(JobType.SCAN, do)
 

core/directories.py

@@ -7,13 +7,13 @@
 import os
 from xml.etree import ElementTree as ET
 import logging
+from pathlib import Path
 
 from hscommon.jobprogress import job
-from hscommon.path import Path
 from hscommon.util import FileOrPath
 from hscommon.trans import tr
 
-from . import fs
+from core import fs
 
 __all__ = [
     "Directories",
@@ -63,7 +63,7 @@ class Directories:
 
     def __contains__(self, path):
         for p in self._dirs:
-            if path in p:
+            if path == p or p in path.parents:
                 return True
         return False
 
@@ -84,64 +84,64 @@ class Directories:
             for denied_path_re in self._exclude_list.compiled:
                 if denied_path_re.match(str(path.name)):
                     return DirectoryState.EXCLUDED
-            # return # We still use the old logic to force state on hidden dirs
+            return DirectoryState.NORMAL
         # Override this in subclasses to specify the state of some special folders.
         if path.name.startswith("."):
             return DirectoryState.EXCLUDED
+        return DirectoryState.NORMAL
 
     def _get_files(self, from_path, fileclasses, j):
-        for root, dirs, files in os.walk(str(from_path)):
-            j.check_if_cancelled()
-            root_path = Path(root)
-            state = self.get_state(root_path)
-            if state == DirectoryState.EXCLUDED and not any(p[: len(root_path)] == root_path for p in self.states):
-                # Recursively get files from folders with lots of subfolder is expensive. However, there
-                # might be a subfolder in this path that is not excluded. What we want to do is to skim
-                # through self.states and see if we must continue, or we can stop right here to save time
-                del dirs[:]
-            try:
-                if state != DirectoryState.EXCLUDED:
-                    # Old logic
-                    if self._exclude_list is None or not self._exclude_list.mark_count:
-                        found_files = [fs.get_file(root_path + f, fileclasses=fileclasses) for f in files]
-                    else:
-                        found_files = []
-                        # print(f"len of files: {len(files)} {files}")
-                        for f in files:
-                            if not self._exclude_list.is_excluded(root, f):
-                                found_files.append(fs.get_file(root_path + f, fileclasses=fileclasses))
-                    found_files = [f for f in found_files if f is not None]
-                    # In some cases, directories can be considered as files by dupeGuru, which is
-                    # why we have this line below. In fact, there only one case: Bundle files under
-                    # OS X... In other situations, this forloop will do nothing.
-                    for d in dirs[:]:
-                        f = fs.get_file(root_path + d, fileclasses=fileclasses)
-                        if f is not None:
-                            found_files.append(f)
-                            dirs.remove(d)
-                    logging.debug(
-                        "Collected %d files in folder %s",
-                        len(found_files),
-                        str(root_path),
-                    )
-                    for file in found_files:
-                        file.is_ref = state == DirectoryState.REFERENCE
-                        yield file
-            except (EnvironmentError, fs.InvalidPath):
-                pass
+        try:
+            with os.scandir(from_path) as iter:
+                root_path = Path(from_path)
+                state = self.get_state(root_path)
+                # if we have no un-excluded dirs under this directory skip going deeper
+                skip_dirs = state == DirectoryState.EXCLUDED and not any(
+                    p.parts[: len(root_path.parts)] == root_path.parts for p in self.states
+                )
+                count = 0
+                for item in iter:
+                    j.check_if_cancelled()
+                    try:
+                        if item.is_dir():
+                            if skip_dirs:
+                                continue
+                            yield from self._get_files(item.path, fileclasses, j)
+                            continue
+                        elif state == DirectoryState.EXCLUDED:
+                            continue
+                        # File excluding or not
+                        if (
+                            self._exclude_list is None
+                            or not self._exclude_list.mark_count
+                            or not self._exclude_list.is_excluded(str(from_path), item.name)
+                        ):
+                            file = fs.get_file(item, fileclasses=fileclasses)
+                            if file:
+                                file.is_ref = state == DirectoryState.REFERENCE
+                                count += 1
+                                yield file
+                    except (OSError, fs.InvalidPath):
+                        pass
+                logging.debug(
+                    "Collected %d files in folder %s",
+                    count,
+                    str(root_path),
+                )
+        except OSError:
+            pass
 
     def _get_folders(self, from_folder, j):
         j.check_if_cancelled()
         try:
             for subfolder in from_folder.subfolders:
-                for folder in self._get_folders(subfolder, j):
-                    yield folder
+                yield from self._get_folders(subfolder, j)
             state = self.get_state(from_folder.path)
             if state != DirectoryState.EXCLUDED:
                 from_folder.is_ref = state == DirectoryState.REFERENCE
                 logging.debug("Yielding Folder %r state: %d", from_folder, state)
                 yield from_folder
-        except (EnvironmentError, fs.InvalidPath):
+        except (OSError, fs.InvalidPath):
             pass
 
     # ---Public
@@ -159,7 +159,7 @@ class Directories:
             raise AlreadyThereError()
         if not path.exists():
             raise InvalidPathError()
-        self._dirs = [p for p in self._dirs if p not in path]
+        self._dirs = [p for p in self._dirs if path not in p.parents]
         self._dirs.append(path)
 
     @staticmethod
@@ -170,10 +170,10 @@ class Directories:
         :rtype: list of Path
         """
         try:
-            subpaths = [p for p in path.listdir() if p.isdir()]
+            subpaths = [p for p in path.glob("*") if p.is_dir()]
             subpaths.sort(key=lambda x: x.name.lower())
             return subpaths
-        except EnvironmentError:
+        except OSError:
             return []
 
     def get_files(self, fileclasses=None, j=job.nulljob):
@@ -187,7 +187,7 @@ class Directories:
         for path in self._dirs:
             for file in self._get_files(path, fileclasses=fileclasses, j=j):
                 file_count += 1
-                if type(j) != job.NullJob:
+                if not isinstance(j, job.NullJob):
                     j.set_progress(-1, tr("Collected {} files to scan").format(file_count))
                 yield file
 
@@ -203,7 +203,7 @@ class Directories:
             from_folder = folderclass(path)
             for folder in self._get_folders(from_folder, j):
                 folder_count += 1
-                if type(j) != job.NullJob:
+                if not isinstance(j, job.NullJob):
                     j.set_progress(-1, tr("Collected {} folders to scan").format(folder_count))
                 yield folder
 
@@ -215,19 +215,16 @@ class Directories:
         # direct match? easy result.
         if path in self.states:
             return self.states[path]
-        state = self._default_state_for_path(path) or DirectoryState.NORMAL
+        state = self._default_state_for_path(path)
         # Save non-default states in cache, necessary for _get_files()
         if state != DirectoryState.NORMAL:
             self.states[path] = state
             return state
-
-        prevlen = 0
-        # we loop through the states to find the longest matching prefix
-        # if the parent has a state in cache, return that state
-        for p, s in self.states.items():
-            if p.is_parent_of(path) and len(p) > prevlen:
-                prevlen = len(p)
-                state = s
+        # find the longest parent path that is in states and return that state if found
+        # NOTE: path.parents is ordered longest to shortest
+        for parent_path in path.parents:
+            if parent_path in self.states:
+                return self.states[parent_path]
         return state
 
     def has_any_file(self):
@@ -296,6 +293,6 @@ class Directories:
         if self.get_state(path) == state:
             return
         for iter_path in list(self.states.keys()):
-            if path.is_parent_of(iter_path):
+            if path in iter_path.parents:
                 del self.states[iter_path]
         self.states[path] = state

core/engine.py

@@ -166,7 +166,7 @@ def reduce_common_words(word_dict, threshold):
     The exception to this removal are the objects where all the words of the object are common.
     Because if we remove them, we will miss some duplicates!
     """
-    uncommon_words = set(word for word, objects in word_dict.items() if len(objects) < threshold)
+    uncommon_words = {word for word, objects in word_dict.items() if len(objects) < threshold}
     for word, objects in list(word_dict.items()):
         if len(objects) < threshold:
             continue
@@ -283,7 +283,7 @@ def getmatches_by_contents(files, bigsize=0, j=job.nulljob):
     """Returns a list of :class:`Match` within ``files`` if their contents is the same.
 
     :param bigsize: The size in bytes over which we consider files big enough to
-                    justify taking samples of md5. If 0, compute md5 as usual.
+                    justify taking samples of the file for hashing. If 0, compute digest as usual.
     :param j: A :ref:`job progress instance <jobs>`.
     """
     size2files = defaultdict(set)
@@ -300,15 +300,16 @@ def getmatches_by_contents(files, bigsize=0, j=job.nulljob):
             if first.is_ref and second.is_ref:
                 continue  # Don't spend time comparing two ref pics together.
             if first.size == 0 and second.size == 0:
-                # skip md5 for zero length files
+                # skip hashing for zero length files
                 result.append(Match(first, second, 100))
                 continue
-            if first.md5partial == second.md5partial:
+            # if digests are the same (and not None) then files match
+            if first.digest_partial is not None and first.digest_partial == second.digest_partial:
                 if bigsize > 0 and first.size > bigsize:
-                    if first.md5samples == second.md5samples:
+                    if first.digest_samples is not None and first.digest_samples == second.digest_samples:
                         result.append(Match(first, second, 100))
                 else:
-                    if first.md5 == second.md5:
+                    if first.digest is not None and first.digest == second.digest:
                         result.append(Match(first, second, 100))
         group_count += 1
         j.add_progress(desc=PROGRESS_MESSAGE % (len(result), group_count))
@@ -409,7 +410,7 @@ class Group:
 
         You can call this after the duplicate scanning process to free a bit of memory.
         """
-        discarded = set(m for m in self.matches if not all(obj in self.unordered for obj in [m.first, m.second]))
+        discarded = {m for m in self.matches if not all(obj in self.unordered for obj in [m.first, m.second])}
         self.matches -= discarded
         self.candidates = defaultdict(set)
         return discarded
@@ -456,7 +457,7 @@ class Group:
             self._matches_for_ref = None
             if (len(self) > 1) and any(not getattr(item, "is_ref", False) for item in self):
                 if discard_matches:
-                    self.matches = set(m for m in self.matches if item not in m)
+                    self.matches = {m for m in self.matches if item not in m}
             else:
                 self._clear()
         except ValueError:
@@ -529,7 +530,7 @@ def get_groups(matches):
         del dupe2group
         del matches
         # should free enough memory to continue
-        logging.warning("Memory Overflow. Groups: {0}".format(len(groups)))
+        logging.warning(f"Memory Overflow. Groups: {len(groups)}")
     # Now that we have a group, we have to discard groups' matches and see if there're any "orphan"
     # matches, that is, matches that were candidate in a group but that none of their 2 files were
     # accepted in the group. With these orphan groups, it's safe to build additional groups

core/exclude.py

@@ -2,7 +2,7 @@
 # which should be included with this package. The terms are also available at
 # http://www.gnu.org/licenses/gpl-3.0.html
 
-from .markable import Markable
+from core.markable import Markable
 from xml.etree import ElementTree as ET
 
 # TODO: perhaps use regex module for better Unicode support? https://pypi.org/project/regex/

core/fs.py

@@ -11,16 +11,28 @@
 # resulting needless complexity and memory usage. It's been a while since I wanted to do that fork,
 # and I'm doing it now.
 
-import hashlib
+import os
+
 from math import floor
 import logging
 import sqlite3
+from sys import platform
 from threading import Lock
-from typing import Any
+from typing import Any, AnyStr, Union, Callable
 
-from hscommon.path import Path
+from pathlib import Path
 from hscommon.util import nonone, get_file_ext
 
+hasher: Callable
+try:
+    import xxhash
+
+    hasher = xxhash.xxh128
+except ImportError:
+    import hashlib
+
+    hasher = hashlib.md5
+
 __all__ = [
     "File",
     "Folder",
@@ -40,9 +52,12 @@ NOT_SET = object()
 # CPU.
 CHUNK_SIZE = 1024 * 1024  # 1 MiB
 
-# Minimum size below which partial hashes don't need to be computed
+# Minimum size below which partial hashing is not used
 MIN_FILE_SIZE = 3 * CHUNK_SIZE  # 3MiB, because we take 3 samples
 
+# Partial hashing offset and size
+PARTIAL_OFFSET_SIZE = (0x4000, 0x4000)
+
 
 class FSError(Exception):
     cls_message = "An error has occured on '{name}' in '{parent}'"
@@ -83,75 +98,101 @@ class OperationError(FSError):
 
 
 class FilesDB:
+    schema_version = 1
+    schema_version_description = "Changed from md5 to xxhash if available."
 
-    create_table_query = "CREATE TABLE IF NOT EXISTS files (path TEXT PRIMARY KEY, size INTEGER, mtime_ns INTEGER, entry_dt DATETIME, md5 BLOB, md5partial BLOB)"
-    drop_table_query = "DROP TABLE files;"
+    create_table_query = """CREATE TABLE IF NOT EXISTS files (path TEXT PRIMARY KEY, size INTEGER, mtime_ns INTEGER,
+        entry_dt DATETIME, digest BLOB, digest_partial BLOB, digest_samples BLOB)"""
+    drop_table_query = "DROP TABLE IF EXISTS files;"
     select_query = "SELECT {key} FROM files WHERE path=:path AND size=:size and mtime_ns=:mtime_ns"
+    select_query_ignore_mtime = "SELECT {key} FROM files WHERE path=:path AND size=:size"
     insert_query = """
-        INSERT INTO files (path, size, mtime_ns, entry_dt, {key}) VALUES (:path, :size, :mtime_ns, datetime('now'), :value)
+        INSERT INTO files (path, size, mtime_ns, entry_dt, {key})
+        VALUES (:path, :size, :mtime_ns, datetime('now'), :value)
         ON CONFLICT(path) DO UPDATE SET size=:size, mtime_ns=:mtime_ns, entry_dt=datetime('now'), {key}=:value;
     """
 
+    ignore_mtime = False
+
     def __init__(self):
         self.conn = None
-        self.cur = None
         self.lock = None
 
-    def connect(self, path):
-        # type: (str, ) -> None
-
-        self.conn = sqlite3.connect(path, check_same_thread=False)
-        self.cur = self.conn.cursor()
-        self.cur.execute(self.create_table_query)
+    def connect(self, path: Union[AnyStr, os.PathLike]) -> None:
+        if platform.startswith("gnu0"):
+            self.conn = sqlite3.connect(path, check_same_thread=False, isolation_level=None)
+        else:
+            self.conn = sqlite3.connect(path, check_same_thread=False)
         self.lock = Lock()
+        self._check_upgrade()
 
-    def clear(self):
-        # type: () -> None
+    def _check_upgrade(self) -> None:
+        with self.lock, self.conn as conn:
+            has_schema = conn.execute(
+                "SELECT NAME FROM sqlite_master WHERE type='table' AND name='schema_version'"
+            ).fetchall()
+            version = None
+            if has_schema:
+                version = conn.execute("SELECT version FROM schema_version ORDER BY version DESC").fetchone()[0]
+            else:
+                conn.execute("CREATE TABLE schema_version (version int PRIMARY KEY, description TEXT)")
+            if version != self.schema_version:
+                conn.execute(self.drop_table_query)
+                conn.execute(
+                    "INSERT OR REPLACE INTO schema_version VALUES (:version, :description)",
+                    {"version": self.schema_version, "description": self.schema_version_description},
+                )
+            conn.execute(self.create_table_query)
 
-        with self.lock:
-            self.cur.execute(self.drop_table_query)
-            self.cur.execute(self.create_table_query)
-
-    def get(self, path, key):
-        # type: (Path, str) -> bytes
+    def clear(self) -> None:
+        with self.lock, self.conn as conn:
+            conn.execute(self.drop_table_query)
+            conn.execute(self.create_table_query)
 
+    def get(self, path: Path, key: str) -> Union[bytes, None]:
         stat = path.stat()
         size = stat.st_size
         mtime_ns = stat.st_mtime_ns
+        try:
+            with self.conn as conn:
+                if self.ignore_mtime:
+                    cursor = conn.execute(
+                        self.select_query_ignore_mtime.format(key=key), {"path": str(path), "size": size}
+                    )
+                else:
+                    cursor = conn.execute(
+                        self.select_query.format(key=key),
+                        {"path": str(path), "size": size, "mtime_ns": mtime_ns},
+                    )
+                result = cursor.fetchone()
+                cursor.close()
 
-        with self.lock:
-            self.cur.execute(self.select_query.format(key=key), {"path": str(path), "size": size, "mtime_ns": mtime_ns})
-            result = self.cur.fetchone()
-
-        if result:
-            return result[0]
+            if result:
+                return result[0]
+        except Exception as ex:
+            logging.warning(f"Couldn't get {key} for {path} w/{size}, {mtime_ns}: {ex}")
 
         return None
 
-    def put(self, path, key, value):
-        # type: (Path, str, Any) -> None
-
+    def put(self, path: Path, key: str, value: Any) -> None:
         stat = path.stat()
         size = stat.st_size
         mtime_ns = stat.st_mtime_ns
+        try:
+            with self.lock, self.conn as conn:
+                conn.execute(
+                    self.insert_query.format(key=key),
+                    {"path": str(path), "size": size, "mtime_ns": mtime_ns, "value": value},
+                )
+        except Exception as ex:
+            logging.warning(f"Couldn't put {key} for {path} w/{size}, {mtime_ns}: {ex}")
 
-        with self.lock:
-            self.cur.execute(
-                self.insert_query.format(key=key),
-                {"path": str(path), "size": size, "mtime_ns": mtime_ns, "value": value},
-            )
-
-    def commit(self):
-        # type: () -> None
-
+    def commit(self) -> None:
         with self.lock:
             self.conn.commit()
 
-    def close(self):
-        # type: () -> None
-
+    def close(self) -> None:
         with self.lock:
-            self.cur.close()
             self.conn.close()
 
 
@@ -161,19 +202,26 @@ filesdb = FilesDB()  # Singleton
 class File:
     """Represents a file and holds metadata to be used for scanning."""
 
-    INITIAL_INFO = {"size": 0, "mtime": 0, "md5": b"", "md5partial": b"", "md5samples": b""}
+    INITIAL_INFO = {"size": 0, "mtime": 0, "digest": b"", "digest_partial": b"", "digest_samples": b""}
     # Slots for File make us save quite a bit of memory. In a memory test I've made with a lot of
     # files, I saved 35% memory usage with "unread" files (no _read_info() call) and gains become
     # even greater when we take into account read attributes (70%!). Yeah, it's worth it.
-    __slots__ = ("path", "is_ref", "words") + tuple(INITIAL_INFO.keys())
+    __slots__ = ("path", "unicode_path", "is_ref", "words") + tuple(INITIAL_INFO.keys())
 
     def __init__(self, path):
-        self.path = path
         for attrname in self.INITIAL_INFO:
             setattr(self, attrname, NOT_SET)
+        if type(path) is os.DirEntry:
+            self.path = Path(path.path)
+            self.size = nonone(path.stat().st_size, 0)
+            self.mtime = nonone(path.stat().st_mtime, 0)
+        else:
+            self.path = path
+        if self.path:
+            self.unicode_path = str(self.path)
 
     def __repr__(self):
-        return "<{} {}>".format(self.__class__.__name__, str(self.path))
+        return f"<{self.__class__.__name__} {str(self.path)}>"
 
     def __getattribute__(self, attrname):
         result = object.__getattribute__(self, attrname)
@@ -187,32 +235,46 @@ class File:
                 result = self.INITIAL_INFO[attrname]
         return result
 
-    def _calc_md5(self):
+    def _calc_digest(self):
         # type: () -> bytes
 
         with self.path.open("rb") as fp:
-            md5 = hashlib.md5()
+            file_hash = hasher()
             # The goal here is to not run out of memory on really big files. However, the chunk
             # size has to be large enough so that the python loop isn't too costly in terms of
             # CPU.
             CHUNK_SIZE = 1024 * 1024  # 1 mb
             filedata = fp.read(CHUNK_SIZE)
             while filedata:
-                md5.update(filedata)
+                file_hash.update(filedata)
                 filedata = fp.read(CHUNK_SIZE)
-            return md5.digest()
+            return file_hash.digest()
 
-    def _calc_md5partial(self):
+    def _calc_digest_partial(self):
         # type: () -> bytes
-
-        # This offset is where we should start reading the file to get a partial md5
-        # For audio file, it should be where audio data starts
-        offset, size = (0x4000, 0x4000)
-
         with self.path.open("rb") as fp:
-            fp.seek(offset)
-            partialdata = fp.read(size)
-            return hashlib.md5(partialdata).digest()
+            fp.seek(PARTIAL_OFFSET_SIZE[0])
+            partial_data = fp.read(PARTIAL_OFFSET_SIZE[1])
+            return hasher(partial_data).digest()
+
+    def _calc_digest_samples(self) -> bytes:
+        size = self.size
+        with self.path.open("rb") as fp:
+            # Chunk at 25% of the file
+            fp.seek(floor(size * 25 / 100), 0)
+            file_data = fp.read(CHUNK_SIZE)
+            file_hash = hasher(file_data)
+
+            # Chunk at 60% of the file
+            fp.seek(floor(size * 60 / 100), 0)
+            file_data = fp.read(CHUNK_SIZE)
+            file_hash.update(file_data)
+
+            # Last chunk of the file
+            fp.seek(-CHUNK_SIZE, 2)
+            file_data = fp.read(CHUNK_SIZE)
+            file_hash.update(file_data)
+            return file_hash.digest()
 
     def _read_info(self, field):
         # print(f"_read_info({field}) for {self}")
@@ -220,48 +282,30 @@ class File:
             stats = self.path.stat()
             self.size = nonone(stats.st_size, 0)
             self.mtime = nonone(stats.st_mtime, 0)
-        elif field == "md5partial":
-            try:
-                self.md5partial = filesdb.get(self.path, "md5partial")
-                if self.md5partial is None:
-                    self.md5partial = self._calc_md5partial()
-                    filesdb.put(self.path, "md5partial", self.md5partial)
-            except Exception as e:
-                logging.warning("Couldn't get md5partial for %s: %s", self.path, e)
-        elif field == "md5":
-            try:
-                self.md5 = filesdb.get(self.path, "md5")
-                if self.md5 is None:
-                    self.md5 = self._calc_md5()
-                    filesdb.put(self.path, "md5", self.md5)
-            except Exception as e:
-                logging.warning("Couldn't get md5 for %s: %s", self.path, e)
-        elif field == "md5samples":
-            try:
-                with self.path.open("rb") as fp:
-                    size = self.size
-                    # Might as well hash such small files entirely.
-                    if size <= MIN_FILE_SIZE:
-                        setattr(self, field, self.md5)
-                        return
-
-                    # Chunk at 25% of the file
-                    fp.seek(floor(size * 25 / 100), 0)
-                    filedata = fp.read(CHUNK_SIZE)
-                    md5 = hashlib.md5(filedata)
-
-                    # Chunk at 60% of the file
-                    fp.seek(floor(size * 60 / 100), 0)
-                    filedata = fp.read(CHUNK_SIZE)
-                    md5.update(filedata)
-
-                    # Last chunk of the file
-                    fp.seek(-CHUNK_SIZE, 2)
-                    filedata = fp.read(CHUNK_SIZE)
-                    md5.update(filedata)
-                    setattr(self, field, md5.digest())
-            except Exception as e:
-                logging.error(f"Error computing md5samples: {e}")
+        elif field == "digest_partial":
+            self.digest_partial = filesdb.get(self.path, "digest_partial")
+            if self.digest_partial is None:
+                # If file is smaller than partial requirements just use the full digest
+                if self.size < PARTIAL_OFFSET_SIZE[0] + PARTIAL_OFFSET_SIZE[1]:
+                    self.digest_partial = self.digest
+                else:
+                    self.digest_partial = self._calc_digest_partial()
+                filesdb.put(self.path, "digest_partial", self.digest_partial)
+        elif field == "digest":
+            self.digest = filesdb.get(self.path, "digest")
+            if self.digest is None:
+                self.digest = self._calc_digest()
+                filesdb.put(self.path, "digest", self.digest)
+        elif field == "digest_samples":
+            size = self.size
+            # Might as well hash such small files entirely.
+            if size <= MIN_FILE_SIZE:
+                self.digest_samples = self.digest
+                return
+            self.digest_samples = filesdb.get(self.path, "digest_samples")
+            if self.digest_samples is None:
+                self.digest_samples = self._calc_digest_samples()
+                filesdb.put(self.path, "digest_samples", self.digest_samples)
 
     def _read_all_info(self, attrnames=None):
         """Cache all possible info.
@@ -277,17 +321,25 @@ class File:
     @classmethod
     def can_handle(cls, path):
         """Returns whether this file wrapper class can handle ``path``."""
-        return not path.islink() and path.isfile()
+        return not path.is_symlink() and path.is_file()
+
+    def exists(self) -> bool:
+        """Safely check if the underlying file exists, treat error as non-existent"""
+        try:
+            return self.path.exists()
+        except OSError as ex:
+            logging.warning(f"Checking {self.path} raised: {ex}")
+            return False
 
     def rename(self, newname):
         if newname == self.name:
             return
-        destpath = self.path.parent()[newname]
+        destpath = self.path.parent.joinpath(newname)
         if destpath.exists():
-            raise AlreadyExistsError(newname, self.path.parent())
+            raise AlreadyExistsError(newname, self.path.parent)
         try:
             self.path.rename(destpath)
-        except EnvironmentError:
+        except OSError:
             raise OperationError(self)
         if not destpath.exists():
             raise OperationError(self)
@@ -308,19 +360,20 @@ class File:
 
     @property
     def folder_path(self):
-        return self.path.parent()
+        return self.path.parent
 
 
 class Folder(File):
     """A wrapper around a folder path.
 
-    It has the size/md5 info of a File, but its value is the sum of its subitems.
+    It has the size/digest info of a File, but its value is the sum of its subitems.
     """
 
     __slots__ = File.__slots__ + ("_subfolders",)
 
     def __init__(self, path):
         File.__init__(self, path)
+        self.size = NOT_SET
         self._subfolders = None
 
     def _all_items(self):
@@ -335,31 +388,31 @@ class Folder(File):
             self.size = size
             stats = self.path.stat()
             self.mtime = nonone(stats.st_mtime, 0)
-        elif field in {"md5", "md5partial", "md5samples"}:
+        elif field in {"digest", "digest_partial", "digest_samples"}:
             # What's sensitive here is that we must make sure that subfiles'
-            # md5 are always added up in the same order, but we also want a
-            # different md5 if a file gets moved in a different subdirectory.
+            # digest are always added up in the same order, but we also want a
+            # different digest if a file gets moved in a different subdirectory.
 
-            def get_dir_md5_concat():
+            def get_dir_digest_concat():
                 items = self._all_items()
                 items.sort(key=lambda f: f.path)
-                md5s = [getattr(f, field) for f in items]
-                return b"".join(md5s)
+                digests = [getattr(f, field) for f in items]
+                return b"".join(digests)
 
-            md5 = hashlib.md5(get_dir_md5_concat())
-            digest = md5.digest()
+            digest = hasher(get_dir_digest_concat()).digest()
             setattr(self, field, digest)
 
     @property
     def subfolders(self):
         if self._subfolders is None:
-            subfolders = [p for p in self.path.listdir() if not p.islink() and p.isdir()]
+            with os.scandir(self.path) as iter:
+                subfolders = [p for p in iter if not p.is_symlink() and p.is_dir()]
             self._subfolders = [self.__class__(p) for p in subfolders]
         return self._subfolders
 
     @classmethod
     def can_handle(cls, path):
-        return not path.islink() and path.isdir()
+        return not path.is_symlink() and path.is_dir()
 
 
 def get_file(path, fileclasses=[File]):
@@ -384,10 +437,11 @@ def get_files(path, fileclasses=[File]):
     assert all(issubclass(fileclass, File) for fileclass in fileclasses)
     try:
         result = []
-        for path in path.listdir():
-            file = get_file(path, fileclasses=fileclasses)
-            if file is not None:
-                result.append(file)
+        with os.scandir(path) as iter:
+            for item in iter:
+                file = get_file(item, fileclasses=fileclasses)
+                if file is not None:
+                    result.append(file)
         return result
-    except EnvironmentError:
+    except OSError:
         raise InvalidPath(path)

core/gui/details_panel.py

@@ -7,7 +7,7 @@
 # http://www.gnu.org/licenses/gpl-3.0.html
 
 from hscommon.gui.base import GUIObject
-from .base import DupeGuruGUIObject
+from core.gui.base import DupeGuruGUIObject
 
 
 class DetailsPanel(GUIObject, DupeGuruGUIObject):

core/gui/directory_tree.py

@@ -8,8 +8,8 @@
 
 from hscommon.gui.tree import Tree, Node
 
-from ..directories import DirectoryState
-from .base import DupeGuruGUIObject
+from core.directories import DirectoryState
+from core.gui.base import DupeGuruGUIObject
 
 STATE_ORDER = [DirectoryState.NORMAL, DirectoryState.REFERENCE, DirectoryState.EXCLUDED]
 

core/gui/exclude_list_dialog.py

@@ -5,7 +5,7 @@
 # which should be included with this package. The terms are also available at
 # http://www.gnu.org/licenses/gpl-3.0.html
 
-from .exclude_list_table import ExcludeListTable
+from core.gui.exclude_list_table import ExcludeListTable
 from core.exclude import has_sep
 from os import sep
 import logging

core/gui/exclude_list_table.py

@@ -2,7 +2,7 @@
 # which should be included with this package. The terms are also available at
 # http://www.gnu.org/licenses/gpl-3.0.html
 
-from .base import DupeGuruGUIObject
+from core.gui.base import DupeGuruGUIObject
 from hscommon.gui.table import GUITable, Row
 from hscommon.gui.column import Column, Columns
 from hscommon.trans import trget

core/gui/ignore_list_dialog.py

@@ -6,7 +6,7 @@
 # http://www.gnu.org/licenses/gpl-3.0.html
 
 from hscommon.trans import tr
-from .ignore_list_table import IgnoreListTable
+from core.gui.ignore_list_table import IgnoreListTable
 
 
 class IgnoreListDialog:

core/gui/problem_dialog.py

@@ -8,7 +8,7 @@
 
 from hscommon import desktop
 
-from .problem_table import ProblemTable
+from core.gui.problem_table import ProblemTable
 
 
 class ProblemDialog:

core/gui/result_table.py

@@ -11,7 +11,7 @@ from operator import attrgetter
 from hscommon.gui.table import GUITable, Row
 from hscommon.gui.column import Columns
 
-from .base import DupeGuruGUIObject
+from core.gui.base import DupeGuruGUIObject
 
 
 class DupeRow(Row):

core/gui/stats_label.py

@@ -6,7 +6,7 @@
 # which should be included with this package. The terms are also available at
 # http://www.gnu.org/licenses/gpl-3.0.html
 
-from .base import DupeGuruGUIObject
+from core.gui.base import DupeGuruGUIObject
 
 
 class StatsLabel(DupeGuruGUIObject):

core/me/__init__.py

@@ -1 +1 @@
-from . import fs, prioritize, result_table, scanner  # noqa
+from core.me import fs, prioritize, result_table, scanner  # noqa

core/me/fs.py

@@ -97,11 +97,6 @@ class MusicFile(fs.File):
             "dupe_count": format_dupe_count(dupe_count),
         }
 
-    def _get_md5partial_offset_and_size(self):
-        # No longer calculating the offset and audio size, just whole file
-        size = self.path.stat().st_size
-        return (0, size)
-
     def _read_info(self, field):
         fs.File._read_info(self, field)
         if field in TAG_FIELDS:

core/pe/__init__.py

@@ -1,8 +1,7 @@
-from . import (  # noqa
+from core.pe import (  # noqa
     block,
     cache,
     exif,
-    iphoto_plist,
     matchblock,
     matchexif,
     photo,

core/pe/block.py

@@ -6,7 +6,7 @@
 # which should be included with this package. The terms are also available at
 # http://www.gnu.org/licenses/gpl-3.0.html
 
-from ._block import NoBlocksError, DifferentBlockCountError, avgdiff, getblocks2  # NOQA
+from core.pe._block import NoBlocksError, DifferentBlockCountError, avgdiff, getblocks2  # NOQA
 
 # Converted to C
 # def getblock(image):

core/pe/block.pyi

@@ -0,0 +1,13 @@
+from typing import Tuple, List, Union, Sequence
+
+_block = Tuple[int, int, int]
+
+class NoBlocksError(Exception): ...  # noqa: E302, E701
+class DifferentBlockCountError(Exception): ...  # noqa E701
+
+def getblock(image: object) -> Union[_block, None]: ...  # noqa: E302
+def getblocks2(image: object, block_count_per_side: int) -> Union[List[_block], None]: ...
+def diff(first: _block, second: _block) -> int: ...
+def avgdiff(  # noqa: E302
+    first: Sequence[_block], second: Sequence[_block], limit: int = 768, min_iterations: int = 1
+) -> Union[int, None]: ...

core/pe/cache.py

@@ -4,24 +4,13 @@
 # which should be included with this package. The terms are also available at
 # http://www.gnu.org/licenses/gpl-3.0.html
 
-from ._cache import string_to_colors  # noqa
+from core.pe._cache import bytes_to_colors  # noqa
 
 
-def colors_to_string(colors):
-    """Transform the 3 sized tuples 'colors' into a hex string.
+def colors_to_bytes(colors):
+    """Transform the 3 sized tuples 'colors' into a bytes string.
 
-    [(0,100,255)] --> 0064ff
-    [(1,2,3),(4,5,6)] --> 010203040506
+    [(0,100,255)] --> b'\x00d\xff'
+    [(1,2,3),(4,5,6)] --> b'\x01\x02\x03\x04\x05\x06'
     """
-    return "".join("%02x%02x%02x" % (r, g, b) for r, g, b in colors)
-
-
-# This function is an important bottleneck of dupeGuru PE. It has been converted to C.
-# def string_to_colors(s):
-#     """Transform the string 's' in a list of 3 sized tuples.
-#     """
-#     result = []
-#     for i in xrange(0, len(s), 6):
-#         number = int(s[i:i+6], 16)
-#         result.append((number >> 16, (number >> 8) & 0xff, number & 0xff))
-#     return result
+    return b"".join(map(bytes, colors))

core/pe/cache.pyi

@@ -0,0 +1,6 @@
+from typing import Union, Tuple, List
+
+_block = Tuple[int, int, int]
+
+def colors_to_bytes(colors: List[_block]) -> bytes: ...  # noqa: E302
+def bytes_to_colors(s: bytes) -> Union[List[_block], None]: ...

core/pe/cache_shelve.py

@@ -1,141 +0,0 @@
-# Copyright 2016 Virgil Dupras
-#
-# This software is licensed under the "GPLv3" License as described in the "LICENSE" file,
-# which should be included with this package. The terms are also available at
-# http://www.gnu.org/licenses/gpl-3.0.html
-
-import os
-import os.path as op
-import shelve
-import tempfile
-from collections import namedtuple
-
-from .cache import string_to_colors, colors_to_string
-
-
-def wrap_path(path):
-    return "path:{}".format(path)
-
-
-def unwrap_path(key):
-    return key[5:]
-
-
-def wrap_id(path):
-    return "id:{}".format(path)
-
-
-def unwrap_id(key):
-    return int(key[3:])
-
-
-CacheRow = namedtuple("CacheRow", "id path blocks mtime")
-
-
-class ShelveCache:
-    """A class to cache picture blocks in a shelve backend."""
-
-    def __init__(self, db=None, readonly=False):
-        self.istmp = db is None
-        if self.istmp:
-            self.dtmp = tempfile.mkdtemp()
-            self.ftmp = db = op.join(self.dtmp, "tmpdb")
-        flag = "r" if readonly else "c"
-        self.shelve = shelve.open(db, flag)
-        self.maxid = self._compute_maxid()
-
-    def __contains__(self, key):
-        return wrap_path(key) in self.shelve
-
-    def __delitem__(self, key):
-        row = self.shelve[wrap_path(key)]
-        del self.shelve[wrap_path(key)]
-        del self.shelve[wrap_id(row.id)]
-
-    def __getitem__(self, key):
-        if isinstance(key, int):
-            skey = self.shelve[wrap_id(key)]
-        else:
-            skey = wrap_path(key)
-        return string_to_colors(self.shelve[skey].blocks)
-
-    def __iter__(self):
-        return (unwrap_path(k) for k in self.shelve if k.startswith("path:"))
-
-    def __len__(self):
-        return sum(1 for k in self.shelve if k.startswith("path:"))
-
-    def __setitem__(self, path_str, blocks):
-        blocks = colors_to_string(blocks)
-        if op.exists(path_str):
-            mtime = int(os.stat(path_str).st_mtime)
-        else:
-            mtime = 0
-        if path_str in self:
-            rowid = self.shelve[wrap_path(path_str)].id
-        else:
-            rowid = self._get_new_id()
-        row = CacheRow(rowid, path_str, blocks, mtime)
-        self.shelve[wrap_path(path_str)] = row
-        self.shelve[wrap_id(rowid)] = wrap_path(path_str)
-
-    def _compute_maxid(self):
-        return max((unwrap_id(k) for k in self.shelve if k.startswith("id:")), default=1)
-
-    def _get_new_id(self):
-        self.maxid += 1
-        return self.maxid
-
-    def clear(self):
-        self.shelve.clear()
-
-    def close(self):
-        if self.shelve is not None:
-            self.shelve.close()
-            if self.istmp:
-                os.remove(self.ftmp)
-                os.rmdir(self.dtmp)
-        self.shelve = None
-
-    def filter(self, func):
-        to_delete = [key for key in self if not func(key)]
-        for key in to_delete:
-            del self[key]
-
-    def get_id(self, path):
-        if path in self:
-            return self.shelve[wrap_path(path)].id
-        else:
-            raise ValueError(path)
-
-    def get_multiple(self, rowids):
-        for rowid in rowids:
-            try:
-                skey = self.shelve[wrap_id(rowid)]
-            except KeyError:
-                continue
-            yield (rowid, string_to_colors(self.shelve[skey].blocks))
-
-    def purge_outdated(self):
-        """Go through the cache and purge outdated records.
-
-        A record is outdated if the picture doesn't exist or if its mtime is greater than the one in
-        the db.
-        """
-        todelete = []
-        for path in self:
-            row = self.shelve[wrap_path(path)]
-            if row.mtime and op.exists(path):
-                picture_mtime = os.stat(path).st_mtime
-                if int(picture_mtime) <= row.mtime:
-                    # not outdated
-                    continue
-            todelete.append(path)
-        for path in todelete:
-            try:
-                del self[path]
-            except KeyError:
-                # I have no idea why a KeyError sometimes happen, but it does, as we can see in
-                # #402 and #439. I don't think it hurts to silently ignore the error, so that's
-                # what we do
-                pass

core/pe/cache_sqlite.py

@@ -9,12 +9,24 @@ import os.path as op
 import logging
 import sqlite3 as sqlite
 
-from .cache import string_to_colors, colors_to_string
+from core.pe.cache import bytes_to_colors, colors_to_bytes
 
 
 class SqliteCache:
     """A class to cache picture blocks in a sqlite backend."""
 
+    schema_version = 2
+    schema_version_description = "Added blocks for all 8 orientations."
+
+    create_table_query = (
+        "CREATE TABLE IF NOT EXISTS "
+        "pictures(path TEXT, mtime_ns INTEGER, blocks BLOB, blocks2 BLOB, blocks3 BLOB, "
+        "blocks4 BLOB, blocks5 BLOB, blocks6 BLOB, blocks7 BLOB, blocks8 BLOB)"
+    )
+    create_index_query = "CREATE INDEX IF NOT EXISTS idx_path on pictures (path)"
+    drop_table_query = "DROP TABLE IF EXISTS pictures"
+    drop_index_query = "DROP INDEX IF EXISTS idx_path"
+
     def __init__(self, db=":memory:", readonly=False):
         # readonly is not used in the sqlite version of the cache
         self.dbname = db
@@ -35,12 +47,20 @@ class SqliteCache:
     # Optimized
     def __getitem__(self, key):
         if isinstance(key, int):
-            sql = "select blocks from pictures where rowid = ?"
+            sql = (
+                "select blocks, blocks2, blocks3, blocks4, blocks5, blocks6, blocks7, blocks8 "
+                "from pictures "
+                "where rowid = ?"
+            )
         else:
-            sql = "select blocks from pictures where path = ?"
-        result = self.con.execute(sql, [key]).fetchone()
-        if result:
-            result = string_to_colors(result[0])
+            sql = (
+                "select blocks, blocks2, blocks3, blocks4, blocks5, blocks6, blocks7, blocks8 "
+                "from pictures "
+                "where path = ?"
+            )
+        blocks = self.con.execute(sql, [key]).fetchone()
+        if blocks:
+            result = [bytes_to_colors(block) for block in blocks]
             return result
         else:
             raise KeyError(key)
@@ -56,35 +76,33 @@ class SqliteCache:
         return result[0][0]
 
     def __setitem__(self, path_str, blocks):
-        blocks = colors_to_string(blocks)
+        blocks = [colors_to_bytes(block) for block in blocks]
         if op.exists(path_str):
             mtime = int(os.stat(path_str).st_mtime)
         else:
             mtime = 0
         if path_str in self:
-            sql = "update pictures set blocks = ?, mtime = ? where path = ?"
+            sql = (
+                "update pictures set blocks = ?, blocks2 = ?, blocks3 = ?, blocks4 = ?, blocks5 = ?, blocks6 = ?, "
+                "blocks7 = ?, blocks8 = ?, mtime_ns = ?"
+                "where path = ?"
+            )
         else:
-            sql = "insert into pictures(blocks,mtime,path) values(?,?,?)"
+            sql = (
+                "insert into pictures(blocks,blocks2,blocks3,blocks4,blocks5,blocks6,blocks7,blocks8,mtime_ns,path) "
+                "values(?,?,?,?,?,?,?,?,?,?)"
+            )
         try:
-            self.con.execute(sql, [blocks, mtime, path_str])
+            self.con.execute(sql, blocks + [mtime, path_str])
         except sqlite.OperationalError:
             logging.warning("Picture cache could not set value for key %r", path_str)
         except sqlite.DatabaseError as e:
             logging.warning("DatabaseError while setting value for key %r: %s", path_str, str(e))
 
     def _create_con(self, second_try=False):
-        def create_tables():
-            logging.debug("Creating picture cache tables.")
-            self.con.execute("drop table if exists pictures")
-            self.con.execute("drop index if exists idx_path")
-            self.con.execute("create table pictures(path TEXT, mtime INTEGER, blocks TEXT)")
-            self.con.execute("create index idx_path on pictures (path)")
-
-        self.con = sqlite.connect(self.dbname, isolation_level=None)
         try:
-            self.con.execute("select path, mtime, blocks from pictures where 1=2")
-        except sqlite.OperationalError:  # new db
-            create_tables()
+            self.con = sqlite.connect(self.dbname, isolation_level=None)
+            self._check_upgrade()
         except sqlite.DatabaseError as e:  # corrupted db
             if second_try:
                 raise  # Something really strange is happening
@@ -93,6 +111,25 @@ class SqliteCache:
             os.remove(self.dbname)
             self._create_con(second_try=True)
 
+    def _check_upgrade(self) -> None:
+        with self.con as conn:
+            has_schema = conn.execute(
+                "SELECT NAME FROM sqlite_master WHERE type='table' AND name='schema_version'"
+            ).fetchall()
+            version = None
+            if has_schema:
+                version = conn.execute("SELECT version FROM schema_version ORDER BY version DESC").fetchone()[0]
+            else:
+                conn.execute("CREATE TABLE schema_version (version int PRIMARY KEY, description TEXT)")
+            if version != self.schema_version:
+                conn.execute(self.drop_table_query)
+                conn.execute(
+                    "INSERT OR REPLACE INTO schema_version VALUES (:version, :description)",
+                    {"version": self.schema_version, "description": self.schema_version_description},
+                )
+            conn.execute(self.create_table_query)
+            conn.execute(self.create_index_query)
+
     def clear(self):
         self.close()
         if self.dbname != ":memory:":
@@ -118,9 +155,28 @@ class SqliteCache:
             raise ValueError(path)
 
     def get_multiple(self, rowids):
-        sql = "select rowid, blocks from pictures where rowid in (%s)" % ",".join(map(str, rowids))
+        ids = ",".join(map(str, rowids))
+        sql = (
+            "select rowid, blocks, blocks2, blocks3, blocks4, blocks5, blocks6, blocks7, blocks8 "
+            f"from pictures where rowid in ({ids})"
+        )
         cur = self.con.execute(sql)
-        return ((rowid, string_to_colors(blocks)) for rowid, blocks in cur)
+        return (
+            (
+                rowid,
+                [
+                    bytes_to_colors(blocks),
+                    bytes_to_colors(blocks2),
+                    bytes_to_colors(blocks3),
+                    bytes_to_colors(blocks4),
+                    bytes_to_colors(blocks5),
+                    bytes_to_colors(blocks6),
+                    bytes_to_colors(blocks7),
+                    bytes_to_colors(blocks8),
+                ],
+            )
+            for rowid, blocks, blocks2, blocks3, blocks4, blocks5, blocks6, blocks7, blocks8 in cur
+        )
 
     def purge_outdated(self):
         """Go through the cache and purge outdated records.
@@ -129,12 +185,12 @@ class SqliteCache:
         the db.
         """
         todelete = []
-        sql = "select rowid, path, mtime from pictures"
+        sql = "select rowid, path, mtime_ns from pictures"
         cur = self.con.execute(sql)
-        for rowid, path_str, mtime in cur:
-            if mtime and op.exists(path_str):
+        for rowid, path_str, mtime_ns in cur:
+            if mtime_ns and op.exists(path_str):
                 picture_mtime = os.stat(path_str).st_mtime
-                if int(picture_mtime) <= mtime:
+                if int(picture_mtime) <= mtime_ns:
                     # not outdated
                     continue
             todelete.append(rowid)

core/pe/iphoto_plist.py

@@ -1,33 +0,0 @@
-# Created By: Virgil Dupras
-# Created On: 2014-03-15
-# Copyright 2015 Hardcoded Software (http://www.hardcoded.net)
-#
-# This software is licensed under the "GPLv3" License as described in the "LICENSE" file,
-# which should be included with this package. The terms are also available at
-# http://www.gnu.org/licenses/gpl-3.0.html
-
-import plistlib
-
-
-class IPhotoPlistParser(plistlib._PlistParser):
-    """A parser for iPhoto plists.
-
-    iPhoto plists tend to be malformed, so we have to subclass the built-in parser to be a bit more
-    lenient.
-    """
-
-    def __init__(self):
-        plistlib._PlistParser.__init__(self, use_builtin_types=True, dict_type=dict)
-        # For debugging purposes, we remember the last bit of data to be analyzed so that we can
-        # log it in case of an exception
-        self.lastdata = ""
-
-    def get_data(self):
-        self.lastdata = plistlib._PlistParser.get_data(self)
-        return self.lastdata
-
-    def end_integer(self):
-        try:
-            self.add_object(int(self.get_data()))
-        except ValueError:
-            self.add_object(0)

core/pe/matchblock.py

@@ -15,7 +15,8 @@ from hscommon.trans import tr
 from hscommon.jobprogress import job
 
 from core.engine import Match
-from .block import avgdiff, DifferentBlockCountError, NoBlocksError
+from core.pe.block import avgdiff, DifferentBlockCountError, NoBlocksError
+from core.pe.cache_sqlite import SqliteCache
 
 # OPTIMIZATION NOTES:
 # The bottleneck of the matching phase is CPU, which is why we use multiprocessing. However, another
@@ -27,7 +28,7 @@ from .block import avgdiff, DifferentBlockCountError, NoBlocksError
 # to files in other chunks. So chunkifying doesn't save us any actual comparison, but the advantage
 # is that instead of reading blocks from disk number_of_files**2 times, we read it
 # number_of_files*number_of_chunks times.
-# Determining the right chunk size is tricky, bceause if it's too big, too many blocks will be in
+# Determining the right chunk size is tricky, because if it's too big, too many blocks will be in
 # memory at the same time and we might end up with memory trashing, which is awfully slow. So,
 # because our *real* bottleneck is CPU, the chunk size must simply be enough so that the CPU isn't
 # starved by Disk IOs.
@@ -50,17 +51,10 @@ except Exception:
 
 
 def get_cache(cache_path, readonly=False):
-    if cache_path.endswith("shelve"):
-        from .cache_shelve import ShelveCache
-
-        return ShelveCache(cache_path, readonly=readonly)
-    else:
-        from .cache_sqlite import SqliteCache
-
-        return SqliteCache(cache_path, readonly=readonly)
+    return SqliteCache(cache_path, readonly=readonly)
 
 
-def prepare_pictures(pictures, cache_path, with_dimensions, j=job.nulljob):
+def prepare_pictures(pictures, cache_path, with_dimensions, match_rotated, j=job.nulljob):
     # The MemoryError handlers in there use logging without first caring about whether or not
     # there is enough memory left to carry on the operation because it is assumed that the
     # MemoryError happens when trying to read an image file, which is freed from memory by the
@@ -78,16 +72,21 @@ def prepare_pictures(pictures, cache_path, with_dimensions, j=job.nulljob):
                 # entry in iPhoto library.
                 logging.warning("We have a picture with a null path here")
                 continue
-            picture.unicode_path = str(picture.path)
             logging.debug("Analyzing picture at %s", picture.unicode_path)
             if with_dimensions:
                 picture.dimensions  # pre-read dimensions
             try:
-                if picture.unicode_path not in cache:
-                    blocks = picture.get_blocks(BLOCK_COUNT_PER_SIDE)
+                if picture.unicode_path not in cache or (
+                    match_rotated and any(block == [] for block in cache[picture.unicode_path])
+                ):
+                    if match_rotated:
+                        blocks = [picture.get_blocks(BLOCK_COUNT_PER_SIDE, orientation) for orientation in range(1, 9)]
+                    else:
+                        blocks = [[]] * 8
+                        blocks[max(picture.get_orientation() - 1, 0)] = picture.get_blocks(BLOCK_COUNT_PER_SIDE)
                     cache[picture.unicode_path] = blocks
                 prepared.append(picture)
-            except (IOError, ValueError) as e:
+            except (OSError, ValueError) as e:
                 logging.warning(str(e))
             except MemoryError:
                 logging.warning(
@@ -125,13 +124,13 @@ def get_match(first, second, percentage):
     return Match(first, second, percentage)
 
 
-def async_compare(ref_ids, other_ids, dbname, threshold, picinfo):
+def async_compare(ref_ids, other_ids, dbname, threshold, picinfo, match_rotated=False):
     # The list of ids in ref_ids have to be compared to the list of ids in other_ids. other_ids
     # can be None. In this case, ref_ids has to be compared with itself
     # picinfo is a dictionary {pic_id: (dimensions, is_ref)}
     cache = get_cache(dbname, readonly=True)
     limit = 100 - threshold
-    ref_pairs = list(cache.get_multiple(ref_ids))
+    ref_pairs = list(cache.get_multiple(ref_ids))  # (rowid, [b, b2, ..., b8])
     if other_ids is not None:
         other_pairs = list(cache.get_multiple(other_ids))
         comparisons_to_do = [(r, o) for r in ref_pairs for o in other_pairs]
@@ -144,22 +143,35 @@ def async_compare(ref_ids, other_ids, dbname, threshold, picinfo):
         if ref_is_ref and other_is_ref:
             continue
         if ref_dimensions != other_dimensions:
-            continue
-        try:
-            diff = avgdiff(ref_blocks, other_blocks, limit, MIN_ITERATIONS)
-            percentage = 100 - diff
-        except (DifferentBlockCountError, NoBlocksError):
-            percentage = 0
-        if percentage >= threshold:
-            results.append((ref_id, other_id, percentage))
+            if match_rotated:
+                rotated_ref_dimensions = (ref_dimensions[1], ref_dimensions[0])
+                if rotated_ref_dimensions != other_dimensions:
+                    continue
+            else:
+                continue
+
+        orientation_range = 1
+        if match_rotated:
+            orientation_range = 8
+
+        for orientation_ref in range(orientation_range):
+            try:
+                diff = avgdiff(ref_blocks[orientation_ref], other_blocks[0], limit, MIN_ITERATIONS)
+                percentage = 100 - diff
+            except (DifferentBlockCountError, NoBlocksError):
+                percentage = 0
+            if percentage >= threshold:
+                results.append((ref_id, other_id, percentage))
+                break
+
     cache.close()
     return results
 
 
-def getmatches(pictures, cache_path, threshold, match_scaled=False, j=job.nulljob):
+def getmatches(pictures, cache_path, threshold, match_scaled=False, match_rotated=False, j=job.nulljob):
     def get_picinfo(p):
         if match_scaled:
-            return (None, p.is_ref)
+            return ((None, None), p.is_ref)
         else:
             return (p.dimensions, p.is_ref)
 
@@ -181,7 +193,7 @@ def getmatches(pictures, cache_path, threshold, match_scaled=False, j=job.nulljo
         j.set_progress(comparison_count, progress_msg)
 
     j = j.start_subjob([3, 7])
-    pictures = prepare_pictures(pictures, cache_path, with_dimensions=not match_scaled, j=j)
+    pictures = prepare_pictures(pictures, cache_path, not match_scaled, match_rotated, j=j)
     j = j.start_subjob([9, 1], tr("Preparing for matching"))
     cache = get_cache(cache_path)
     id2picture = {}
@@ -211,7 +223,7 @@ def getmatches(pictures, cache_path, threshold, match_scaled=False, j=job.nulljo
                 picinfo.update({p.cache_id: get_picinfo(p) for p in other_chunk})
             else:
                 other_ids = None
-            args = (ref_ids, other_ids, cache_path, threshold, picinfo)
+            args = (ref_ids, other_ids, cache_path, threshold, picinfo, match_rotated)
             async_results.append(pool.apply_async(async_compare, args))
             collect_results()
         collect_results(collect_all=True)
@@ -238,7 +250,7 @@ def getmatches(pictures, cache_path, threshold, match_scaled=False, j=job.nulljo
     for ref_id, other_id, percentage in myiter:
         ref = id2picture[ref_id]
         other = id2picture[other_id]
-        if percentage == 100 and ref.md5 != other.md5:
+        if percentage == 100 and ref.digest != other.digest:
             percentage = 99
         if percentage >= threshold:
             ref.dimensions  # pre-read dimensions for display in results

core/pe/modules/block.c

@@ -245,4 +245,4 @@ PyObject *PyInit__block(void) {
   PyModule_AddObject(m, "DifferentBlockCountError", DifferentBlockCountError);
 
   return m;
-}
+}

core/pe/modules/block_osx.m

@@ -2,8 +2,8 @@
  * Created On: 2010-02-04
  * Copyright 2015 Hardcoded Software (http://www.hardcoded.net)
  *
- * This software is licensed under the "GPLv3" License as described in the "LICENSE" file, 
- * which should be included with this package. The terms are also available at 
+ * This software is licensed under the "GPLv3" License as described in the "LICENSE" file,
+ * which should be included with this package. The terms are also available at
  * http://www.gnu.org/licenses/gpl-3.0.html
 **/
 
@@ -22,7 +22,7 @@ pystring2cfstring(PyObject *pystring)
     UInt8 *s;
     CFIndex size;
     CFStringRef result;
-    
+
     if (PyUnicode_Check(pystring)) {
         encoded = PyUnicode_AsUTF8String(pystring);
         if (encoded == NULL) {
@@ -32,7 +32,7 @@ pystring2cfstring(PyObject *pystring)
         encoded = pystring;
         Py_INCREF(encoded);
     }
-    
+
     s = (UInt8*)PyBytes_AS_STRING(encoded);
     size = PyBytes_GET_SIZE(encoded);
     result = CFStringCreateWithBytes(NULL, s, size, kCFStringEncodingUTF8, FALSE);
@@ -50,20 +50,20 @@ static PyObject* block_osx_get_image_size(PyObject *self, PyObject *args)
     long width, height;
     PyObject *pwidth, *pheight;
     PyObject *result;
-    
+
     width = 0;
     height = 0;
     if (!PyArg_ParseTuple(args, "O", &path)) {
         return NULL;
     }
-    
+
     image_path = pystring2cfstring(path);
     if (image_path == NULL) {
         return PyErr_NoMemory();
     }
     image_url = CFURLCreateWithFileSystemPath(NULL, image_path, kCFURLPOSIXPathStyle, FALSE);
     CFRelease(image_path);
-    
+
     source = CGImageSourceCreateWithURL(image_url, NULL);
     CFRelease(image_url);
     if (source != NULL) {
@@ -75,7 +75,7 @@ static PyObject* block_osx_get_image_size(PyObject *self, PyObject *args)
         }
         CFRelease(source);
     }
-    
+
     pwidth = PyLong_FromLong(width);
     if (pwidth == NULL) {
         return NULL;
@@ -91,19 +91,19 @@ static PyObject* block_osx_get_image_size(PyObject *self, PyObject *args)
 }
 
 static CGContextRef
-MyCreateBitmapContext(int width, int height) 
+MyCreateBitmapContext(int width, int height)
 {
     CGContextRef context = NULL;
     CGColorSpaceRef colorSpace;
     void *bitmapData;
     int bitmapByteCount;
     int bitmapBytesPerRow;
-    
+
     bitmapBytesPerRow = (width * 4);
     bitmapByteCount = (bitmapBytesPerRow * height);
-    
+
     colorSpace = CGColorSpaceCreateWithName(kCGColorSpaceGenericRGB);
-    
+
     // calloc() must be used to allocate bitmapData here because the buffer has to be zeroed.
     // If it's not zeroes, when images with transparency are drawn in the context, this buffer
     // will stay with undefined pixels, which means that two pictures with the same pixels will
@@ -113,7 +113,7 @@ MyCreateBitmapContext(int width, int height)
         fprintf(stderr, "Memory not allocated!");
         return NULL;
     }
-    
+
     context = CGBitmapContextCreate(bitmapData, width, height, 8, bitmapBytesPerRow, colorSpace,
         (CGBitmapInfo)kCGImageAlphaNoneSkipLast);
     if (context== NULL) {
@@ -128,7 +128,7 @@ MyCreateBitmapContext(int width, int height)
 static PyObject* getblock(unsigned char *imageData, int imageWidth, int imageHeight, int boxX, int boxY, int boxW, int boxH)
 {
     int i,j, totalR, totalG, totalB;
-    
+
     totalR = totalG = totalB = 0;
     for(i=boxY; i<boxY+boxH; i++) {
         for(j=boxX; j<boxX+boxW; j++) {
@@ -142,7 +142,7 @@ static PyObject* getblock(unsigned char *imageData, int imageWidth, int imageHei
     totalR /= pixelCount;
     totalG /= pixelCount;
     totalB /= pixelCount;
-    
+
     return inttuple(3, totalR, totalG, totalB);
 }
 
@@ -155,27 +155,27 @@ static PyObject* block_osx_getblocks(PyObject *self, PyObject *args)
     CGImageRef image;
     size_t width, height, image_width, image_height;
     int block_count, block_width, block_height, orientation, i;
-    
+
     if (!PyArg_ParseTuple(args, "Oii", &path, &block_count, &orientation)) {
         return NULL;
     }
-    
+
     if (PySequence_Length(path) == 0) {
         PyErr_SetString(PyExc_ValueError, "empty path");
         return NULL;
     }
-    
+
     if ((orientation > 8) || (orientation < 0)) {
         orientation = 0; // simplifies checks later since we can only have values in 0-8
     }
-    
+
     image_path = pystring2cfstring(path);
     if (image_path == NULL) {
         return PyErr_NoMemory();
     }
     image_url = CFURLCreateWithFileSystemPath(NULL, image_path, kCFURLPOSIXPathStyle, FALSE);
     CFRelease(image_path);
-    
+
     source = CGImageSourceCreateWithURL(image_url, NULL);
     CFRelease(image_url);
     if (source == NULL) {
@@ -187,8 +187,8 @@ static PyObject* block_osx_getblocks(PyObject *self, PyObject *args)
         CFRelease(source);
         return PyErr_NoMemory();
     }
-    
-    
+
+
     width = image_width = CGImageGetWidth(image);
     height = image_height = CGImageGetHeight(image);
     if (orientation >= 5) {
@@ -196,9 +196,9 @@ static PyObject* block_osx_getblocks(PyObject *self, PyObject *args)
         width = image_height;
         height = image_width;
     }
-    
+
     CGContextRef context = MyCreateBitmapContext(width, height);
-    
+
     if (orientation == 2) {
         // Flip X
         CGContextTranslateCTM(context, width, 0);
@@ -207,7 +207,7 @@ static PyObject* block_osx_getblocks(PyObject *self, PyObject *args)
     else if (orientation == 3) {
         // Rot 180
         CGContextTranslateCTM(context, width, height);
-        CGContextRotateCTM(context, RADIANS(180)); 
+        CGContextRotateCTM(context, RADIANS(180));
     }
     else if (orientation == 4) {
         // Flip Y
@@ -242,21 +242,21 @@ static PyObject* block_osx_getblocks(PyObject *self, PyObject *args)
     CGContextDrawImage(context, myBoundingBox, image);
     unsigned char *bitmapData = CGBitmapContextGetData(context);
     CGContextRelease(context);
-    
+
     CGImageRelease(image);
     CFRelease(source);
     if (bitmapData == NULL) {
         return PyErr_NoMemory();
     }
-    
+
     block_width = max(width/block_count, 1);
     block_height = max(height/block_count, 1);
-    
+
     result = PyList_New(block_count * block_count);
     if (result == NULL) {
         return NULL;
     }
-    
+
     for(i=0; i<block_count; i++) {
         int j, top;
         top = min(i*block_height, height-block_height);
@@ -271,8 +271,8 @@ static PyObject* block_osx_getblocks(PyObject *self, PyObject *args)
             PyList_SET_ITEM(result, i*block_count+j, block);
         }
     }
-    
-    free(bitmapData); 
+
+    free(bitmapData);
     return result;
 }
 
@@ -302,4 +302,4 @@ PyInit__block_osx(void)
         return NULL;
     }
     return m;
-}
+}

core/pe/modules/cache.c

@@ -2,94 +2,68 @@
  * Created On: 2010-01-30
  * Copyright 2014 Hardcoded Software (http://www.hardcoded.net)
  *
- * This software is licensed under the "BSD" License as described in the "LICENSE" file, 
- * which should be included with this package. The terms are also available at 
- * http://www.hardcoded.net/licenses/bsd_license
+ * This software is licensed under the "BSD" License as described in the
+ * "LICENSE" file, which should be included with this package. The terms are
+ * also available at http://www.hardcoded.net/licenses/bsd_license
  */
 
 #include "common.h"
 
-/* I know that there strtol out there, but it requires a pointer to
- * a char, which would in turn require me to buffer my chars around,
- * making the whole process slower.
- */
-static long
-xchar_to_long(char c)
-{
-    if ((c >= 48) && (c <= 57)) { /* 0-9 */
-        return c - 48;
-    }
-    else if ((c >= 65) && (c <= 70)) { /* A-F */
-        return c - 55;
-    }
-    else if ((c >= 97) && (c <= 102)) { /* a-f */
-        return c - 87;
-    }
-    return 0;
-}
+static PyObject *cache_bytes_to_colors(PyObject *self, PyObject *args) {
+  char *y;
+  Py_ssize_t char_count, i, color_count;
+  PyObject *result;
+  unsigned long r, g, b;
+  Py_ssize_t ci;
+  PyObject *color_tuple;
 
-static PyObject*
-cache_string_to_colors(PyObject *self, PyObject *args)
-{
-    char *s;
-    Py_ssize_t char_count, color_count, i;
-    PyObject *result;
-    
-    if (!PyArg_ParseTuple(args, "s#", &s, &char_count)) {
-        return NULL;
+  if (!PyArg_ParseTuple(args, "y#", &y, &char_count)) {
+    return NULL;
+  }
+
+  color_count = char_count / 3;
+  result = PyList_New(color_count);
+  if (result == NULL) {
+    return NULL;
+  }
+
+  for (i = 0; i < color_count; i++) {
+    ci = i * 3;
+    r = (unsigned char)y[ci];
+    g = (unsigned char)y[ci + 1];
+    b = (unsigned char)y[ci + 2];
+
+    color_tuple = inttuple(3, r, g, b);
+    if (color_tuple == NULL) {
+      Py_DECREF(result);
+      return NULL;
     }
-    
-    color_count = (char_count / 6);
-    result = PyList_New(color_count);
-    if (result == NULL) {
-        return NULL;
-    }
-    
-    for (i=0; i<color_count; i++) {
-        long r, g, b;
-        Py_ssize_t ci;
-        PyObject *color_tuple;
-        
-        ci = i * 6;
-        r = (xchar_to_long(s[ci]) << 4) + xchar_to_long(s[ci+1]);
-        g = (xchar_to_long(s[ci+2]) << 4) + xchar_to_long(s[ci+3]);
-        b = (xchar_to_long(s[ci+4]) << 4) + xchar_to_long(s[ci+5]);
-        
-        color_tuple = inttuple(3, r, g, b);
-        if (color_tuple == NULL) {
-            Py_DECREF(result);
-            return NULL;
-        }
-        PyList_SET_ITEM(result, i, color_tuple);
-    }
-    
-    return result;
+    PyList_SET_ITEM(result, i, color_tuple);
+  }
+
+  return result;
 }
 
 static PyMethodDef CacheMethods[] = {
-    {"string_to_colors",  cache_string_to_colors, METH_VARARGS,
-     "Transform the string 's' in a list of 3 sized tuples."},
-    {NULL, NULL, 0, NULL}        /* Sentinel */
+    {"bytes_to_colors", cache_bytes_to_colors, METH_VARARGS,
+     "Transform the bytes 's' into a list of 3 sized tuples."},
+    {NULL, NULL, 0, NULL} /* Sentinel */
 };
 
-static struct PyModuleDef CacheDef = {
-    PyModuleDef_HEAD_INIT,
-    "_cache",
-    NULL,
-    -1,
-    CacheMethods,
-    NULL,
-    NULL,
-    NULL,
-    NULL
-};
+static struct PyModuleDef CacheDef = {PyModuleDef_HEAD_INIT,
+                                      "_cache",
+                                      NULL,
+                                      -1,
+                                      CacheMethods,
+                                      NULL,
+                                      NULL,
+                                      NULL,
+                                      NULL};
 
-PyObject *
-PyInit__cache(void)
-{
-    PyObject *m = PyModule_Create(&CacheDef);
-    if (m == NULL) {
-        return NULL;
-    }
-    return m;
-}
+PyObject *PyInit__cache(void) {
+  PyObject *m = PyModule_Create(&CacheDef);
+  if (m == NULL) {
+    return NULL;
+  }
+  return m;
+}

core/pe/modules/common.c

@@ -2,8 +2,8 @@
  * Created On: 2010-02-04
  * Copyright 2014 Hardcoded Software (http://www.hardcoded.net)
  *
- * This software is licensed under the "BSD" License as described in the "LICENSE" file, 
- * which should be included with this package. The terms are also available at 
+ * This software is licensed under the "BSD" License as described in the "LICENSE" file,
+ * which should be included with this package. The terms are also available at
  * http://www.hardcoded.net/licenses/bsd_license
  */
 
@@ -27,19 +27,19 @@ PyObject* inttuple(int n, ...)
     PyObject *pnumber;
     PyObject *result;
     va_list numbers;
-    
+
     va_start(numbers, n);
     result = PyTuple_New(n);
-    
+
     for (i=0; i<n; i++) {
-        pnumber = PyLong_FromLong(va_arg(numbers, long));
+        pnumber = PyLong_FromUnsignedLong(va_arg(numbers, long));
         if (pnumber == NULL) {
             Py_DECREF(result);
             return NULL;
         }
         PyTuple_SET_ITEM(result, i, pnumber);
     }
-    
+
     va_end(numbers);
     return result;
 }

core/pe/modules/common.h

@@ -2,8 +2,8 @@
  * Created On: 2010-02-04
  * Copyright 2015 Hardcoded Software (http://www.hardcoded.net)
  *
- * This software is licensed under the "GPLv3" License as described in the "LICENSE" file, 
- * which should be included with this package. The terms are also available at 
+ * This software is licensed under the "GPLv3" License as described in the "LICENSE" file,
+ * which should be included with this package. The terms are also available at
  * http://www.gnu.org/licenses/gpl-3.0.html
  */
 
@@ -17,4 +17,4 @@ int min(int a, int b);
 #endif
 
 /* Create a tuple out of an array of integers. */
-PyObject* inttuple(int n, ...);
+PyObject* inttuple(int n, ...);

core/pe/photo.py

@@ -9,7 +9,7 @@ from hscommon.util import get_file_ext, format_size
 
 from core.util import format_timestamp, format_perc, format_dupe_count
 from core import fs
-from . import exif
+from core.pe import exif
 
 # This global value is set by the platform-specific subclasser of the Photo base class
 PLAT_SPECIFIC_PHOTO_CLASS = None
@@ -29,7 +29,7 @@ class Photo(fs.File):
     __slots__ = fs.File.__slots__ + tuple(INITIAL_INFO.keys())
 
     # These extensions are supported on all platforms
-    HANDLED_EXTS = {"png", "jpg", "jpeg", "gif", "bmp", "tiff", "tif"}
+    HANDLED_EXTS = {"png", "jpg", "jpeg", "gif", "bmp", "tiff", "tif", "webp"}
 
     def _plat_get_dimensions(self):
         raise NotImplementedError()
@@ -37,7 +37,7 @@ class Photo(fs.File):
     def _plat_get_blocks(self, block_count_per_side, orientation):
         raise NotImplementedError()
 
-    def _get_orientation(self):
+    def get_orientation(self):
         if not hasattr(self, "_cached_orientation"):
             try:
                 with self.path.open("rb") as fp:
@@ -95,10 +95,13 @@ class Photo(fs.File):
         fs.File._read_info(self, field)
         if field == "dimensions":
             self.dimensions = self._plat_get_dimensions()
-            if self._get_orientation() in {5, 6, 7, 8}:
+            if self.get_orientation() in {5, 6, 7, 8}:
                 self.dimensions = (self.dimensions[1], self.dimensions[0])
         elif field == "exif_timestamp":
             self.exif_timestamp = self._get_exif_timestamp()
 
-    def get_blocks(self, block_count_per_side):
-        return self._plat_get_blocks(block_count_per_side, self._get_orientation())
+    def get_blocks(self, block_count_per_side, orientation: int = None):
+        if orientation is None:
+            return self._plat_get_blocks(block_count_per_side, self.get_orientation())
+        else:
+            return self._plat_get_blocks(block_count_per_side, orientation)

core/pe/scanner.py

@@ -8,12 +8,13 @@ from hscommon.trans import tr
 
 from core.scanner import Scanner, ScanType, ScanOption
 
-from . import matchblock, matchexif
+from core.pe import matchblock, matchexif
 
 
 class ScannerPE(Scanner):
     cache_path = None
     match_scaled = False
+    match_rotated = False
 
     @staticmethod
     def get_scan_options():
@@ -29,6 +30,7 @@ class ScannerPE(Scanner):
                 cache_path=self.cache_path,
                 threshold=self.min_match_percentage,
                 match_scaled=self.match_scaled,
+                match_rotated=self.match_rotated,
                 j=j,
             )
         elif self.scan_type == ScanType.EXIFTIMESTAMP:

core/prioritize.py

@@ -43,7 +43,7 @@ class Criterion:
 
     @property
     def display(self):
-        return "{} ({})".format(self.category.NAME, self.display_value)
+        return f"{self.category.NAME} ({self.display_value})"
 
 
 class ValueListCategory(CriterionCategory):
@@ -82,10 +82,12 @@ class FolderCategory(ValueListCategory):
 
     def sort_key(self, dupe, crit_value):
         value = self.extract_value(dupe)
-        if value[: len(crit_value)] == crit_value:
-            return 0
-        else:
+        # This is instead of using is_relative_to() which was added in py 3.9
+        try:
+            value.relative_to(crit_value)
+        except ValueError:
             return 1
+        return 0
 
 
 class FilenameCategory(CriterionCategory):

core/results.py

@@ -10,6 +10,7 @@ import logging
 import re
 import os
 import os.path as op
+from errno import EISDIR, EACCES
 from xml.etree import ElementTree as ET
 
 from hscommon.jobprogress.job import nulljob
@@ -17,8 +18,8 @@ from hscommon.conflict import get_conflicted_name
 from hscommon.util import flatten, nonone, FileOrPath, format_size
 from hscommon.trans import tr
 
-from . import engine
-from .markable import Markable
+from core import engine
+from core.markable import Markable
 
 
 class Results(Markable):
@@ -191,7 +192,7 @@ class Results(Markable):
             self.__filters.append(filter_str)
             if self.__filtered_dupes is None:
                 self.__filtered_dupes = flatten(g[:] for g in self.groups)
-            self.__filtered_dupes = set(dupe for dupe in self.__filtered_dupes if filter_re.search(str(dupe.path)))
+            self.__filtered_dupes = {dupe for dupe in self.__filtered_dupes if filter_re.search(str(dupe.path))}
             filtered_groups = set()
             for dupe in self.__filtered_dupes:
                 filtered_groups.add(self.get_group_of_duplicate(dupe))
@@ -301,7 +302,7 @@ class Results(Markable):
             try:
                 func(dupe)
                 to_remove.append(dupe)
-            except (EnvironmentError, UnicodeEncodeError) as e:
+            except (OSError, UnicodeEncodeError) as e:
                 self.problems.append((dupe, str(e)))
         if remove_from_results:
             self.remove_duplicates(to_remove)
@@ -374,10 +375,10 @@ class Results(Markable):
 
         try:
             do_write(outfile)
-        except IOError as e:
-            # If our IOError is because dest is already a directory, we want to handle that. 21 is
-            # the code we get on OS X and Linux, 13 is what we get on Windows.
-            if e.errno in {21, 13}:
+        except OSError as e:
+            # If our OSError is because dest is already a directory, we want to handle that. 21 is
+            # the code we get on OS X and Linux (EISDIR), 13 is what we get on Windows (EACCES).
+            if e.errno in (EISDIR, EACCES):
                 p = str(outfile)
                 dirname, basename = op.split(p)
                 otherfiles = os.listdir(dirname)

core/scanner.py

@@ -13,7 +13,7 @@ from hscommon.jobprogress import job
 from hscommon.util import dedupe, rem_file_ext, get_file_ext
 from hscommon.trans import tr
 
-from . import engine
+from core import engine
 
 # It's quite ugly to have scan types from all editions all put in the same class, but because there's
 # there will be some nasty bugs popping up (ScanType is used in core when in should exclusively be
@@ -87,8 +87,6 @@ class Scanner:
             }
         ):
             j = j.start_subjob([2, 8])
-            for f in j.iter_with_progress(files, tr("Read size of %d/%d files")):
-                f.size  # pre-read, makes a smoother progress if read here (especially for bundles)
             if self.size_threshold:
                 files = [f for f in files if f.size >= self.size_threshold]
             if self.large_size_threshold:
@@ -134,7 +132,7 @@ class Scanner:
             return False
         if is_same_with_digit(refname, dupename):
             return True
-        return len(dupe.path) > len(ref.path)
+        return len(dupe.path.parts) > len(ref.path.parts)
 
     @staticmethod
     def get_scan_options():
@@ -164,15 +162,18 @@ class Scanner:
             toremove = set()
             last_parent_path = sortedpaths[0]
             for p in sortedpaths[1:]:
-                if p in last_parent_path:
+                if last_parent_path in p.parents:
                     toremove.add(p)
                 else:
                     last_parent_path = p
             matches = [m for m in matches if m.first.path not in toremove or m.second.path not in toremove]
         if not self.mix_file_kind:
             matches = [m for m in matches if get_file_ext(m.first.name) == get_file_ext(m.second.name)]
-        matches = [m for m in matches if m.first.path.exists() and m.second.path.exists()]
-        matches = [m for m in matches if not (m.first.is_ref and m.second.is_ref)]
+        if self.include_exists_check:
+            matches = [m for m in matches if m.first.exists() and m.second.exists()]
+        # Contents already handles ref checks, other scan types might not catch during scan
+        if self.scan_type != ScanType.CONTENTS:
+            matches = [m for m in matches if not (m.first.is_ref and m.second.is_ref)]
         if ignore_list:
             matches = [m for m in matches if not ignore_list.are_ignored(str(m.first.path), str(m.second.path))]
         logging.info("Grouping matches")
@@ -212,3 +213,4 @@ class Scanner:
     large_size_threshold = 0
     big_file_size_threshold = 0
     word_weighting = False
+    include_exists_check = True

core/se/__init__.py

@@ -1 +1 @@
-from . import fs, result_table, scanner  # noqa
+from core.se import fs, result_table, scanner  # noqa

core/tests/app_test.py

@@ -7,18 +7,19 @@
 import os
 import os.path as op
 import logging
+import tempfile
 
 import pytest
-from hscommon.path import Path
+from pathlib import Path
 import hscommon.conflict
 import hscommon.util
 from hscommon.testutil import eq_, log_calls
 from hscommon.jobprogress.job import Job
 
-from .base import TestApp
-from .results_test import GetTestGroups
-from .. import app, fs, engine
-from ..scanner import ScanType
+from core.tests.base import TestApp
+from core.tests.results_test import GetTestGroups
+from core import app, fs, engine
+from core.scanner import ScanType
 
 
 def add_fake_files_to_directories(directories, files):
@@ -56,7 +57,7 @@ class TestCaseDupeGuru:
         # for this unit is pathetic. What's done is done. My approach now is to add tests for
         # every change I want to make. The blowup was caused by a missing import.
         p = Path(str(tmpdir))
-        p["foo"].open("w").close()
+        p.joinpath("foo").touch()
         monkeypatch.setattr(
             hscommon.conflict,
             "smart_copy",
@@ -68,22 +69,23 @@ class TestCaseDupeGuru:
         dgapp = TestApp().app
         dgapp.directories.add_path(p)
         [f] = dgapp.directories.get_files()
-        dgapp.copy_or_move(f, True, "some_destination", 0)
-        eq_(1, len(hscommon.conflict.smart_copy.calls))
-        call = hscommon.conflict.smart_copy.calls[0]
-        eq_(call["dest_path"], op.join("some_destination", "foo"))
-        eq_(call["source_path"], f.path)
+        with tempfile.TemporaryDirectory() as tmp_dir:
+            dgapp.copy_or_move(f, True, tmp_dir, 0)
+            eq_(1, len(hscommon.conflict.smart_copy.calls))
+            call = hscommon.conflict.smart_copy.calls[0]
+            eq_(call["dest_path"], Path(tmp_dir, "foo"))
+            eq_(call["source_path"], f.path)
 
     def test_copy_or_move_clean_empty_dirs(self, tmpdir, monkeypatch):
         tmppath = Path(str(tmpdir))
-        sourcepath = tmppath["source"]
+        sourcepath = tmppath.joinpath("source")
         sourcepath.mkdir()
-        sourcepath["myfile"].open("w")
+        sourcepath.joinpath("myfile").touch()
         app = TestApp().app
         app.directories.add_path(tmppath)
         [myfile] = app.directories.get_files()
         monkeypatch.setattr(app, "clean_empty_dirs", log_calls(lambda path: None))
-        app.copy_or_move(myfile, False, tmppath["dest"], 0)
+        app.copy_or_move(myfile, False, tmppath.joinpath("dest"), 0)
         calls = app.clean_empty_dirs.calls
         eq_(1, len(calls))
         eq_(sourcepath, calls[0]["path"])
@@ -95,7 +97,7 @@ class TestCaseDupeGuru:
 
         # At some point, any() was used in a wrong way that made Scan() wrongly return 1
         app = TestApp().app
-        f1, f2 = [FakeFile("foo") for _ in range(2)]
+        f1, f2 = (FakeFile("foo") for _ in range(2))
         f1.is_ref, f2.is_ref = (False, False)
         assert not (bool(f1) and bool(f2))
         add_fake_files_to_directories(app.directories, [f1, f2])
@@ -106,8 +108,8 @@ class TestCaseDupeGuru:
         # If the ignore_hardlink_matches option is set, don't match files hardlinking to the same
         # inode.
         tmppath = Path(str(tmpdir))
-        tmppath["myfile"].open("w").write("foo")
-        os.link(str(tmppath["myfile"]), str(tmppath["hardlink"]))
+        tmppath.joinpath("myfile").open("wt").write("foo")
+        os.link(str(tmppath.joinpath("myfile")), str(tmppath.joinpath("hardlink")))
         app = TestApp().app
         app.directories.add_path(tmppath)
         app.options["scan_type"] = ScanType.CONTENTS
@@ -153,7 +155,7 @@ class TestCaseDupeGuruCleanEmptyDirs:
         # delete_if_empty must be recursively called up in the path until it returns False
         @log_calls
         def mock_delete_if_empty(path, files_to_delete=[]):
-            return len(path) > 1
+            return len(path.parts) > 1
 
         monkeypatch.setattr(hscommon.util, "delete_if_empty", mock_delete_if_empty)
         # XXX This monkeypatch is temporary. will be fixed in a better monkeypatcher.
@@ -180,8 +182,8 @@ class TestCaseDupeGuruWithResults:
         self.rtable.refresh()
         tmpdir = request.getfixturevalue("tmpdir")
         tmppath = Path(str(tmpdir))
-        tmppath["foo"].mkdir()
-        tmppath["bar"].mkdir()
+        tmppath.joinpath("foo").mkdir()
+        tmppath.joinpath("bar").mkdir()
         self.app.directories.add_path(tmppath)
 
     def test_get_objects(self, do_setup):
@@ -424,12 +426,9 @@ class TestCaseDupeGuruRenameSelected:
     def do_setup(self, request):
         tmpdir = request.getfixturevalue("tmpdir")
         p = Path(str(tmpdir))
-        fp = open(str(p["foo bar 1"]), mode="w")
-        fp.close()
-        fp = open(str(p["foo bar 2"]), mode="w")
-        fp.close()
-        fp = open(str(p["foo bar 3"]), mode="w")
-        fp.close()
+        p.joinpath("foo bar 1").touch()
+        p.joinpath("foo bar 2").touch()
+        p.joinpath("foo bar 3").touch()
         files = fs.get_files(p)
         for f in files:
             f.is_ref = False
@@ -451,7 +450,7 @@ class TestCaseDupeGuruRenameSelected:
         g = self.groups[0]
         self.rtable.select([1])
         assert app.rename_selected("renamed")
-        names = [p.name for p in self.p.listdir()]
+        names = [p.name for p in self.p.glob("*")]
         assert "renamed" in names
         assert "foo bar 2" not in names
         eq_(g.dupes[0].name, "renamed")
@@ -464,7 +463,7 @@ class TestCaseDupeGuruRenameSelected:
         assert not app.rename_selected("renamed")
         msg = logging.warning.calls[0]["msg"]
         eq_("dupeGuru Warning: list index out of range", msg)
-        names = [p.name for p in self.p.listdir()]
+        names = [p.name for p in self.p.glob("*")]
         assert "renamed" not in names
         assert "foo bar 2" in names
         eq_(g.dupes[0].name, "foo bar 2")
@@ -477,7 +476,7 @@ class TestCaseDupeGuruRenameSelected:
         assert not app.rename_selected("foo bar 1")
         msg = logging.warning.calls[0]["msg"]
         assert msg.startswith("dupeGuru Warning: 'foo bar 1' already exists in")
-        names = [p.name for p in self.p.listdir()]
+        names = [p.name for p in self.p.glob("*")]
         assert "foo bar 1" in names
         assert "foo bar 2" in names
         eq_(g.dupes[0].name, "foo bar 2")
@@ -488,9 +487,9 @@ class TestAppWithDirectoriesInTree:
     def do_setup(self, request):
         tmpdir = request.getfixturevalue("tmpdir")
         p = Path(str(tmpdir))
-        p["sub1"].mkdir()
-        p["sub2"].mkdir()
-        p["sub3"].mkdir()
+        p.joinpath("sub1").mkdir()
+        p.joinpath("sub2").mkdir()
+        p.joinpath("sub3").mkdir()
         app = TestApp()
         self.app = app.app
         self.dtree = app.dtree

core/tests/base.py

@@ -5,17 +5,16 @@
 # http://www.gnu.org/licenses/gpl-3.0.html
 
 from hscommon.testutil import TestApp as TestAppBase, CallLogger, eq_, with_app  # noqa
-from hscommon.path import Path
+from pathlib import Path
 from hscommon.util import get_file_ext, format_size
 from hscommon.gui.column import Column
 from hscommon.jobprogress.job import nulljob, JobCancelled
 
-from .. import engine
-from .. import prioritize
-from ..engine import getwords
-from ..app import DupeGuru as DupeGuruBase
-from ..gui.result_table import ResultTable as ResultTableBase
-from ..gui.prioritize_dialog import PrioritizeDialog
+from core import engine, prioritize
+from core.engine import getwords
+from core.app import DupeGuru as DupeGuruBase
+from core.gui.result_table import ResultTable as ResultTableBase
+from core.gui.prioritize_dialog import PrioritizeDialog
 
 
 class DupeGuruView:
@@ -86,9 +85,9 @@ class NamedObject:
             folder = "basepath"
         self._folder = Path(folder)
         self.size = size
-        self.md5partial = name
-        self.md5 = name
-        self.md5samples = name
+        self.digest_partial = name
+        self.digest = name
+        self.digest_samples = name
         if with_words:
             self.words = getwords(name)
         self.is_ref = False
@@ -111,11 +110,11 @@ class NamedObject:
 
     @property
     def path(self):
-        return self._folder[self.name]
+        return self._folder.joinpath(self.name)
 
     @property
     def folder_path(self):
-        return self.path.parent()
+        return self.path.parent
 
     @property
     def extension(self):

core/tests/block_test.py

@@ -9,7 +9,7 @@ from pytest import raises, skip
 from hscommon.testutil import eq_
 
 try:
-    from ..pe.block import avgdiff, getblocks2, NoBlocksError, DifferentBlockCountError
+    from core.pe.block import avgdiff, getblocks2, NoBlocksError, DifferentBlockCountError
 except ImportError:
     skip("Can't import the block module, probably hasn't been compiled.")
 

core/tests/cache_test.py

@@ -10,41 +10,41 @@ from pytest import raises, skip
 from hscommon.testutil import eq_
 
 try:
-    from ..pe.cache import colors_to_string, string_to_colors
-    from ..pe.cache_sqlite import SqliteCache
-    from ..pe.cache_shelve import ShelveCache
+    from core.pe.cache import colors_to_bytes, bytes_to_colors
+    from core.pe.cache_sqlite import SqliteCache
 except ImportError:
     skip("Can't import the cache module, probably hasn't been compiled.")
 
 
 class TestCaseColorsToString:
     def test_no_color(self):
-        eq_("", colors_to_string([]))
+        eq_(b"", colors_to_bytes([]))
 
     def test_single_color(self):
-        eq_("000000", colors_to_string([(0, 0, 0)]))
-        eq_("010101", colors_to_string([(1, 1, 1)]))
-        eq_("0a141e", colors_to_string([(10, 20, 30)]))
+        eq_(b"\x00\x00\x00", colors_to_bytes([(0, 0, 0)]))
+        eq_(b"\x01\x01\x01", colors_to_bytes([(1, 1, 1)]))
+        eq_(b"\x0a\x14\x1e", colors_to_bytes([(10, 20, 30)]))
 
     def test_two_colors(self):
-        eq_("000102030405", colors_to_string([(0, 1, 2), (3, 4, 5)]))
+        eq_(b"\x00\x01\x02\x03\x04\x05", colors_to_bytes([(0, 1, 2), (3, 4, 5)]))
 
 
 class TestCaseStringToColors:
     def test_empty(self):
-        eq_([], string_to_colors(""))
+        eq_([], bytes_to_colors(b""))
 
     def test_single_color(self):
-        eq_([(0, 0, 0)], string_to_colors("000000"))
-        eq_([(2, 3, 4)], string_to_colors("020304"))
-        eq_([(10, 20, 30)], string_to_colors("0a141e"))
+        eq_([(0, 0, 0)], bytes_to_colors(b"\x00\x00\x00"))
+        eq_([(2, 3, 4)], bytes_to_colors(b"\x02\x03\x04"))
+        eq_([(10, 20, 30)], bytes_to_colors(b"\x0a\x14\x1e"))
 
     def test_two_colors(self):
-        eq_([(10, 20, 30), (40, 50, 60)], string_to_colors("0a141e28323c"))
+        eq_([(10, 20, 30), (40, 50, 60)], bytes_to_colors(b"\x0a\x14\x1e\x28\x32\x3c"))
 
     def test_incomplete_color(self):
         # don't return anything if it's not a complete color
-        eq_([], string_to_colors("102"))
+        eq_([], bytes_to_colors(b"\x01"))
+        eq_([(1, 2, 3)], bytes_to_colors(b"\x01\x02\x03\x04"))
 
 
 class BaseTestCaseCache:
@@ -59,13 +59,13 @@ class BaseTestCaseCache:
 
     def test_set_then_retrieve_blocks(self):
         c = self.get_cache()
-        b = [(0, 0, 0), (1, 2, 3)]
+        b = [[(0, 0, 0), (1, 2, 3)]] * 8
         c["foo"] = b
         eq_(b, c["foo"])
 
     def test_delitem(self):
         c = self.get_cache()
-        c["foo"] = ""
+        c["foo"] = [[]] * 8
         del c["foo"]
         assert "foo" not in c
         with raises(KeyError):
@@ -74,16 +74,16 @@ class BaseTestCaseCache:
     def test_persistance(self, tmpdir):
         DBNAME = tmpdir.join("hstest.db")
         c = self.get_cache(str(DBNAME))
-        c["foo"] = [(1, 2, 3)]
+        c["foo"] = [[(1, 2, 3)]] * 8
         del c
         c = self.get_cache(str(DBNAME))
-        eq_([(1, 2, 3)], c["foo"])
+        eq_([[(1, 2, 3)]] * 8, c["foo"])
 
     def test_filter(self):
         c = self.get_cache()
-        c["foo"] = ""
-        c["bar"] = ""
-        c["baz"] = ""
+        c["foo"] = [[]] * 8
+        c["bar"] = [[]] * 8
+        c["baz"] = [[]] * 8
         c.filter(lambda p: p != "bar")  # only 'bar' is removed
         eq_(2, len(c))
         assert "foo" in c
@@ -92,9 +92,9 @@ class BaseTestCaseCache:
 
     def test_clear(self):
         c = self.get_cache()
-        c["foo"] = ""
-        c["bar"] = ""
-        c["baz"] = ""
+        c["foo"] = [[]] * 8
+        c["bar"] = [[]] * 8
+        c["baz"] = [[]] * 8
         c.clear()
         eq_(0, len(c))
         assert "foo" not in c
@@ -104,7 +104,7 @@ class BaseTestCaseCache:
     def test_by_id(self):
         # it's possible to use the cache by referring to the files by their row_id
         c = self.get_cache()
-        b = [(0, 0, 0), (1, 2, 3)]
+        b = [[(0, 0, 0), (1, 2, 3)]] * 8
         c["foo"] = b
         foo_id = c.get_id("foo")
         eq_(c[foo_id], b)
@@ -127,15 +127,10 @@ class TestCaseSqliteCache(BaseTestCaseCache):
         fp.write("invalid sqlite content")
         fp.close()
         c = self.get_cache(dbname)  # should not raise a DatabaseError
-        c["foo"] = [(1, 2, 3)]
+        c["foo"] = [[(1, 2, 3)]] * 8
         del c
         c = self.get_cache(dbname)
-        eq_(c["foo"], [(1, 2, 3)])
-
-
-class TestCaseShelveCache(BaseTestCaseCache):
-    def get_cache(self, dbname=None):
-        return ShelveCache(dbname)
+        eq_(c["foo"], [[(1, 2, 3)]] * 8)
 
 
 class TestCaseCacheSQLEscape:
@@ -157,7 +152,7 @@ class TestCaseCacheSQLEscape:
 
     def test_delitem(self):
         c = self.get_cache()
-        c["foo'bar"] = []
+        c["foo'bar"] = [[]] * 8
         try:
             del c["foo'bar"]
         except KeyError:

core/tests/directories_test.py

@@ -10,45 +10,39 @@ import tempfile
 import shutil
 
 from pytest import raises
-from hscommon.path import Path
+from pathlib import Path
 from hscommon.testutil import eq_
 from hscommon.plat import ISWINDOWS
 
-from ..fs import File
-from ..directories import (
+from core.fs import File
+from core.directories import (
     Directories,
     DirectoryState,
     AlreadyThereError,
     InvalidPathError,
 )
-from ..exclude import ExcludeList, ExcludeDict
+from core.exclude import ExcludeList, ExcludeDict
 
 
 def create_fake_fs(rootpath):
     # We have it as a separate function because other units are using it.
-    rootpath = rootpath["fs"]
+    rootpath = rootpath.joinpath("fs")
     rootpath.mkdir()
-    rootpath["dir1"].mkdir()
-    rootpath["dir2"].mkdir()
-    rootpath["dir3"].mkdir()
-    fp = rootpath["file1.test"].open("w")
-    fp.write("1")
-    fp.close()
-    fp = rootpath["file2.test"].open("w")
-    fp.write("12")
-    fp.close()
-    fp = rootpath["file3.test"].open("w")
-    fp.write("123")
-    fp.close()
-    fp = rootpath["dir1"]["file1.test"].open("w")
-    fp.write("1")
-    fp.close()
-    fp = rootpath["dir2"]["file2.test"].open("w")
-    fp.write("12")
-    fp.close()
-    fp = rootpath["dir3"]["file3.test"].open("w")
-    fp.write("123")
-    fp.close()
+    rootpath.joinpath("dir1").mkdir()
+    rootpath.joinpath("dir2").mkdir()
+    rootpath.joinpath("dir3").mkdir()
+    with rootpath.joinpath("file1.test").open("wt") as fp:
+        fp.write("1")
+    with rootpath.joinpath("file2.test").open("wt") as fp:
+        fp.write("12")
+    with rootpath.joinpath("file3.test").open("wt") as fp:
+        fp.write("123")
+    with rootpath.joinpath("dir1", "file1.test").open("wt") as fp:
+        fp.write("1")
+    with rootpath.joinpath("dir2", "file2.test").open("wt") as fp:
+        fp.write("12")
+    with rootpath.joinpath("dir3", "file3.test").open("wt") as fp:
+        fp.write("123")
     return rootpath
 
 
@@ -60,11 +54,10 @@ def setup_module(module):
     # and another with a more complex structure.
     testpath = Path(tempfile.mkdtemp())
     module.testpath = testpath
-    rootpath = testpath["onefile"]
+    rootpath = testpath.joinpath("onefile")
     rootpath.mkdir()
-    fp = rootpath["test.txt"].open("w")
-    fp.write("test_data")
-    fp.close()
+    with rootpath.joinpath("test.txt").open("wt") as fp:
+        fp.write("test_data")
     create_fake_fs(testpath)
 
 
@@ -80,13 +73,13 @@ def test_empty():
 
 def test_add_path():
     d = Directories()
-    p = testpath["onefile"]
+    p = testpath.joinpath("onefile")
     d.add_path(p)
     eq_(1, len(d))
     assert p in d
-    assert (p["foobar"]) in d
-    assert p.parent() not in d
-    p = testpath["fs"]
+    assert (p.joinpath("foobar")) in d
+    assert p.parent not in d
+    p = testpath.joinpath("fs")
     d.add_path(p)
     eq_(2, len(d))
     assert p in d
@@ -94,18 +87,18 @@ def test_add_path():
 
 def test_add_path_when_path_is_already_there():
     d = Directories()
-    p = testpath["onefile"]
+    p = testpath.joinpath("onefile")
     d.add_path(p)
     with raises(AlreadyThereError):
         d.add_path(p)
     with raises(AlreadyThereError):
-        d.add_path(p["foobar"])
+        d.add_path(p.joinpath("foobar"))
     eq_(1, len(d))
 
 
 def test_add_path_containing_paths_already_there():
     d = Directories()
-    d.add_path(testpath["onefile"])
+    d.add_path(testpath.joinpath("onefile"))
     eq_(1, len(d))
     d.add_path(testpath)
     eq_(len(d), 1)
@@ -114,7 +107,7 @@ def test_add_path_containing_paths_already_there():
 
 def test_add_path_non_latin(tmpdir):
     p = Path(str(tmpdir))
-    to_add = p["unicode\u201a"]
+    to_add = p.joinpath("unicode\u201a")
     os.mkdir(str(to_add))
     d = Directories()
     try:
@@ -125,25 +118,25 @@ def test_add_path_non_latin(tmpdir):
 
 def test_del():
     d = Directories()
-    d.add_path(testpath["onefile"])
+    d.add_path(testpath.joinpath("onefile"))
     try:
         del d[1]
         assert False
     except IndexError:
         pass
-    d.add_path(testpath["fs"])
+    d.add_path(testpath.joinpath("fs"))
     del d[1]
     eq_(1, len(d))
 
 
 def test_states():
     d = Directories()
-    p = testpath["onefile"]
+    p = testpath.joinpath("onefile")
     d.add_path(p)
     eq_(DirectoryState.NORMAL, d.get_state(p))
     d.set_state(p, DirectoryState.REFERENCE)
     eq_(DirectoryState.REFERENCE, d.get_state(p))
-    eq_(DirectoryState.REFERENCE, d.get_state(p["dir1"]))
+    eq_(DirectoryState.REFERENCE, d.get_state(p.joinpath("dir1")))
     eq_(1, len(d.states))
     eq_(p, list(d.states.keys())[0])
     eq_(DirectoryState.REFERENCE, d.states[p])
@@ -152,7 +145,7 @@ def test_states():
 def test_get_state_with_path_not_there():
     # When the path's not there, just return DirectoryState.Normal
     d = Directories()
-    d.add_path(testpath["onefile"])
+    d.add_path(testpath.joinpath("onefile"))
     eq_(d.get_state(testpath), DirectoryState.NORMAL)
 
 
@@ -160,26 +153,26 @@ def test_states_overwritten_when_larger_directory_eat_smaller_ones():
     # ref #248
     # When setting the state of a folder, we overwrite previously set states for subfolders.
     d = Directories()
-    p = testpath["onefile"]
+    p = testpath.joinpath("onefile")
     d.add_path(p)
     d.set_state(p, DirectoryState.EXCLUDED)
     d.add_path(testpath)
     d.set_state(testpath, DirectoryState.REFERENCE)
     eq_(d.get_state(p), DirectoryState.REFERENCE)
-    eq_(d.get_state(p["dir1"]), DirectoryState.REFERENCE)
+    eq_(d.get_state(p.joinpath("dir1")), DirectoryState.REFERENCE)
     eq_(d.get_state(testpath), DirectoryState.REFERENCE)
 
 
 def test_get_files():
     d = Directories()
-    p = testpath["fs"]
+    p = testpath.joinpath("fs")
     d.add_path(p)
-    d.set_state(p["dir1"], DirectoryState.REFERENCE)
-    d.set_state(p["dir2"], DirectoryState.EXCLUDED)
+    d.set_state(p.joinpath("dir1"), DirectoryState.REFERENCE)
+    d.set_state(p.joinpath("dir2"), DirectoryState.EXCLUDED)
     files = list(d.get_files())
     eq_(5, len(files))
     for f in files:
-        if f.path.parent() == p["dir1"]:
+        if f.path.parent == p.joinpath("dir1"):
             assert f.is_ref
         else:
             assert not f.is_ref
@@ -193,7 +186,7 @@ def test_get_files_with_folders():
             return True
 
     d = Directories()
-    p = testpath["fs"]
+    p = testpath.joinpath("fs")
     d.add_path(p)
     files = list(d.get_files(fileclasses=[FakeFile]))
     # We have the 3 root files and the 3 root dirs
@@ -202,23 +195,23 @@ def test_get_files_with_folders():
 
 def test_get_folders():
     d = Directories()
-    p = testpath["fs"]
+    p = testpath.joinpath("fs")
     d.add_path(p)
-    d.set_state(p["dir1"], DirectoryState.REFERENCE)
-    d.set_state(p["dir2"], DirectoryState.EXCLUDED)
+    d.set_state(p.joinpath("dir1"), DirectoryState.REFERENCE)
+    d.set_state(p.joinpath("dir2"), DirectoryState.EXCLUDED)
     folders = list(d.get_folders())
     eq_(len(folders), 3)
     ref = [f for f in folders if f.is_ref]
     not_ref = [f for f in folders if not f.is_ref]
     eq_(len(ref), 1)
-    eq_(ref[0].path, p["dir1"])
+    eq_(ref[0].path, p.joinpath("dir1"))
     eq_(len(not_ref), 2)
     eq_(ref[0].size, 1)
 
 
 def test_get_files_with_inherited_exclusion():
     d = Directories()
-    p = testpath["onefile"]
+    p = testpath.joinpath("onefile")
     d.add_path(p)
     d.set_state(p, DirectoryState.EXCLUDED)
     eq_([], list(d.get_files()))
@@ -234,13 +227,13 @@ def test_save_and_load(tmpdir):
     d1.add_path(p1)
     d1.add_path(p2)
     d1.set_state(p1, DirectoryState.REFERENCE)
-    d1.set_state(p1["dir1"], DirectoryState.EXCLUDED)
+    d1.set_state(p1.joinpath("dir1"), DirectoryState.EXCLUDED)
     tmpxml = str(tmpdir.join("directories_testunit.xml"))
     d1.save_to_file(tmpxml)
     d2.load_from_file(tmpxml)
     eq_(2, len(d2))
     eq_(DirectoryState.REFERENCE, d2.get_state(p1))
-    eq_(DirectoryState.EXCLUDED, d2.get_state(p1["dir1"]))
+    eq_(DirectoryState.EXCLUDED, d2.get_state(p1.joinpath("dir1")))
 
 
 def test_invalid_path():
@@ -268,7 +261,7 @@ def test_load_from_file_with_invalid_path(tmpdir):
     # This test simulates a load from file resulting in a
     # InvalidPath raise. Other directories must be loaded.
     d1 = Directories()
-    d1.add_path(testpath["onefile"])
+    d1.add_path(testpath.joinpath("onefile"))
     # Will raise InvalidPath upon loading
     p = Path(str(tmpdir.join("toremove")))
     p.mkdir()
@@ -283,11 +276,11 @@ def test_load_from_file_with_invalid_path(tmpdir):
 
 def test_unicode_save(tmpdir):
     d = Directories()
-    p1 = Path(str(tmpdir))["hello\xe9"]
+    p1 = Path(str(tmpdir), "hello\xe9")
     p1.mkdir()
-    p1["foo\xe9"].mkdir()
+    p1.joinpath("foo\xe9").mkdir()
     d.add_path(p1)
-    d.set_state(p1["foo\xe9"], DirectoryState.EXCLUDED)
+    d.set_state(p1.joinpath("foo\xe9"), DirectoryState.EXCLUDED)
     tmpxml = str(tmpdir.join("directories_testunit.xml"))
     try:
         d.save_to_file(tmpxml)
@@ -297,12 +290,12 @@ def test_unicode_save(tmpdir):
 
 def test_get_files_refreshes_its_directories():
     d = Directories()
-    p = testpath["fs"]
+    p = testpath.joinpath("fs")
     d.add_path(p)
     files = d.get_files()
     eq_(6, len(list(files)))
     time.sleep(1)
-    os.remove(str(p["dir1"]["file1.test"]))
+    os.remove(str(p.joinpath("dir1", "file1.test")))
     files = d.get_files()
     eq_(5, len(list(files)))
 
@@ -311,15 +304,15 @@ def test_get_files_does_not_choke_on_non_existing_directories(tmpdir):
     d = Directories()
     p = Path(str(tmpdir))
     d.add_path(p)
-    p.rmtree()
+    shutil.rmtree(str(p))
     eq_([], list(d.get_files()))
 
 
 def test_get_state_returns_excluded_by_default_for_hidden_directories(tmpdir):
     d = Directories()
     p = Path(str(tmpdir))
-    hidden_dir_path = p[".foo"]
-    p[".foo"].mkdir()
+    hidden_dir_path = p.joinpath(".foo")
+    p.joinpath(".foo").mkdir()
     d.add_path(p)
     eq_(d.get_state(hidden_dir_path), DirectoryState.EXCLUDED)
     # But it can be overriden
@@ -331,22 +324,23 @@ def test_default_path_state_override(tmpdir):
     # It's possible for a subclass to override the default state of a path
     class MyDirectories(Directories):
         def _default_state_for_path(self, path):
-            if "foobar" in path:
+            if "foobar" in path.parts:
                 return DirectoryState.EXCLUDED
+            return DirectoryState.NORMAL
 
     d = MyDirectories()
     p1 = Path(str(tmpdir))
-    p1["foobar"].mkdir()
-    p1["foobar/somefile"].open("w").close()
-    p1["foobaz"].mkdir()
-    p1["foobaz/somefile"].open("w").close()
+    p1.joinpath("foobar").mkdir()
+    p1.joinpath("foobar/somefile").touch()
+    p1.joinpath("foobaz").mkdir()
+    p1.joinpath("foobaz/somefile").touch()
     d.add_path(p1)
-    eq_(d.get_state(p1["foobaz"]), DirectoryState.NORMAL)
-    eq_(d.get_state(p1["foobar"]), DirectoryState.EXCLUDED)
+    eq_(d.get_state(p1.joinpath("foobaz")), DirectoryState.NORMAL)
+    eq_(d.get_state(p1.joinpath("foobar")), DirectoryState.EXCLUDED)
     eq_(len(list(d.get_files())), 1)  # only the 'foobaz' file is there
     # However, the default state can be changed
-    d.set_state(p1["foobar"], DirectoryState.NORMAL)
-    eq_(d.get_state(p1["foobar"]), DirectoryState.NORMAL)
+    d.set_state(p1.joinpath("foobar"), DirectoryState.NORMAL)
+    eq_(d.get_state(p1.joinpath("foobar")), DirectoryState.NORMAL)
     eq_(len(list(d.get_files())), 2)
 
 
@@ -372,42 +366,42 @@ files: {self.d._exclude_list.compiled_files} all: {self.d._exclude_list.compiled
         self.d._exclude_list.add(regex)
         self.d._exclude_list.mark(regex)
         p1 = Path(str(tmpdir))
-        p1["$Recycle.Bin"].mkdir()
-        p1["$Recycle.Bin"]["subdir"].mkdir()
+        p1.joinpath("$Recycle.Bin").mkdir()
+        p1.joinpath("$Recycle.Bin", "subdir").mkdir()
         self.d.add_path(p1)
-        eq_(self.d.get_state(p1["$Recycle.Bin"]), DirectoryState.EXCLUDED)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin")), DirectoryState.EXCLUDED)
         # By default, subdirs should be excluded too, but this can be overridden separately
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdir"]), DirectoryState.EXCLUDED)
-        self.d.set_state(p1["$Recycle.Bin"]["subdir"], DirectoryState.NORMAL)
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdir"]), DirectoryState.NORMAL)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdir")), DirectoryState.EXCLUDED)
+        self.d.set_state(p1.joinpath("$Recycle.Bin", "subdir"), DirectoryState.NORMAL)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdir")), DirectoryState.NORMAL)
 
     def test_exclude_refined(self, tmpdir):
         regex1 = r"^\$Recycle\.Bin$"
         self.d._exclude_list.add(regex1)
         self.d._exclude_list.mark(regex1)
         p1 = Path(str(tmpdir))
-        p1["$Recycle.Bin"].mkdir()
-        p1["$Recycle.Bin"]["somefile.png"].open("w").close()
-        p1["$Recycle.Bin"]["some_unwanted_file.jpg"].open("w").close()
-        p1["$Recycle.Bin"]["subdir"].mkdir()
-        p1["$Recycle.Bin"]["subdir"]["somesubdirfile.png"].open("w").close()
-        p1["$Recycle.Bin"]["subdir"]["unwanted_subdirfile.gif"].open("w").close()
-        p1["$Recycle.Bin"]["subdar"].mkdir()
-        p1["$Recycle.Bin"]["subdar"]["somesubdarfile.jpeg"].open("w").close()
-        p1["$Recycle.Bin"]["subdar"]["unwanted_subdarfile.png"].open("w").close()
-        self.d.add_path(p1["$Recycle.Bin"])
+        p1.joinpath("$Recycle.Bin").mkdir()
+        p1.joinpath("$Recycle.Bin", "somefile.png").touch()
+        p1.joinpath("$Recycle.Bin", "some_unwanted_file.jpg").touch()
+        p1.joinpath("$Recycle.Bin", "subdir").mkdir()
+        p1.joinpath("$Recycle.Bin", "subdir", "somesubdirfile.png").touch()
+        p1.joinpath("$Recycle.Bin", "subdir", "unwanted_subdirfile.gif").touch()
+        p1.joinpath("$Recycle.Bin", "subdar").mkdir()
+        p1.joinpath("$Recycle.Bin", "subdar", "somesubdarfile.jpeg").touch()
+        p1.joinpath("$Recycle.Bin", "subdar", "unwanted_subdarfile.png").touch()
+        self.d.add_path(p1.joinpath("$Recycle.Bin"))
 
         # Filter should set the default state to Excluded
-        eq_(self.d.get_state(p1["$Recycle.Bin"]), DirectoryState.EXCLUDED)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin")), DirectoryState.EXCLUDED)
         # The subdir should inherit its parent state
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdir"]), DirectoryState.EXCLUDED)
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdar"]), DirectoryState.EXCLUDED)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdir")), DirectoryState.EXCLUDED)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdar")), DirectoryState.EXCLUDED)
         # Override a child path's state
-        self.d.set_state(p1["$Recycle.Bin"]["subdir"], DirectoryState.NORMAL)
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdir"]), DirectoryState.NORMAL)
+        self.d.set_state(p1.joinpath("$Recycle.Bin", "subdir"), DirectoryState.NORMAL)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdir")), DirectoryState.NORMAL)
         # Parent should keep its default state, and the other child too
-        eq_(self.d.get_state(p1["$Recycle.Bin"]), DirectoryState.EXCLUDED)
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdar"]), DirectoryState.EXCLUDED)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin")), DirectoryState.EXCLUDED)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdar")), DirectoryState.EXCLUDED)
         # print(f"get_folders(): {[x for x in self.d.get_folders()]}")
 
         # only the 2 files directly under the Normal directory
@@ -419,8 +413,8 @@ files: {self.d._exclude_list.compiled_files} all: {self.d._exclude_list.compiled
         assert "somesubdirfile.png" in files
         assert "unwanted_subdirfile.gif" in files
         # Overriding the parent should enable all children
-        self.d.set_state(p1["$Recycle.Bin"], DirectoryState.NORMAL)
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdar"]), DirectoryState.NORMAL)
+        self.d.set_state(p1.joinpath("$Recycle.Bin"), DirectoryState.NORMAL)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdar")), DirectoryState.NORMAL)
         # all files there
         files = self.get_files_and_expect_num_result(6)
         assert "somefile.png" in files
@@ -444,7 +438,7 @@ files: {self.d._exclude_list.compiled_files} all: {self.d._exclude_list.compiled
         assert self.d._exclude_list.error(regex3) is None
         # print(f"get_folders(): {[x for x in self.d.get_folders()]}")
         # Directory shouldn't change its state here, unless explicitely done by user
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdir"]), DirectoryState.NORMAL)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdir")), DirectoryState.NORMAL)
         files = self.get_files_and_expect_num_result(5)
         assert "unwanted_subdirfile.gif" not in files
         assert "unwanted_subdarfile.png" in files
@@ -453,15 +447,15 @@ files: {self.d._exclude_list.compiled_files} all: {self.d._exclude_list.compiled
         regex4 = r".*subdir$"
         self.d._exclude_list.rename(regex3, regex4)
         assert self.d._exclude_list.error(regex4) is None
-        p1["$Recycle.Bin"]["subdar"]["file_ending_with_subdir"].open("w").close()
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdir"]), DirectoryState.EXCLUDED)
+        p1.joinpath("$Recycle.Bin", "subdar", "file_ending_with_subdir").touch()
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdir")), DirectoryState.EXCLUDED)
         files = self.get_files_and_expect_num_result(4)
         assert "file_ending_with_subdir" not in files
         assert "somesubdarfile.jpeg" in files
         assert "somesubdirfile.png" not in files
         assert "unwanted_subdirfile.gif" not in files
-        self.d.set_state(p1["$Recycle.Bin"]["subdir"], DirectoryState.NORMAL)
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdir"]), DirectoryState.NORMAL)
+        self.d.set_state(p1.joinpath("$Recycle.Bin", "subdir"), DirectoryState.NORMAL)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdir")), DirectoryState.NORMAL)
         # print(f"get_folders(): {[x for x in self.d.get_folders()]}")
         files = self.get_files_and_expect_num_result(6)
         assert "file_ending_with_subdir" not in files
@@ -471,9 +465,9 @@ files: {self.d._exclude_list.compiled_files} all: {self.d._exclude_list.compiled
         regex5 = r".*subdir.*"
         self.d._exclude_list.rename(regex4, regex5)
         # Files containing substring should be filtered
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdir"]), DirectoryState.NORMAL)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdir")), DirectoryState.NORMAL)
         # The path should not match, only the filename, the "subdir" in the directory name shouldn't matter
-        p1["$Recycle.Bin"]["subdir"]["file_which_shouldnt_match"].open("w").close()
+        p1.joinpath("$Recycle.Bin", "subdir", "file_which_shouldnt_match").touch()
         files = self.get_files_and_expect_num_result(5)
         assert "somesubdirfile.png" not in files
         assert "unwanted_subdirfile.gif" not in files
@@ -493,7 +487,7 @@ files: {self.d._exclude_list.compiled_files} all: {self.d._exclude_list.compiled
         assert self.d._exclude_list.error(regex6) is None
         assert regex6 in self.d._exclude_list
         # This still should not be affected
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["subdir"]), DirectoryState.NORMAL)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "subdir")), DirectoryState.NORMAL)
         files = self.get_files_and_expect_num_result(5)
         # These files are under the "/subdir" directory
         assert "somesubdirfile.png" not in files
@@ -505,20 +499,20 @@ files: {self.d._exclude_list.compiled_files} all: {self.d._exclude_list.compiled
 
     def test_japanese_unicode(self, tmpdir):
         p1 = Path(str(tmpdir))
-        p1["$Recycle.Bin"].mkdir()
-        p1["$Recycle.Bin"]["somerecycledfile.png"].open("w").close()
-        p1["$Recycle.Bin"]["some_unwanted_file.jpg"].open("w").close()
-        p1["$Recycle.Bin"]["subdir"].mkdir()
-        p1["$Recycle.Bin"]["subdir"]["過去白濁物語～]_カラー.jpg"].open("w").close()
-        p1["$Recycle.Bin"]["思叫物語"].mkdir()
-        p1["$Recycle.Bin"]["思叫物語"]["なししろ会う前"].open("w").close()
-        p1["$Recycle.Bin"]["思叫物語"]["堂～ロ"].open("w").close()
-        self.d.add_path(p1["$Recycle.Bin"])
+        p1.joinpath("$Recycle.Bin").mkdir()
+        p1.joinpath("$Recycle.Bin", "somerecycledfile.png").touch()
+        p1.joinpath("$Recycle.Bin", "some_unwanted_file.jpg").touch()
+        p1.joinpath("$Recycle.Bin", "subdir").mkdir()
+        p1.joinpath("$Recycle.Bin", "subdir", "過去白濁物語～]_カラー.jpg").touch()
+        p1.joinpath("$Recycle.Bin", "思叫物語").mkdir()
+        p1.joinpath("$Recycle.Bin", "思叫物語", "なししろ会う前").touch()
+        p1.joinpath("$Recycle.Bin", "思叫物語", "堂～ロ").touch()
+        self.d.add_path(p1.joinpath("$Recycle.Bin"))
         regex3 = r".*物語.*"
         self.d._exclude_list.add(regex3)
         self.d._exclude_list.mark(regex3)
         # print(f"get_folders(): {[x for x in self.d.get_folders()]}")
-        eq_(self.d.get_state(p1["$Recycle.Bin"]["思叫物語"]), DirectoryState.EXCLUDED)
+        eq_(self.d.get_state(p1.joinpath("$Recycle.Bin", "思叫物語")), DirectoryState.EXCLUDED)
         files = self.get_files_and_expect_num_result(2)
         assert "過去白濁物語～]_カラー.jpg" not in files
         assert "なししろ会う前" not in files
@@ -527,7 +521,7 @@ files: {self.d._exclude_list.compiled_files} all: {self.d._exclude_list.compiled
         regex4 = r".*物語$"
         self.d._exclude_list.rename(regex3, regex4)
         assert self.d._exclude_list.error(regex4) is None
-        self.d.set_state(p1["$Recycle.Bin"]["思叫物語"], DirectoryState.NORMAL)
+        self.d.set_state(p1.joinpath("$Recycle.Bin", "思叫物語"), DirectoryState.NORMAL)
         files = self.get_files_and_expect_num_result(5)
         assert "過去白濁物語～]_カラー.jpg" in files
         assert "なししろ会う前" in files
@@ -539,15 +533,15 @@ files: {self.d._exclude_list.compiled_files} all: {self.d._exclude_list.compiled
         self.d._exclude_list.add(regex)
         self.d._exclude_list.mark(regex)
         p1 = Path(str(tmpdir))
-        p1["foobar"].mkdir()
-        p1["foobar"][".hidden_file.txt"].open("w").close()
-        p1["foobar"][".hidden_dir"].mkdir()
-        p1["foobar"][".hidden_dir"]["foobar.jpg"].open("w").close()
-        p1["foobar"][".hidden_dir"][".hidden_subfile.png"].open("w").close()
-        self.d.add_path(p1["foobar"])
+        p1.joinpath("foobar").mkdir()
+        p1.joinpath("foobar", ".hidden_file.txt").touch()
+        p1.joinpath("foobar", ".hidden_dir").mkdir()
+        p1.joinpath("foobar", ".hidden_dir", "foobar.jpg").touch()
+        p1.joinpath("foobar", ".hidden_dir", ".hidden_subfile.png").touch()
+        self.d.add_path(p1.joinpath("foobar"))
         # It should not inherit its parent's state originally
-        eq_(self.d.get_state(p1["foobar"][".hidden_dir"]), DirectoryState.EXCLUDED)
-        self.d.set_state(p1["foobar"][".hidden_dir"], DirectoryState.NORMAL)
+        eq_(self.d.get_state(p1.joinpath("foobar", ".hidden_dir")), DirectoryState.EXCLUDED)
+        self.d.set_state(p1.joinpath("foobar", ".hidden_dir"), DirectoryState.NORMAL)
         # The files should still be filtered
         files = self.get_files_and_expect_num_result(1)
         eq_(len(self.d._exclude_list.compiled_paths), 0)

core/tests/engine_test.py

@@ -10,9 +10,9 @@ from hscommon.jobprogress import job
 from hscommon.util import first
 from hscommon.testutil import eq_, log_calls
 
-from .base import NamedObject
-from .. import engine
-from ..engine import (
+from core.tests.base import NamedObject
+from core import engine
+from core.engine import (
     get_match,
     getwords,
     Group,
@@ -71,7 +71,10 @@ class TestCasegetwords:
 
     def test_unicode(self):
         eq_(["e", "c", "0", "a", "o", "u", "e", "u"], getwords("é ç 0 à ö û è ¤ ù"))
-        eq_(["02", "君のこころは輝いてるかい？", "国木田花丸", "solo", "ver"], getwords("02 君のこころは輝いてるかい？ 国木田花丸 Solo Ver"))
+        eq_(
+            ["02", "君のこころは輝いてるかい？", "国木田花丸", "solo", "ver"],
+            getwords("02 君のこころは輝いてるかい？ 国木田花丸 Solo Ver"),
+        )
 
     def test_splitter_chars(self):
         eq_(
@@ -271,9 +274,9 @@ class TestCaseBuildWordDict:
 class TestCaseMergeSimilarWords:
     def test_some_similar_words(self):
         d = {
-            "foobar": set([1]),
-            "foobar1": set([2]),
-            "foobar2": set([3]),
+            "foobar": {1},
+            "foobar1": {2},
+            "foobar2": {3},
         }
         merge_similar_words(d)
         eq_(1, len(d))
@@ -283,8 +286,8 @@ class TestCaseMergeSimilarWords:
 class TestCaseReduceCommonWords:
     def test_typical(self):
         d = {
-            "foo": set([NamedObject("foo bar", True) for _ in range(50)]),
-            "bar": set([NamedObject("foo bar", True) for _ in range(49)]),
+            "foo": {NamedObject("foo bar", True) for _ in range(50)},
+            "bar": {NamedObject("foo bar", True) for _ in range(49)},
         }
         reduce_common_words(d, 50)
         assert "foo" not in d
@@ -293,7 +296,7 @@ class TestCaseReduceCommonWords:
     def test_dont_remove_objects_with_only_common_words(self):
         d = {
             "common": set([NamedObject("common uncommon", True) for _ in range(50)] + [NamedObject("common", True)]),
-            "uncommon": set([NamedObject("common uncommon", True)]),
+            "uncommon": {NamedObject("common uncommon", True)},
         }
         reduce_common_words(d, 50)
         eq_(1, len(d["common"]))
@@ -302,7 +305,7 @@ class TestCaseReduceCommonWords:
     def test_values_still_are_set_instances(self):
         d = {
             "common": set([NamedObject("common uncommon", True) for _ in range(50)] + [NamedObject("common", True)]),
-            "uncommon": set([NamedObject("common uncommon", True)]),
+            "uncommon": {NamedObject("common uncommon", True)},
         }
         reduce_common_words(d, 50)
         assert isinstance(d["common"], set)
@@ -312,9 +315,9 @@ class TestCaseReduceCommonWords:
         # If a word has been removed by the reduce, an object in a subsequent common word that
         # contains the word that has been removed would cause a KeyError.
         d = {
-            "foo": set([NamedObject("foo bar baz", True) for _ in range(50)]),
-            "bar": set([NamedObject("foo bar baz", True) for _ in range(50)]),
-            "baz": set([NamedObject("foo bar baz", True) for _ in range(49)]),
+            "foo": {NamedObject("foo bar baz", True) for _ in range(50)},
+            "bar": {NamedObject("foo bar baz", True) for _ in range(50)},
+            "baz": {NamedObject("foo bar baz", True) for _ in range(49)},
         }
         try:
             reduce_common_words(d, 50)
@@ -328,7 +331,7 @@ class TestCaseReduceCommonWords:
             o.words = [["foo", "bar"], ["baz"]]
             return o
 
-        d = {"foo": set([create_it() for _ in range(50)])}
+        d = {"foo": {create_it() for _ in range(50)}}
         try:
             reduce_common_words(d, 50)
         except TypeError:
@@ -343,7 +346,7 @@ class TestCaseReduceCommonWords:
         d = {
             "foo": set([NamedObject("foo bar baz", True) for _ in range(49)] + [only_common]),
             "bar": set([NamedObject("foo bar baz", True) for _ in range(49)] + [only_common]),
-            "baz": set([NamedObject("foo bar baz", True) for _ in range(49)]),
+            "baz": {NamedObject("foo bar baz", True) for _ in range(49)},
         }
         reduce_common_words(d, 50)
         eq_(1, len(d["foo"]))
@@ -530,7 +533,7 @@ class TestCaseGetMatches:
 
 
 class TestCaseGetMatchesByContents:
-    def test_big_file_partial_hashes(self):
+    def test_big_file_partial_hashing(self):
         smallsize = 1
         bigsize = 100 * 1024 * 1024  # 100MB
         f = [
@@ -539,17 +542,17 @@ class TestCaseGetMatchesByContents:
             no("smallfoo", size=smallsize),
             no("smallbar", size=smallsize),
         ]
-        f[0].md5 = f[0].md5partial = f[0].md5samples = "foobar"
-        f[1].md5 = f[1].md5partial = f[1].md5samples = "foobar"
-        f[2].md5 = f[2].md5partial = "bleh"
-        f[3].md5 = f[3].md5partial = "bleh"
+        f[0].digest = f[0].digest_partial = f[0].digest_samples = "foobar"
+        f[1].digest = f[1].digest_partial = f[1].digest_samples = "foobar"
+        f[2].digest = f[2].digest_partial = "bleh"
+        f[3].digest = f[3].digest_partial = "bleh"
         r = getmatches_by_contents(f, bigsize=bigsize)
         eq_(len(r), 2)
-        # User disabled optimization for big files, compute hashes as usual
+        # User disabled optimization for big files, compute digests as usual
         r = getmatches_by_contents(f, bigsize=0)
         eq_(len(r), 2)
-        # Other file is now slightly different, md5partial is still the same
-        f[1].md5 = f[1].md5samples = "foobardiff"
+        # Other file is now slightly different, digest_partial is still the same
+        f[1].digest = f[1].digest_samples = "foobardiff"
         r = getmatches_by_contents(f, bigsize=bigsize)
         # Successfully filter it out
         eq_(len(r), 1)
@@ -884,7 +887,7 @@ class TestCaseGetGroups:
         # If, with a (A, B, C, D) set, all match with A, but C and D don't match with B and that the
         # (A, B) match is the highest (thus resulting in an (A, B) group), still match C and D
         # in a separate group instead of discarding them.
-        A, B, C, D = [NamedObject() for _ in range(4)]
+        A, B, C, D = (NamedObject() for _ in range(4))
         m1 = Match(A, B, 90)  # This is the strongest "A" match
         m2 = Match(A, C, 80)  # Because C doesn't match with B, it won't be in the group
         m3 = Match(A, D, 80)  # Same thing for D

core/tests/exclude_test.py

@@ -10,8 +10,8 @@ from xml.etree import ElementTree as ET
 from hscommon.testutil import eq_
 from hscommon.plat import ISWINDOWS
 
-from .base import DupeGuru
-from ..exclude import ExcludeList, ExcludeDict, default_regexes, AlreadyThereException
+from core.tests.base import DupeGuru
+from core.exclude import ExcludeList, ExcludeDict, default_regexes, AlreadyThereException
 
 from re import error
 
@@ -289,8 +289,8 @@ class TestCaseListEmptyUnion(TestCaseListEmpty):
         compiled = [x for x in self.exclude_list.compiled]
         assert regex not in compiled
         # Need to escape both to get the same strings after compilation
-        compiled_escaped = set([x.encode("unicode-escape").decode() for x in compiled[0].pattern.split("|")])
-        default_escaped = set([x.encode("unicode-escape").decode() for x in default_regexes])
+        compiled_escaped = {x.encode("unicode-escape").decode() for x in compiled[0].pattern.split("|")}
+        default_escaped = {x.encode("unicode-escape").decode() for x in default_regexes}
         assert compiled_escaped == default_escaped
         eq_(len(default_regexes), len(compiled[0].pattern.split("|")))
 
@@ -366,8 +366,8 @@ class TestCaseDictEmptyUnion(TestCaseDictEmpty):
         compiled = [x for x in self.exclude_list.compiled]
         assert regex not in compiled
         # Need to escape both to get the same strings after compilation
-        compiled_escaped = set([x.encode("unicode-escape").decode() for x in compiled[0].pattern.split("|")])
-        default_escaped = set([x.encode("unicode-escape").decode() for x in default_regexes])
+        compiled_escaped = {x.encode("unicode-escape").decode() for x in compiled[0].pattern.split("|")}
+        default_escaped = {x.encode("unicode-escape").decode() for x in default_regexes}
         assert compiled_escaped == default_escaped
         eq_(len(default_regexes), len(compiled[0].pattern.split("|")))
 

core/tests/fs_test.py

@@ -6,43 +6,47 @@
 # which should be included with this package. The terms are also available at
 # http://www.gnu.org/licenses/gpl-3.0.html
 
-import hashlib
+import typing
 from os import urandom
 
-from hscommon.path import Path
+from pathlib import Path
 from hscommon.testutil import eq_
 from core.tests.directories_test import create_fake_fs
 
-from .. import fs
+from core import fs
+
+hasher: typing.Callable
+try:
+    import xxhash
+
+    hasher = xxhash.xxh128
+except ImportError:
+    import hashlib
+
+    hasher = hashlib.md5
 
 
 def create_fake_fs_with_random_data(rootpath):
-    rootpath = rootpath["fs"]
+    rootpath = rootpath.joinpath("fs")
     rootpath.mkdir()
-    rootpath["dir1"].mkdir()
-    rootpath["dir2"].mkdir()
-    rootpath["dir3"].mkdir()
-    fp = rootpath["file1.test"].open("wb")
+    rootpath.joinpath("dir1").mkdir()
+    rootpath.joinpath("dir2").mkdir()
+    rootpath.joinpath("dir3").mkdir()
     data1 = urandom(200 * 1024)  # 200KiB
     data2 = urandom(1024 * 1024)  # 1MiB
     data3 = urandom(10 * 1024 * 1024)  # 10MiB
-    fp.write(data1)
-    fp.close()
-    fp = rootpath["file2.test"].open("wb")
-    fp.write(data2)
-    fp.close()
-    fp = rootpath["file3.test"].open("wb")
-    fp.write(data3)
-    fp.close()
-    fp = rootpath["dir1"]["file1.test"].open("wb")
-    fp.write(data1)
-    fp.close()
-    fp = rootpath["dir2"]["file2.test"].open("wb")
-    fp.write(data2)
-    fp.close()
-    fp = rootpath["dir3"]["file3.test"].open("wb")
-    fp.write(data3)
-    fp.close()
+    with rootpath.joinpath("file1.test").open("wb") as fp:
+        fp.write(data1)
+    with rootpath.joinpath("file2.test").open("wb") as fp:
+        fp.write(data2)
+    with rootpath.joinpath("file3.test").open("wb") as fp:
+        fp.write(data3)
+    with rootpath.joinpath("dir1", "file1.test").open("wb") as fp:
+        fp.write(data1)
+    with rootpath.joinpath("dir2", "file2.test").open("wb") as fp:
+        fp.write(data2)
+    with rootpath.joinpath("dir3", "file3.test").open("wb") as fp:
+        fp.write(data3)
     return rootpath
 
 
@@ -52,54 +56,54 @@ def test_size_aggregates_subfiles(tmpdir):
     eq_(b.size, 12)
 
 
-def test_md5_aggregate_subfiles_sorted(tmpdir):
-    # dir.allfiles can return child in any order. Thus, bundle.md5 must aggregate
-    # all files' md5 it contains, but it must make sure that it does so in the
+def test_digest_aggregate_subfiles_sorted(tmpdir):
+    # dir.allfiles can return child in any order. Thus, bundle.digest must aggregate
+    # all files' digests it contains, but it must make sure that it does so in the
     # same order everytime.
     p = create_fake_fs_with_random_data(Path(str(tmpdir)))
     b = fs.Folder(p)
-    md51 = fs.File(p["dir1"]["file1.test"]).md5
-    md52 = fs.File(p["dir2"]["file2.test"]).md5
-    md53 = fs.File(p["dir3"]["file3.test"]).md5
-    md54 = fs.File(p["file1.test"]).md5
-    md55 = fs.File(p["file2.test"]).md5
-    md56 = fs.File(p["file3.test"]).md5
-    # The expected md5 is the md5 of md5s for folders and the direct md5 for files
-    folder_md51 = hashlib.md5(md51).digest()
-    folder_md52 = hashlib.md5(md52).digest()
-    folder_md53 = hashlib.md5(md53).digest()
-    md5 = hashlib.md5(folder_md51 + folder_md52 + folder_md53 + md54 + md55 + md56)
-    eq_(b.md5, md5.digest())
+    digest1 = fs.File(p.joinpath("dir1", "file1.test")).digest
+    digest2 = fs.File(p.joinpath("dir2", "file2.test")).digest
+    digest3 = fs.File(p.joinpath("dir3", "file3.test")).digest
+    digest4 = fs.File(p.joinpath("file1.test")).digest
+    digest5 = fs.File(p.joinpath("file2.test")).digest
+    digest6 = fs.File(p.joinpath("file3.test")).digest
+    # The expected digest is the hash of digests for folders and the direct digest for files
+    folder_digest1 = hasher(digest1).digest()
+    folder_digest2 = hasher(digest2).digest()
+    folder_digest3 = hasher(digest3).digest()
+    digest = hasher(folder_digest1 + folder_digest2 + folder_digest3 + digest4 + digest5 + digest6).digest()
+    eq_(b.digest, digest)
 
 
-def test_partial_md5_aggregate_subfile_sorted(tmpdir):
+def test_partial_digest_aggregate_subfile_sorted(tmpdir):
     p = create_fake_fs_with_random_data(Path(str(tmpdir)))
     b = fs.Folder(p)
-    md51 = fs.File(p["dir1"]["file1.test"]).md5partial
-    md52 = fs.File(p["dir2"]["file2.test"]).md5partial
-    md53 = fs.File(p["dir3"]["file3.test"]).md5partial
-    md54 = fs.File(p["file1.test"]).md5partial
-    md55 = fs.File(p["file2.test"]).md5partial
-    md56 = fs.File(p["file3.test"]).md5partial
-    # The expected md5 is the md5 of md5s for folders and the direct md5 for files
-    folder_md51 = hashlib.md5(md51).digest()
-    folder_md52 = hashlib.md5(md52).digest()
-    folder_md53 = hashlib.md5(md53).digest()
-    md5 = hashlib.md5(folder_md51 + folder_md52 + folder_md53 + md54 + md55 + md56)
-    eq_(b.md5partial, md5.digest())
+    digest1 = fs.File(p.joinpath("dir1", "file1.test")).digest_partial
+    digest2 = fs.File(p.joinpath("dir2", "file2.test")).digest_partial
+    digest3 = fs.File(p.joinpath("dir3", "file3.test")).digest_partial
+    digest4 = fs.File(p.joinpath("file1.test")).digest_partial
+    digest5 = fs.File(p.joinpath("file2.test")).digest_partial
+    digest6 = fs.File(p.joinpath("file3.test")).digest_partial
+    # The expected digest is the hash of digests for folders and the direct digest for files
+    folder_digest1 = hasher(digest1).digest()
+    folder_digest2 = hasher(digest2).digest()
+    folder_digest3 = hasher(digest3).digest()
+    digest = hasher(folder_digest1 + folder_digest2 + folder_digest3 + digest4 + digest5 + digest6).digest()
+    eq_(b.digest_partial, digest)
 
-    md51 = fs.File(p["dir1"]["file1.test"]).md5samples
-    md52 = fs.File(p["dir2"]["file2.test"]).md5samples
-    md53 = fs.File(p["dir3"]["file3.test"]).md5samples
-    md54 = fs.File(p["file1.test"]).md5samples
-    md55 = fs.File(p["file2.test"]).md5samples
-    md56 = fs.File(p["file3.test"]).md5samples
-    # The expected md5 is the md5 of md5s for folders and the direct md5 for files
-    folder_md51 = hashlib.md5(md51).digest()
-    folder_md52 = hashlib.md5(md52).digest()
-    folder_md53 = hashlib.md5(md53).digest()
-    md5 = hashlib.md5(folder_md51 + folder_md52 + folder_md53 + md54 + md55 + md56)
-    eq_(b.md5samples, md5.digest())
+    digest1 = fs.File(p.joinpath("dir1", "file1.test")).digest_samples
+    digest2 = fs.File(p.joinpath("dir2", "file2.test")).digest_samples
+    digest3 = fs.File(p.joinpath("dir3", "file3.test")).digest_samples
+    digest4 = fs.File(p.joinpath("file1.test")).digest_samples
+    digest5 = fs.File(p.joinpath("file2.test")).digest_samples
+    digest6 = fs.File(p.joinpath("file3.test")).digest_samples
+    # The expected digest is the digest of digests for folders and the direct digest for files
+    folder_digest1 = hasher(digest1).digest()
+    folder_digest2 = hasher(digest2).digest()
+    folder_digest3 = hasher(digest3).digest()
+    digest = hasher(folder_digest1 + folder_digest2 + folder_digest3 + digest4 + digest5 + digest6).digest()
+    eq_(b.digest_samples, digest)
 
 
 def test_has_file_attrs(tmpdir):

core/tests/ignore_test.py

@@ -10,7 +10,7 @@ from xml.etree import ElementTree as ET
 from pytest import raises
 from hscommon.testutil import eq_
 
-from ..ignore import IgnoreList
+from core.ignore import IgnoreList
 
 
 def test_empty():

core/tests/markable_test.py

@@ -6,7 +6,7 @@
 
 from hscommon.testutil import eq_
 
-from ..markable import MarkableList, Markable
+from core.markable import MarkableList, Markable
 
 
 def gen():

core/tests/prioritize_test.py

@@ -9,8 +9,8 @@
 import os.path as op
 from itertools import combinations
 
-from .base import TestApp, NamedObject, with_app, eq_
-from ..engine import Group, Match
+from core.tests.base import TestApp, NamedObject, with_app, eq_
+from core.engine import Group, Match
 
 no = NamedObject
 

core/tests/result_table_test.py

@@ -6,7 +6,7 @@
 # which should be included with this package. The terms are also available at
 # http://www.gnu.org/licenses/gpl-3.0.html
 
-from .base import TestApp, GetTestGroups
+from core.tests.base import TestApp, GetTestGroups
 
 
 def app_with_results():

core/tests/results_test.py

@@ -12,10 +12,9 @@ from xml.etree import ElementTree as ET
 from pytest import raises
 from hscommon.testutil import eq_
 from hscommon.util import first
-
-from .. import engine
-from .base import NamedObject, GetTestGroups, DupeGuru
-from ..results import Results
+from core import engine
+from core.tests.base import NamedObject, GetTestGroups, DupeGuru
+from core.results import Results
 
 
 class TestCaseResultsEmpty:
@@ -337,7 +336,7 @@ class TestCaseResultsMarkings:
         def log_object(o):
             log.append(o)
             if o is self.objects[1]:
-                raise EnvironmentError("foobar")
+                raise OSError("foobar")
 
         log = []
         self.results.mark_all()
@@ -447,7 +446,7 @@ class TestCaseResultsXML:
         self.results.groups = self.groups
 
     def get_file(self, path):  # use this as a callback for load_from_xml
-        return [o for o in self.objects if o.path == path][0]
+        return [o for o in self.objects if str(o.path) == path][0]
 
     def test_save_to_xml(self):
         self.objects[0].is_ref = True
@@ -464,7 +463,7 @@ class TestCaseResultsXML:
         eq_(6, len(g1))
         eq_(3, len([c for c in g1 if c.tag == "file"]))
         eq_(3, len([c for c in g1 if c.tag == "match"]))
-        d1, d2, d3 = [c for c in g1 if c.tag == "file"]
+        d1, d2, d3 = (c for c in g1 if c.tag == "file")
         eq_(op.join("basepath", "foo bar"), d1.get("path"))
         eq_(op.join("basepath", "bar bleh"), d2.get("path"))
         eq_(op.join("basepath", "foo bleh"), d3.get("path"))
@@ -477,7 +476,7 @@ class TestCaseResultsXML:
         eq_(3, len(g2))
         eq_(2, len([c for c in g2 if c.tag == "file"]))
         eq_(1, len([c for c in g2 if c.tag == "match"]))
-        d1, d2 = [c for c in g2 if c.tag == "file"]
+        d1, d2 = (c for c in g2 if c.tag == "file")
         eq_(op.join("basepath", "ibabtu"), d1.get("path"))
         eq_(op.join("basepath", "ibabtu"), d2.get("path"))
         eq_("n", d1.get("is_ref"))

core/tests/scanner_test.py

@@ -7,29 +7,33 @@
 import pytest
 
 from hscommon.jobprogress import job
-from hscommon.path import Path
+from pathlib import Path
 from hscommon.testutil import eq_
 
-from .. import fs
-from ..engine import getwords, Match
-from ..ignore import IgnoreList
-from ..scanner import Scanner, ScanType
-from ..me.scanner import ScannerME
+from core import fs
+from core.engine import getwords, Match
+from core.ignore import IgnoreList
+from core.scanner import Scanner, ScanType
+from core.me.scanner import ScannerME
 
 
+# TODO update this to be able to inherit from fs.File
 class NamedObject:
     def __init__(self, name="foobar", size=1, path=None):
         if path is None:
             path = Path(name)
         else:
-            path = Path(path)[name]
+            path = Path(path, name)
         self.name = name
         self.size = size
         self.path = path
         self.words = getwords(name)
 
     def __repr__(self):
-        return "<NamedObject %r %r>" % (self.name, self.path)
+        return "<NamedObject {!r} {!r}>".format(self.name, self.path)
+
+    def exists(self):
+        return self.path.exists()
 
 
 no = NamedObject
@@ -123,19 +127,19 @@ def test_content_scan(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.CONTENTS
     f = [no("foo"), no("bar"), no("bleh")]
-    f[0].md5 = f[0].md5partial = f[0].md5samples = "foobar"
-    f[1].md5 = f[1].md5partial = f[1].md5samples = "foobar"
-    f[2].md5 = f[2].md5partial = f[1].md5samples = "bleh"
+    f[0].digest = f[0].digest_partial = f[0].digest_samples = "foobar"
+    f[1].digest = f[1].digest_partial = f[1].digest_samples = "foobar"
+    f[2].digest = f[2].digest_partial = f[1].digest_samples = "bleh"
     r = s.get_dupe_groups(f)
     eq_(len(r), 1)
     eq_(len(r[0]), 2)
-    eq_(s.discarded_file_count, 0)  # don't count the different md5 as discarded!
+    eq_(s.discarded_file_count, 0)  # don't count the different digest as discarded!
 
 
 def test_content_scan_compare_sizes_first(fake_fileexists):
     class MyFile(no):
         @property
-        def md5(self):
+        def digest(self):
             raise AssertionError()
 
     s = Scanner()
@@ -161,14 +165,14 @@ def test_ignore_file_size(fake_fileexists):
         no("largeignore1", large_size + 1),
         no("largeignore2", large_size + 1),
     ]
-    f[0].md5 = f[0].md5partial = f[0].md5samples = "smallignore"
-    f[1].md5 = f[1].md5partial = f[1].md5samples = "smallignore"
-    f[2].md5 = f[2].md5partial = f[2].md5samples = "small"
-    f[3].md5 = f[3].md5partial = f[3].md5samples = "small"
-    f[4].md5 = f[4].md5partial = f[4].md5samples = "large"
-    f[5].md5 = f[5].md5partial = f[5].md5samples = "large"
-    f[6].md5 = f[6].md5partial = f[6].md5samples = "largeignore"
-    f[7].md5 = f[7].md5partial = f[7].md5samples = "largeignore"
+    f[0].digest = f[0].digest_partial = f[0].digest_samples = "smallignore"
+    f[1].digest = f[1].digest_partial = f[1].digest_samples = "smallignore"
+    f[2].digest = f[2].digest_partial = f[2].digest_samples = "small"
+    f[3].digest = f[3].digest_partial = f[3].digest_samples = "small"
+    f[4].digest = f[4].digest_partial = f[4].digest_samples = "large"
+    f[5].digest = f[5].digest_partial = f[5].digest_samples = "large"
+    f[6].digest = f[6].digest_partial = f[6].digest_samples = "largeignore"
+    f[7].digest = f[7].digest_partial = f[7].digest_samples = "largeignore"
 
     r = s.get_dupe_groups(f)
     # No ignores
@@ -197,21 +201,21 @@ def test_big_file_partial_hashes(fake_fileexists):
     s.big_file_size_threshold = bigsize
 
     f = [no("bigfoo", bigsize), no("bigbar", bigsize), no("smallfoo", smallsize), no("smallbar", smallsize)]
-    f[0].md5 = f[0].md5partial = f[0].md5samples = "foobar"
-    f[1].md5 = f[1].md5partial = f[1].md5samples = "foobar"
-    f[2].md5 = f[2].md5partial = "bleh"
-    f[3].md5 = f[3].md5partial = "bleh"
+    f[0].digest = f[0].digest_partial = f[0].digest_samples = "foobar"
+    f[1].digest = f[1].digest_partial = f[1].digest_samples = "foobar"
+    f[2].digest = f[2].digest_partial = "bleh"
+    f[3].digest = f[3].digest_partial = "bleh"
     r = s.get_dupe_groups(f)
     eq_(len(r), 2)
 
-    # md5partial is still the same, but the file is actually different
-    f[1].md5 = f[1].md5samples = "difffoobar"
-    # here we compare the full md5s, as the user disabled the optimization
+    # digest_partial is still the same, but the file is actually different
+    f[1].digest = f[1].digest_samples = "difffoobar"
+    # here we compare the full digests, as the user disabled the optimization
     s.big_file_size_threshold = 0
     r = s.get_dupe_groups(f)
     eq_(len(r), 1)
 
-    # here we should compare the md5samples, and see they are different
+    # here we should compare the digest_samples, and see they are different
     s.big_file_size_threshold = bigsize
     r = s.get_dupe_groups(f)
     eq_(len(r), 1)
@@ -221,9 +225,9 @@ def test_min_match_perc_doesnt_matter_for_content_scan(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.CONTENTS
     f = [no("foo"), no("bar"), no("bleh")]
-    f[0].md5 = f[0].md5partial = f[0].md5samples = "foobar"
-    f[1].md5 = f[1].md5partial = f[1].md5samples = "foobar"
-    f[2].md5 = f[2].md5partial = f[2].md5samples = "bleh"
+    f[0].digest = f[0].digest_partial = f[0].digest_samples = "foobar"
+    f[1].digest = f[1].digest_partial = f[1].digest_samples = "foobar"
+    f[2].digest = f[2].digest_partial = f[2].digest_samples = "bleh"
     s.min_match_percentage = 101
     r = s.get_dupe_groups(f)
     eq_(len(r), 1)
@@ -234,12 +238,16 @@ def test_min_match_perc_doesnt_matter_for_content_scan(fake_fileexists):
     eq_(len(r[0]), 2)
 
 
-def test_content_scan_doesnt_put_md5_in_words_at_the_end(fake_fileexists):
+def test_content_scan_doesnt_put_digest_in_words_at_the_end(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.CONTENTS
     f = [no("foo"), no("bar")]
-    f[0].md5 = f[0].md5partial = f[0].md5samples = "\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0a\x0b\x0c\x0d\x0e\x0f"
-    f[1].md5 = f[1].md5partial = f[1].md5samples = "\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0a\x0b\x0c\x0d\x0e\x0f"
+    f[0].digest = f[0].digest_partial = f[0].digest_samples = (
+        "\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0a\x0b\x0c\x0d\x0e\x0f"
+    )
+    f[1].digest = f[1].digest_partial = f[1].digest_samples = (
+        "\x00\x01\x02\x03\x04\x05\x06\x07\x08\x09\x0a\x0b\x0c\x0d\x0e\x0f"
+    )
     r = s.get_dupe_groups(f)
     # FIXME looks like we are missing something here?
     r[0]
@@ -332,7 +340,7 @@ def test_tag_scan(fake_fileexists):
 def test_tag_with_album_scan(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.TAG
-    s.scanned_tags = set(["artist", "album", "title"])
+    s.scanned_tags = {"artist", "album", "title"}
     o1 = no("foo")
     o2 = no("bar")
     o3 = no("bleh")
@@ -352,7 +360,7 @@ def test_tag_with_album_scan(fake_fileexists):
 def test_that_dash_in_tags_dont_create_new_fields(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.TAG
-    s.scanned_tags = set(["artist", "album", "title"])
+    s.scanned_tags = {"artist", "album", "title"}
     s.min_match_percentage = 50
     o1 = no("foo")
     o2 = no("bar")
@@ -369,7 +377,7 @@ def test_that_dash_in_tags_dont_create_new_fields(fake_fileexists):
 def test_tag_scan_with_different_scanned(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.TAG
-    s.scanned_tags = set(["track", "year"])
+    s.scanned_tags = {"track", "year"}
     o1 = no("foo")
     o2 = no("bar")
     o1.artist = "The White Stripes"
@@ -387,7 +395,7 @@ def test_tag_scan_with_different_scanned(fake_fileexists):
 def test_tag_scan_only_scans_existing_tags(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.TAG
-    s.scanned_tags = set(["artist", "foo"])
+    s.scanned_tags = {"artist", "foo"}
     o1 = no("foo")
     o2 = no("bar")
     o1.artist = "The White Stripes"
@@ -401,7 +409,7 @@ def test_tag_scan_only_scans_existing_tags(fake_fileexists):
 def test_tag_scan_converts_to_str(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.TAG
-    s.scanned_tags = set(["track"])
+    s.scanned_tags = {"track"}
     o1 = no("foo")
     o2 = no("bar")
     o1.track = 42
@@ -416,7 +424,7 @@ def test_tag_scan_converts_to_str(fake_fileexists):
 def test_tag_scan_non_ascii(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.TAG
-    s.scanned_tags = set(["title"])
+    s.scanned_tags = {"title"}
     o1 = no("foo")
     o2 = no("bar")
     o1.title = "foobar\u00e9"
@@ -568,12 +576,14 @@ def test_dont_group_files_that_dont_exist(tmpdir):
     s = Scanner()
     s.scan_type = ScanType.CONTENTS
     p = Path(str(tmpdir))
-    p["file1"].open("w").write("foo")
-    p["file2"].open("w").write("foo")
+    with p.joinpath("file1").open("w") as fp:
+        fp.write("foo")
+    with p.joinpath("file2").open("w") as fp:
+        fp.write("foo")
     file1, file2 = fs.get_files(p)
 
     def getmatches(*args, **kw):
-        file2.path.remove()
+        file2.path.unlink()
         return [Match(file1, file2, 100)]
 
     s._getmatches = getmatches
@@ -587,21 +597,21 @@ def test_folder_scan_exclude_subfolder_matches(fake_fileexists):
     s = Scanner()
     s.scan_type = ScanType.FOLDERS
     topf1 = no("top folder 1", size=42)
-    topf1.md5 = topf1.md5partial = topf1.md5samples = b"some_md5_1"
+    topf1.digest = topf1.digest_partial = topf1.digest_samples = b"some_digest__1"
     topf1.path = Path("/topf1")
     topf2 = no("top folder 2", size=42)
-    topf2.md5 = topf2.md5partial = topf2.md5samples = b"some_md5_1"
+    topf2.digest = topf2.digest_partial = topf2.digest_samples = b"some_digest__1"
     topf2.path = Path("/topf2")
     subf1 = no("sub folder 1", size=41)
-    subf1.md5 = subf1.md5partial = subf1.md5samples = b"some_md5_2"
+    subf1.digest = subf1.digest_partial = subf1.digest_samples = b"some_digest__2"
     subf1.path = Path("/topf1/sub")
     subf2 = no("sub folder 2", size=41)
-    subf2.md5 = subf2.md5partial = subf2.md5samples = b"some_md5_2"
+    subf2.digest = subf2.digest_partial = subf2.digest_samples = b"some_digest__2"
     subf2.path = Path("/topf2/sub")
     eq_(len(s.get_dupe_groups([topf1, topf2, subf1, subf2])), 1)  # only top folders
     # however, if another folder matches a subfolder, keep in in the matches
     otherf = no("other folder", size=41)
-    otherf.md5 = otherf.md5partial = otherf.md5samples = b"some_md5_2"
+    otherf.digest = otherf.digest_partial = otherf.digest_samples = b"some_digest__2"
     otherf.path = Path("/otherfolder")
     eq_(len(s.get_dupe_groups([topf1, topf2, subf1, subf2, otherf])), 2)
 
@@ -624,9 +634,9 @@ def test_dont_count_ref_files_as_discarded(fake_fileexists):
     o1 = no("foo", path="p1")
     o2 = no("foo", path="p2")
     o3 = no("foo", path="p3")
-    o1.md5 = o1.md5partial = o1.md5samples = "foobar"
-    o2.md5 = o2.md5partial = o2.md5samples = "foobar"
-    o3.md5 = o3.md5partial = o3.md5samples = "foobar"
+    o1.digest = o1.digest_partial = o1.digest_samples = "foobar"
+    o2.digest = o2.digest_partial = o2.digest_samples = "foobar"
+    o3.digest = o3.digest_partial = o3.digest_samples = "foobar"
     o1.is_ref = True
     o2.is_ref = True
     eq_(len(s.get_dupe_groups([o1, o2, o3])), 1)

core/util.py

@@ -7,6 +7,12 @@
 import time
 import sys
 import os
+import urllib.request
+import urllib.error
+import json
+import semantic_version
+import logging
+from typing import Union
 
 from hscommon.util import format_time_decimal
 
@@ -64,3 +70,34 @@ def fix_surrogate_encoding(s, encoding="utf-8"):
 
 def executable_folder():
     return os.path.dirname(os.path.abspath(sys.argv[0]))
+
+
+def check_for_update(current_version: str, include_prerelease: bool = False) -> Union[None, dict]:
+    request = urllib.request.Request(
+        "https://api.github.com/repos/arsenetar/dupeguru/releases",
+        headers={"Accept": "application/vnd.github.v3+json"},
+    )
+    try:
+        with urllib.request.urlopen(request) as response:
+            if response.status != 200:
+                logging.warn(f"Error retriving updates. Status: {response.status}")
+                return None
+            try:
+                response_json = json.loads(response.read())
+            except json.JSONDecodeError as ex:
+                logging.warn(f"Error parsing updates. {ex.msg}")
+                return None
+    except urllib.error.URLError as ex:
+        logging.warn(f"Error retriving updates. {ex.reason}")
+        return None
+    new_version = semantic_version.Version(current_version)
+    new_url = None
+    for release in response_json:
+        release_version = semantic_version.Version(release["name"])
+        if new_version < release_version and (include_prerelease or not release_version.prerelease):
+            new_version = release_version
+            new_url = release["html_url"]
+    if new_url is not None:
+        return {"version": new_version, "url": new_url}
+    else:
+        return None

help/changelog

@@ -1,3 +1,21 @@
+=== 4.3.1 (2022-07-08)
+* Fix issue where cache db exceptions could prevent files being hashed (#1015)
+* Add extra guard for non-zero length files without digests to prevent false duplicates
+* Update Italian translations
+
+=== 4.3.0 (2022-07-01)
+* Redirect stdout from custom command to the log files (#1008)
+* Update translations
+* Fix typo in debian control file (#989)
+* Add option to profile scans
+* Update fs.py to optimize stat() calls
+* Fix Error when delete after scan (#988)
+* Update directory scanning to use os.scandir() and DirEntry objects
+* Improve performance of Directories.get_state()
+* Migrate from hscommon.path to pathlib
+* Switch file hashing to xxhash with fallback to md5
+* Add update check feature to about box
+
 === 4.2.1 (2022-03-25)
 * Default to English on unsupported system language (#976)
 * Fix image viewer zoom datatype issue (#978)
@@ -39,7 +57,7 @@
 
 === 4.1.1 (2021-03-21)
 
-* Add Japanese 
+* Add Japanese
 * Update internationalization and translations to be up to date with current UI.
 * Minor translation and UI language updates
 * Fix language selection issues on Windows (#760)
@@ -401,7 +419,7 @@
 
 === 2.6.1 (2009-03-27)
 * **Fixed** an occasional crash caused by permission issues.
-* **Fixed** a bug where the "X discarded" notice would show a too large number of discarded 
+* **Fixed** a bug where the "X discarded" notice would show a too large number of discarded
       duplicates.
 
 === 2.6.0 (2008-09-10)
@@ -435,14 +453,14 @@
 * **Added** the "Remove empty folders" option.
 * **Fixed** results load/save issues.
 * **Fixed** occasional status bar inaccuracies when the results are filtered.
-           
+
 
 === 2.5.0 (2007-09-15)
 
 * **Added** post scan filtering.
 * **Fixed** issues with the rename feature under Windows
 * **Fixed** some user interface annoyances under Windows
-                
+
 
 === 2.4.8 (2007-04-14)
 
@@ -458,7 +476,7 @@
 
 * **Added** Re-orderable columns. In fact, I re-added the feature which was lost in the C# conversion in 2.4.0 (Windows).
 * **Changed** the behavior of the scanning engine when setting the hardness to 100. It will now only match files that have their words in the same order.
-* **Fixed** a bug with all the Delete/Move/Copy actions with certain kinds of files.            
+* **Fixed** a bug with all the Delete/Move/Copy actions with certain kinds of files.
 
 === 2.4.5 (2007-01-11)
 
@@ -496,7 +514,7 @@
 
 === 2.3.4 (2006-11-07)
 
-* **Improved** speed and memory usage of the scanning engine, again. Does it mean there was a lot of improvements to be made? Nah...            
+* **Improved** speed and memory usage of the scanning engine, again. Does it mean there was a lot of improvements to be made? Nah...
 
 === 2.3.3 (2006-11-02)
 
@@ -554,7 +572,7 @@
 === 2.2.3 (2006-06-15)
 
 * **Improved** duplicate scanning speed.
-* **Added** a warning that a file couldn't be renamed if a file with the same name already exists.            
+* **Added** a warning that a file couldn't be renamed if a file with the same name already exists.
 
 === 2.2.2 (2006-06-07)
 
@@ -598,9 +616,9 @@
 
 === 2.0.0 (2006-03-17)
 
-* Complete rewrite.         
+* Complete rewrite.
 * Now runs on Mac OS X.
 
 === 1.0.0 (2004-09-24)
 
-* Initial release.     
+* Initial release.

help/de/faq.rst

@@ -71,7 +71,7 @@ Häufig gestellte Fragen
         * Klicken Sie **Markieren --> Alle Markieren**.
 
 .. only:: edition_me
-    
+
     .. topic:: Ich möchte alle Stücke markieren, die mehr als 3 Sekunden von ihrer Referenz verschieden sind. Was kann ich tun?
 
         * Aktivieren Sie den :doc:`Nur Duplikate <results>` Modus.
@@ -83,7 +83,7 @@ Häufig gestellte Fragen
         * Klicken Sie auf **Entferne Ausgewählte von den Ergebnissen**.
 
     .. topic:: Ich möchte meine Stücke mit der höchsten Bitrate zur Referenz machen. Was kann ich tun?
-    
+
         * Aktivieren Sie den :doc:`Nur Duplikate <results>` Modus.
         * Aktivieren Sie den **Deltawerte** Modus.
         * Klicken Sie auf die "Bitrate" Spalte, um nach Bitrate zu sortieren.
@@ -92,9 +92,9 @@ Häufig gestellte Fragen
         * Klicken Sie auf **Mache Ausgewählte zur Referenz**.
 
     .. topic:: Ich möchte nicht das [live] und [remix] Versionen meiner Stücke als Duplikate erkannt werden. Was kann ich tun?
-    
+
         Ist Ihre Vergleichsschwelle niedrig genug, werden möglicherweise die live und remix Versionen in der Ergebnisliste landen. Das kann nicht verhindert werden, aber es gibt die Möglichkeit die Ergebnisse nach dem Scan zu entfernen, mittels dem Filter. Möchten Sie jedes Stück mit irgendetwas in eckigen Klammern [] im Dateinamen entfernen, so:
-    
+
         * **Windows**: Klicken Sie auf **Aktionen --> Filter anwenden**, geben "[*]" ein und klicken OK.
         * **Mac OS X**: Geben Sie "[*]" in das "Filter" Feld der Werkzeugleiste ein.
         * Klicken Sie auf **Markieren --> Alle Markieren**.

help/de/folders.rst

@@ -16,7 +16,7 @@ Jeder Ordner kann in einem von 3 Zuständen sein:
 * **Referenz:** Duplikate in diesem Ordner können **nicht** gelöscht werden. Dateien dieses Ordners können sich nur in der **Referenz** Position einer Duplikatgruppe befinden. Ist mehr als eine Datei des Referenzordners in derselben Duplikatgruppe, so wird nur Eine behalten. Die Anderen werden aus der Gruppe entfernt.
 * **Ausgeschlossen:** Dateien in diesem Verzeichnis sind nicht im Scan eingeschlossen.
 
-Der Standardzustand eines Ordners ist natürlich **Normal**. Sie können den **Referenz** Zustand für Ordner nutzen, in denen auf keinen Fall eine Datei gelöscht werden soll. 
+Der Standardzustand eines Ordners ist natürlich **Normal**. Sie können den **Referenz** Zustand für Ordner nutzen, in denen auf keinen Fall eine Datei gelöscht werden soll.
 
 Wenn sie einen Zustand für ein Verzeichnis setzen, erben alle Unterordner automatisch diesen Zustand, es sei denn Sie ändern den Zustand der Unterordner explizit.
 

help/de/index.rst

@@ -21,7 +21,7 @@ Inhalte:
 
 .. toctree::
     :maxdepth: 2
-    
+
     quick_start
     folders
     preferences

help/de/preferences.rst

@@ -4,9 +4,9 @@ Einstellungen
 .. only:: edition_se
 
     **Scan Typ:** Diese Option bestimmt nach welcher Eigenschaft die Dateien in einem Duplikate Scan verglichen werden. Wenn Sie **Dateiname** auswählen, wird dupeGuru jeden Dateinamen Wort für Wort vergleichen und, abhängig von den unteren Einstellungen, feststellen ob genügend Wörter übereinstimmen, um 2 Dateien als Duplikate zu betrachten. Wenn Sie **Inhalt** wählen, werden nur Dateien mit dem exakt gleichen Inhalt zusammenpassen.
-    
+
     Der **Ordner** Scan Typ ist etwas speziell. Wird er ausgewählt, scannt dupeGuru nach doppelten Ordnern anstelle von Dateien. Um festzustellen ob 2 Ordner identisch sind, werden alle Datein im Ordner gescannt und wenn die Inhalte aller Dateien der Ordner übereinstimmen, werden die Ordner als Duplikate erkannt.
-    
+
     **Filterempfindlichkeit:** Wenn Sie den **Dateiname** Scan Typ wählen, bestimmt diese Option wie ähnlich 2 Dateinamen für dupeGuru sein müssen, um Duplikate zu sein. Ist die Empfindlichkeit zum Beispiel 80, müssen 80% der Worte der 2 Dateinamen übereinstimmen. Um den Übereinstimmungsanteil herauszufinden, zählt dupeGuru zuerst die Gesamtzahl der Wörter **beider** Dateinamen, dann werden die gleichen Wörter gezählt (jedes Wort zählt als 2) und durch die Gesamtzahl der Wörter dividiert. Ist das Resultat größer oder gleich der Filterempfindlichkeit, haben wir ein Duplikat. Zum Beispiel, "a b c d" und "c d e" haben einen Übereinstimmungsanteil von 57 (4 gleiche Wörter, insgesamt 7 Wörter).
 
 .. only:: edition_me
@@ -33,7 +33,7 @@ Einstellungen
 .. only:: edition_pe
 
     **Scan Typ:** Diese option bestimmt, welcher Scan Typ bei Ihren Bildern angewendet wird. Der **Inhalte** Scan Typ vergleicht den Inhalt der Bilder auf eine ungenaue Art und Weise (so werden nicht nur exakte Duplikate gefunden, sondern auch Ähnliche). Der **EXIF Zeitstempel** Scan Typ schaut auf die EXIF Metadaten der Bilder (wenn vorhanden) und erkennt Bilder die den Selben haben. Er ist viel schneller als der Inhalte Scan. **Warnung:** Veränderte Bilder behalten oft den selben EXIF Zeitstempel, also achten Sie auf Falschpositive bei der Nutzung dieses Scans.
-    
+
     **Filterempfindlichkeit:** *Nur Inhalte Scan.* Je höher diese Einstellung, desto strenger ist der Filter (Mit anderen Worten, desto weniger Ergebnisse erhalten Sie). Die meisten Bilder der selben Qualität stimmen zu 100% überein, selbst wenn das Format anders ist (PNG und JPG zum Beispiel). Wie auch immer, wenn ein PNG mit einem JPG niederiger Qualität übereinstimmen soll, muss die Filterempfindlichkeit kleiner als 100 sein. Die Voreinstellung, 95, ist eine gute Wahl.
 
     **Bilder unterschiedlicher Abmessung gleich:** Wird diese Box gewählt, dürfen Bilder unterschiedlicher Abmessung in einer Duplikategruppe sein..
@@ -57,7 +57,7 @@ Auf jeden Fall behandelt dupeGuru Namenskonflikte indem es dem Ziel-Dateinamen e
 **Eigener Befehl:** Diese Einstellung bestimmt den Befehl der durch "Führe eigenen Befehl aus" ausgeführt wird. Sie können jede externe Anwendung durch diese Aktion aufrufen. Dies ist zum Beispiel hilfreich, wenn Sie eine gute diff-Anwendung installiert haben.
 
 Das Format des Befehls ist das Selbe wie in einer Befehlszeile, außer das 2 Platzhalter vorhanden sind: **%d** und **%r**. Diese Platzhalter werden durch den Pfad des markierten Duplikates (%d) und dem Pfad der Duplikatereferenz ersetzt (%r).
-  
+
 Wenn der Pfad Ihrer ausführbaren Datei Leerzeichen enthält, so schließen sie ihn bitte mit "" Zeichen ein. Sie sollten auch Platzhalter mit den Zitatzeichen einschließen, denn es ist möglich, das die Pfade der Duplikate und Referenzen ebenfalls Leerzeichen enthalten. Hier ist ein Beispiel eines eigenen Befehls::
-  
+
     "C:\Program Files\SuperDiffProg\SuperDiffProg.exe" "%d" "%r"

help/de/reprioritize.rst

@@ -22,4 +22,4 @@ criterion is used and so on and so on. For example, if your arguments are "Size
 "Filename (Doesn't end with a number)", the reference file that will be picked in a group will be
 the biggest file, and if two or more files have the same size, the one that has a filename that
 doesn't end with a number will be used. When all criteria result in ties, the order in which dupes
-previously were in the group will be used.
+previously were in the group will be used.

help/de/results.rst

@@ -98,4 +98,4 @@ Aktionen Menü
 * **Ausgewählte umbenennen:** Fragt nach einem neuen Namen und benennt die ausgewählte Datei um.
 
 .. todo:: Add Move and iPhoto/iTunes warning
-.. todo:: Add "Deletion Options" section.
+.. todo:: Add "Deletion Options" section.

help/en/contribute.rst

@@ -24,7 +24,7 @@ Development process
 * `Issue Tracker`_
 * `Issue labels meaning`_
 
-dupeGuru's source code is on Github and thus managed in a Git repository. At all times, you should
+dupeGuru's source code is on GitHub and thus managed in a Git repository. At all times, you should
 be able to build from source a fresh checkout of the ``master`` branch using instructions from the
 ``README.md`` file at the root of this project. If you can't, it's a bug. Please report it.
 
@@ -61,7 +61,7 @@ It's the same thing with feature requests. Description of a feature request, whe
 already been given to how such a feature would fit in the current design, are precious to developers
 and help them figure out a clear roadmap for the project.
 
-So, even if you're not a developer, you can always open a Github account and create/comment issues.
+So, even if you're not a developer, you can always open a GitHub account and create/comment issues.
 Your contribution will be much appreciated.
 
 **Documentation**. This is a bit trickier because dupeGuru's documentation is written with a rather

help/en/developer/core/engine.rst

@@ -2,12 +2,12 @@ core.engine
 ===========
 
 .. automodule:: core.engine
-    
+
     .. autoclass:: Match
-    
+
     .. autoclass:: Group
         :members:
-    
+
     .. autofunction:: build_word_dict
     .. autofunction:: compare
     .. autofunction:: compare_fields
@@ -16,7 +16,7 @@ core.engine
     .. autofunction:: get_groups
     .. autofunction:: merge_similar_words
     .. autofunction:: reduce_common_words
-    
+
 .. _fields:
 
 Fields

help/en/developer/core/gui/index.rst

@@ -6,5 +6,5 @@ core.gui
 
 .. toctree::
     :maxdepth: 2
-    
+
     deletion_options

help/en/developer/core/index.rst

@@ -3,7 +3,7 @@ core
 
 .. toctree::
     :maxdepth: 2
-    
+
     app
     fs
     engine

help/en/developer/hscommon/gui/base.rst

@@ -4,9 +4,9 @@ hscommon.gui.base
 .. automodule:: hscommon.gui.base
 
     .. autosummary::
-        
+
         GUIObject
-    
+
     .. autoclass:: GUIObject
         :members:
         :private-members:

help/en/developer/hscommon/gui/column.rst

@@ -4,22 +4,22 @@ hscommon.gui.column
 .. automodule:: hscommon.gui.column
 
     .. autosummary::
-        
+
         Columns
         Column
         ColumnsView
         PrefAccessInterface
-    
+
     .. autoclass:: Columns
         :members:
         :private-members:
-    
+
     .. autoclass:: Column
         :members:
         :private-members:
-    
+
     .. autoclass:: ColumnsView
         :members:
-    
+
     .. autoclass:: PrefAccessInterface
         :members:

help/en/developer/hscommon/gui/progress_window.rst

@@ -4,15 +4,14 @@ hscommon.gui.progress_window
 .. automodule:: hscommon.gui.progress_window
 
     .. autosummary::
-        
+
         ProgressWindow
         ProgressWindowView
-    
+
     .. autoclass:: ProgressWindow
         :members:
         :private-members:
-    
+
     .. autoclass:: ProgressWindowView
         :members:
         :private-members:
-    

help/en/developer/hscommon/gui/selectable_list.rst

@@ -4,23 +4,23 @@ hscommon.gui.selectable_list
 .. automodule:: hscommon.gui.selectable_list
 
     .. autosummary::
-        
+
         Selectable
         SelectableList
         GUISelectableList
         GUISelectableListView
-    
+
     .. autoclass:: Selectable
         :members:
         :private-members:
-    
+
     .. autoclass:: SelectableList
         :members:
         :private-members:
-    
+
     .. autoclass:: GUISelectableList
         :members:
         :private-members:
-    
+
     .. autoclass:: GUISelectableListView
         :members:

help/en/developer/hscommon/gui/table.rst

@@ -2,18 +2,18 @@ hscommon.gui.table
 ==================
 
 .. automodule:: hscommon.gui.table
-    
+
     .. autosummary::
-        
+
         Table
         Row
         GUITable
         GUITableView
-    
+
     .. autoclass:: Table
         :members:
         :private-members:
-    
+
     .. autoclass:: Row
         :members:
         :private-members:
@@ -21,6 +21,6 @@ hscommon.gui.table
     .. autoclass:: GUITable
         :members:
         :private-members:
-    
+
     .. autoclass:: GUITableView
         :members:

help/en/developer/hscommon/gui/text_field.rst

@@ -4,10 +4,10 @@ hscommon.gui.text_field
 .. automodule:: hscommon.gui.text_field
 
     .. autosummary::
-        
+
         TextField
         TextFieldView
-    
+
     .. autoclass:: TextField
         :members:
         :private-members:

help/en/developer/hscommon/gui/tree.rst

@@ -2,17 +2,16 @@ hscommon.gui.tree
 =================
 
 .. automodule:: hscommon.gui.tree
-    
+
     .. autosummary::
-        
+
         Tree
         Node
-    
+
     .. autoclass:: Tree
         :members:
         :private-members:
-    
+
     .. autoclass:: Node
         :members:
         :private-members:
-

help/en/developer/hscommon/index.rst

@@ -4,7 +4,7 @@ hscommon
 .. toctree::
     :maxdepth: 2
     :glob:
-    
+
     build
     conflict
     desktop
@@ -13,4 +13,3 @@ hscommon
     util
     jobprogress/*
     gui/*
-

help/en/developer/hscommon/jobprogress/job.rst

@@ -4,14 +4,13 @@ hscommon.jobprogress.job
 .. automodule:: hscommon.jobprogress.job
 
     .. autosummary::
-        
+
         Job
         NullJob
-    
+
     .. autoclass:: Job
         :members:
         :private-members:
 
     .. autoclass:: NullJob
         :members:
-

help/en/developer/hscommon/jobprogress/performer.rst

@@ -4,9 +4,8 @@ hscommon.jobprogress.performer
 .. automodule:: hscommon.jobprogress.performer
 
     .. autosummary::
-        
+
         ThreadedJobPerformer
-    
+
     .. autoclass:: ThreadedJobPerformer
         :members:
-

help/en/developer/index.rst

@@ -69,6 +69,6 @@ API
 
 .. toctree::
     :maxdepth: 2
-    
+
     core/index
     hscommon/index

help/en/faq.rst

@@ -30,8 +30,8 @@ that makes sure that you will **always** keep at least one member of the duplica
 How can I report a bug a suggest a feature?
 -------------------------------------------
 
-dupeGuru is hosted on `Github`_ and it's also where issues are tracked. The best way to report a
-bug or suggest a feature is to sign up on Github and `open an issue`_. 
+dupeGuru is hosted on `GitHub`_ and it's also where issues are tracked. The best way to report a
+bug or suggest a feature is to sign up on GitHub and `open an issue`_.
 
 The mark box of a file I want to delete is disabled. What must I do?
 --------------------------------------------------------------------
@@ -176,6 +176,5 @@ Preferences are stored elsewhere:
 * Linux: ``~/.config/Hardcoded Software/dupeGuru.conf``
 * Mac OS X: In the built-in ``defaults`` system, as ``com.hardcoded-software.dupeguru``
 
-.. _Github: https://github.com/arsenetar/dupeguru
+.. _GitHub: https://github.com/arsenetar/dupeguru
 .. _open an issue: https://github.com/arsenetar/dupeguru/wiki/issue-labels
-

help/en/index.rst

@@ -24,7 +24,7 @@ Contents:
 
 .. toctree::
     :maxdepth: 2
-    
+
     contribute
     quick_start
     folders

help/en/preferences.rst

@@ -14,6 +14,10 @@ Preferences
     If you check this box, pictures of different dimensions will be allowed in the same
     duplicate group.
 
+**Match pictures of different rotations:**
+    If you check this box, pictures of different rotations will be allowed in the same
+    duplicate group.
+
 .. _filter-hardness:
 
 **Filter Hardness:**
@@ -67,11 +71,11 @@ filename if the filename already exists in the destination.
 The format of the command is the same as what you would write in the command line, except that there
 are 2 placeholders: **%d** and **%r**. These placeholders will be replaced by the path of the
 selected dupe (%d) and the path of the selected dupe's reference file (%r).
-  
+
 If the path to your executable contains space characters, you should enclose it in "" quotes. You
 should also enclose placeholders in quotes because it's very possible that paths to dupes and refs
 will contain spaces. Here's an example custom command::
-  
+
     "C:\Program Files\SuperDiffProg\SuperDiffProg.exe" "%d" "%r"
 
 .. _inode: http://en.wikipedia.org/wiki/Inode

help/en/reprioritize.rst

@@ -22,4 +22,4 @@ criterion is used and so on and so on. For example, if your arguments are "Size
 "Filename (Doesn't end with a number)", the reference file that will be picked in a group will be
 the biggest file, and if two or more files have the same size, the one that has a filename that
 doesn't end with a number will be used. When all criteria result in ties, the order in which dupes
-previously were in the group will be used.
+previously were in the group will be used.

help/en/results.rst

@@ -180,7 +180,7 @@ any of them.
     the file's path. If the original file is deleted or moved, the link is broken. A hardlink is a
     link to the file *itself*. That link is as good as a "real" file. Only when *all* hardlinks to a
     file are deleted is the file itself deleted.
-  
+
     On OSX and Linux, this feature is supported fully, but under Windows, it's a bit complicated.
     Windows XP doesn't support it, but Vista and up support it. However, for the feature to work,
     dupeGuru has to run with administrative privileges.

help/fr/faq.rst

@@ -51,7 +51,7 @@ Tour groupe de doublons contient au moins un fichier dit "référence" et ce fic
 effacé. Par contre, ce que vous pouvez faire c'est de le remplacer par un autre fichier du groupe.
 Pour ce faire, sélectionnez un fichier du groupe et cliquez sur l'action **Transformer sélectionnés
 en références**.
-    
+
 Notez que si le fichier référence du groupe vient d'un dossier qui a été défini comme dossier
 référence, ce fichier ne peut pas être déplacé de sa position de référence du groupe.
 
@@ -71,7 +71,7 @@ doublons. Example: Nous avons 3 fichiers, A, B et C. Nous les comparons en utili
 de filtre. La comparaison détermine que A est un double de B, A est un double C, mais que B n'est
 **pas** un double de C. dupeGuru a ici un problème. Il ne peut pas créer un groupe avec A, B et C.
 Il décide donc de jeter C hors du groupe. C'est de là que vient la notice '(X hors-groupe)'.
-    
+
 Cette notice veut dire que si jamais vous effacez tout les doubles contenus dans vos résultats et
 que vous faites un nouveau scan, vous pourriez avoir de nouveaux résultats.
 

help/fr/folders.rst

@@ -3,7 +3,7 @@ Sélection de dossiers
 
 La première fenêtre qui apparaît lorsque dupeGuru démarre est la fenêtre de sélection de dossiers à scanner. Elle détermine la liste des dossiers qui seront scannés lorsque vous cliquerez sur **Scan**.
 
-Pour ajouter un dossier, cliquez sur le bouton **+**. Si vous avez ajouté des dossiers dans le passé, un menu vous permettra de rapidement choisir un de ceux ci. Autrement, il vous sera demandé d'indiquer le dossier à ajouter. 
+Pour ajouter un dossier, cliquez sur le bouton **+**. Si vous avez ajouté des dossiers dans le passé, un menu vous permettra de rapidement choisir un de ceux ci. Autrement, il vous sera demandé d'indiquer le dossier à ajouter.
 
 Vous pouvez aussi utiliser le drag & drop pour ajouter des dossiers à la liste.
 
@@ -26,14 +26,14 @@ Le type d'un dossier s'applique à ses sous-dossiers, excepté si un sous-dossie
 
     Bibliothèques iPhoto et Aperture
     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-    
+
     dupeGuru PE supporte iPhoto et Aperture, ce qui veut dire qu'il sait comment lire le contenu de
     ces bibliothèques et comment communiquer avec ces applications pour correctement supprimer des
     photos de celles-ci. Pour utiliser cette fonctionnalité, vous devez ajouter iPhoto et/ou
     Aperture avec les boutons spéciaux "Ajouter librairie iPhoto" et "Ajouter librairie Aperture",
     qui apparaissent quand on clique sur le petit "+". Les dossiers ajoutés seront alors
     correctement interprétés par dupeGuru.
-    
+
     Quand une photo est supprimée d'iPhoto, elle est envoyée dans la corbeille d'iPhoto.
 
     Quand une photo est supprimée d'Aperture, il n'est malheureusement pas possible de l'envoyer
@@ -45,13 +45,13 @@ Le type d'un dossier s'applique à ses sous-dossiers, excepté si un sous-dossie
 
     Bibliothèques iTunes
     ^^^^^^^^^^^^^^^^^^^^
-    
+
     dupeGuru ME supporte iTunes, ce qui veut dire qu'il sait comment lire le contenu de sa
     bibliothèque et comment communiquer avec iTunes pour correctement supprimer des chansons de sa
     bibliothèque. Pour utiliser cette fonctionnalité, vous devez ajouter iTunes avec le bouton
     spécial "Ajouter librairie iTunes", qui apparait quand on clique sur le petit "+". Le dossier
     ajouté sera alors correctement interprété par dupeGuru.
-    
+
     Quand une chanson est supprimée d'iTunes, elle est envoyée à la corebeille du système, comme un
     fichier normal. La différence ici, c'est qu'après la suppression, iTunes est correctement mis au
     fait de cette suppression et retire sa référence à cette chanson de sa bibliothèque.

help/fr/index.rst

@@ -21,7 +21,7 @@ Contents:
 
 .. toctree::
     :maxdepth: 2
-    
+
     quick_start
     folders
     preferences

help/fr/preferences.rst

@@ -4,9 +4,9 @@ Préférences
 .. only:: edition_se
 
     **Type de scan:** Cette option détermine quels aspects du fichier doit être comparé. Un scan par **Nom de fichier** compare les noms de fichiers mot-à-mot et, dépendant des autres préférences ci-dessous, déterminera si les noms se ressemblent assez pour être considérés comme doublons. Un scan par **Contenu** trouvera les doublons qui ont exactement le même contenu.
-    
+
     Le scan **Dossiers** est spécial. Si vous le sélectionnez, dupeGuru cherchera des doublons de *dossiers* plutôt que des doublons de fichiers. Pour déterminer si deux dossiers sont des doublons, dupeGuru regarde le contenu de tous les fichiers dans les dossiers, et si **tous** sont les mêmes, les dossiers sont considérés comme des doublons.
-    
+
     **Seuil du filtre:** Pour les scan de type **Nom de fichier**, cette option détermine le degré de similtude nécessaire afin de considérer deux noms comme doublons. Avec un seuil de 80, 80% des mots doivent être égaux. Pour déterminer ce pourcentage, dupeGuru compte le nombre de mots total des deux noms, puis compte le nombre de mots égaux, puis fait la division des deux. Un résultat égalisant ou dépassant le seuil sera considéré comme un doublon. Exemple: "a b c d" et "c d e" ont un pourcentage de 57 (4 mots égaux, 7 au total).
 
 .. only:: edition_me
@@ -33,7 +33,7 @@ Préférences
 .. only:: edition_pe
 
     **Type de scan:** Détermine le type de scan qui sera fait sur vos images. Le type **Contenu** compare le contenu des images de façon "fuzzy", rendant possible de trouver non seulement les doublons exactes, mais aussi les similaires. Le type **EXIF Timestamp** compare les métadonnées EXIF des images (si existantes) et détermine si le "timestamp" (moment de prise de la photo) est pareille. C'est beaucoup plus rapide que le scan par Contenu. **Attention:** Les photos modifiées gardent souvent le même timestamp, donc faites attention aux faux doublons si vous utilisez cette méthode.
-    
+
     **Seuil du filtre:** *Scan par Contenu seulement.* Plus il est élevé, plus les images doivent être similaires pour être considérées comme des doublons. Le défaut de 95% permet quelques petites différence, comme par exemple une différence de qualité ou bien une légère modification des couleurs.
 
     **Comparer les images de tailles différentes:** Le nom dit tout. Sans cette option, les images de tailles différentes ne sont pas comparées.
@@ -58,6 +58,6 @@ Dans tous les cas, dupeGuru résout les conflits de noms de fichier en ajoutant
 
 Le format de la ligne de commande est la même que celle que vous écrireriez manuellement, excepté pour les arguments, **%d** et **%r**. L'endroit où vous placez ces deux arguments sera remplacé par le chemin du fichier sélectionné (%d) et le chemin de son fichier référence dans le groupe (%r).
 
-Si le chemin de votre executable contient un espace, vous devez le placer entre guillemets "". Vous devriez aussi placer vos arguments %d et %r entre guillemets parce qu'il est très possible d'avoir des chemins de fichier contenant des espaces. Voici un exemple de commande personnelle::  
-  
+Si le chemin de votre executable contient un espace, vous devez le placer entre guillemets "". Vous devriez aussi placer vos arguments %d et %r entre guillemets parce qu'il est très possible d'avoir des chemins de fichier contenant des espaces. Voici un exemple de commande personnelle::
+
     "C:\Program Files\SuperDiffProg\SuperDiffProg.exe" "%d" "%r"

help/fr/reprioritize.rst

@@ -22,4 +22,4 @@ criterion is used and so on and so on. For example, if your arguments are "Size
 "Filename (Doesn't end with a number)", the reference file that will be picked in a group will be
 the biggest file, and if two or more files have the same size, the one that has a filename that
 doesn't end with a number will be used. When all criteria result in ties, the order in which dupes
-previously were in the group will be used.
+previously were in the group will be used.

help/fr/results.rst

@@ -131,7 +131,7 @@ s'exécute. La plupart du temps, ces options n'ont pas a être activées.
   liens (`symlink`_ ou `hardlink`_) vers leur fichiers de référence respectifs. Un symlink est un
   lien symbolique (qui devient caduque si l'original est supprimé) et un hardlink est un lien direct
   au contenu du fichier (même si l'original est supprimé, le lien reste valide).
-  
+
   Sur OS X et Linux, cette fonction est supportée pleinement, mais sur Windows, c'est un peu
   compliqué. Windows XP ne le supporte pas, mais Vista oui. De plus, cette fonction ne peut être
   utilisée que si dupeGuru roule avec les privilèges administratifs. Ouaip, Windows c'est la joie.