├── .codespellignore ├── .dockerignore ├── .github ├── FUNDING.yml ├── ISSUE_TEMPLATE │ ├── bug_report.md │ └── feature_request.md ├── config.yml ├── dependabot.yml ├── reaction.yml ├── release-drafter.yml ├── stale.yml └── workflows │ ├── codeql-analysis.yml │ ├── codespell.yml │ ├── dockerimage-latest.yml │ ├── dockerimage-release.yml │ └── dockerimage-test.yml ├── .gitignore ├── Dockerfile ├── LICENSE ├── MANIFEST.in ├── Makefile ├── README.md ├── docker-compose.yml ├── docker └── entrypoint.sh ├── example.py ├── poetry.lock ├── py_image_dedup ├── __init__.py ├── cli.py ├── config.py ├── library │ ├── __init__.py │ ├── deduplication_result.py │ ├── deduplicator.py │ ├── file_watch.py │ ├── processing_manager.py │ └── progress_manager.py ├── persistence │ ├── __init__.py │ ├── elasticsearchstorebackend.py │ └── metadata_key.py ├── stats.py └── util │ ├── __init__.py │ ├── file.py │ └── image.py ├── py_image_dedup_reference.yaml ├── pyproject.toml ├── pytest.ini └── tests ├── __init__.py ├── images ├── bottles │ ├── IMG_20190903_193151-edited.jpg │ ├── IMG_20190903_193151-grayscale.jpg │ ├── IMG_20190903_193151-telegram-compression.jpg │ └── IMG_20190903_193151.jpg ├── building │ ├── IMG_20190903_193508-edited.jpg │ ├── IMG_20190903_193508-grayscale.jpg │ ├── IMG_20190903_193508-telegram-compression.jpg │ └── IMG_20190903_193508.jpg └── clouds │ ├── IMG_20190903_193537-edited.jpg │ ├── IMG_20190903_193537-grayscale.jpg │ ├── IMG_20190903_193537-telegram-compression.jpg │ └── IMG_20190903_193537.jpg ├── py_image_dedup.yaml ├── test_file_extension.py └── test_select_images_to_delete.py /.codespellignore: -------------------------------------------------------------------------------- 1 | Archiv -------------------------------------------------------------------------------- /.dockerignore: -------------------------------------------------------------------------------- 1 | .github 2 | .pytest_cache 3 | duplicates 4 | mnt 5 | venv 6 | *.yaml -------------------------------------------------------------------------------- /.github/FUNDING.yml: -------------------------------------------------------------------------------- 1 | # These are supported funding model platforms 2 | 3 | github: [markusressel] 4 | # patreon: # Replace with a single Patreon username 5 | # open_collective: # Replace with a single Open Collective username 6 | # ko_fi: # Replace with a single Ko-fi username 7 | # tidelift: # Replace with a single Tidelift platform-name/package-name e.g., npm/babel 8 | # community_bridge: # Replace with a single Community Bridge project-name e.g., cloud-foundry 9 | # liberapay: TheAlgorithms 10 | # issuehunt: # Replace with a single IssueHunt username 11 | # otechie: # Replace with a single Otechie username 12 | # custom: ['https://paypal.me/markusressel/1'] -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/bug_report.md: -------------------------------------------------------------------------------- 1 | --- 2 | name: Bug report 3 | about: Create a report to help us improve 4 | 5 | --- 6 | 7 | **Describe the bug** 8 | A clear and concise description of what the bug is. 9 | 10 | **To Reproduce** 11 | Steps to reproduce the behavior: 12 | 1. Go to '...' 13 | 2. Click on '....' 14 | 3. Scroll down to '....' 15 | 4. See error 16 | 17 | **Expected behavior** 18 | A clear and concise description of what you expected to happen. 19 | 20 | **Screenshots** 21 | If applicable, add screenshots to help explain your problem. 22 | 23 | **Desktop (please complete the following information):** 24 | - OS: [e.g. Linux] 25 | 26 | **Additional context** 27 | Add any other context about the problem here. 28 | -------------------------------------------------------------------------------- /.github/ISSUE_TEMPLATE/feature_request.md: -------------------------------------------------------------------------------- 1 | --- 2 | name: Feature request 3 | about: Suggest an idea for this project 4 | 5 | --- 6 | 7 | **Is your feature request related to a problem? Please describe.** 8 | A clear and concise description of what the problem is. Ex. I'm always frustrated when [...] 9 | 10 | **Describe the solution you'd like** 11 | A clear and concise description of what you want to happen. 12 | 13 | **Describe alternatives you've considered** 14 | A clear and concise description of any alternative solutions or features you've considered. 15 | 16 | **Additional context** 17 | Add any other context or screenshots about the feature request here. 18 | -------------------------------------------------------------------------------- /.github/config.yml: -------------------------------------------------------------------------------- 1 | # Configuration for new-issue-welcome - https://github.com/behaviorbot/new-issue-welcome 2 | 3 | # Comment to be posted to on first time issues 4 | newIssueWelcomeComment: > 5 | Thanks for opening your first issue here! :tada: 6 | 7 | # Configuration for new-pr-welcome - https://github.com/behaviorbot/new-pr-welcome 8 | 9 | # Comment to be posted to on PRs from first time contributors in your repository 10 | newPRWelcomeComment: > 11 | Thanks for opening this pull request! :nerd_face: 12 | 13 | # Configuration for first-pr-merge - https://github.com/behaviorbot/first-pr-merge 14 | 15 | # Comment to be posted to on pull requests merged by a first time user 16 | firstPRMergeComment: > 17 | Congrats on merging your first pull request here! You should be proud of yourself :1st_place_medal: 18 | ![Congratulations](https://media.giphy.com/media/ehhuGD0nByYxO/giphy.gif) 19 | 20 | # It is recommend to include as many gifs and emojis as possible 21 | -------------------------------------------------------------------------------- /.github/dependabot.yml: -------------------------------------------------------------------------------- 1 | version: 2 2 | registries: 3 | python-index-pypi-python-org-simple: 4 | type: python-index 5 | url: https://pypi.python.org/simple/ 6 | username: "${{secrets.PYTHON_INDEX_PYPI_PYTHON_ORG_SIMPLE_USERNAME}}" 7 | password: "${{secrets.PYTHON_INDEX_PYPI_PYTHON_ORG_SIMPLE_PASSWORD}}" 8 | 9 | updates: 10 | - package-ecosystem: github-actions 11 | directory: "/" 12 | schedule: 13 | # Check for updates to GitHub Actions every week 14 | interval: "weekly" 15 | - package-ecosystem: pip 16 | insecure-external-code-execution: allow 17 | directory: "/" 18 | schedule: 19 | interval: daily 20 | time: "16:00" 21 | timezone: Europe/Berlin 22 | open-pull-requests-limit: 10 23 | registries: 24 | - python-index-pypi-python-org-simple 25 | -------------------------------------------------------------------------------- /.github/reaction.yml: -------------------------------------------------------------------------------- 1 | 2 | -------------------------------------------------------------------------------- /.github/release-drafter.yml: -------------------------------------------------------------------------------- 1 | categories: 2 | - title: 🚀 Features and ✨ Enhancements 3 | label: enhancement 4 | - title: 🐛 Bugfixes 5 | label: bug 6 | change-template: "* $TITLE (#$NUMBER) by @$AUTHOR" 7 | template: | 8 | ## What’s Changed 9 | 10 | $CHANGES 11 | -------------------------------------------------------------------------------- /.github/stale.yml: -------------------------------------------------------------------------------- 1 | # Configuration for probot-stale - https://github.com/probot/stale 2 | 3 | # Number of days of inactivity before an Issue or Pull Request becomes stale 4 | daysUntilStale: 60 5 | 6 | # Number of days of inactivity before an Issue or Pull Request with the stale label is closed. 7 | # Set to false to disable. If disabled, issues still need to be closed manually, but will remain marked as stale. 8 | daysUntilClose: 14 9 | 10 | # Issues or Pull Requests with these labels will never be considered stale. Set to `[]` to disable 11 | exemptLabels: 12 | - pinned 13 | - security 14 | - bug 15 | - enhancement 16 | 17 | # Set to true to ignore issues in a project (defaults to false) 18 | exemptProjects: false 19 | 20 | # Set to true to ignore issues in a milestone (defaults to false) 21 | exemptMilestones: false 22 | 23 | # Set to true to ignore issues with an assignee (defaults to false) 24 | exemptAssignees: true 25 | 26 | # Label to use when marking as stale 27 | staleLabel: wontfix 28 | 29 | # Comment to post when marking as stale. Set to `false` to disable 30 | markComment: > 31 | This issue has been automatically marked as stale because it has not had 32 | recent activity. It will be closed if no further activity occurs. Thank you 33 | for your contributions. 34 | 35 | # Comment to post when removing the stale label. 36 | # unmarkComment: > 37 | # Your comment here. 38 | 39 | # Comment to post when closing a stale Issue or Pull Request. 40 | closeComment: > 41 | There has been no incentive by contributors or maintainers to revive this stale issue and it will now be closed. 42 | 43 | # Limit the number of actions per hour, from 1-30. Default is 30 44 | limitPerRun: 30 45 | 46 | # Limit to only `issues` or `pulls` 47 | only: issues 48 | 49 | # Optionally, specify configuration settings that are specific to just 'issues' or 'pulls': 50 | # pulls: 51 | # daysUntilStale: 30 52 | # markComment: > 53 | # This pull request has been automatically marked as stale because it has not had 54 | # recent activity. It will be closed if no further activity occurs. Thank you 55 | # for your contributions. 56 | 57 | # issues: 58 | # exemptLabels: 59 | # - confirmed 60 | -------------------------------------------------------------------------------- /.github/workflows/codeql-analysis.yml: -------------------------------------------------------------------------------- 1 | name: "CodeQL" 2 | 3 | on: 4 | push: 5 | branches: [master, ] 6 | pull_request: 7 | # The branches below must be a subset of the branches above 8 | branches: [master] 9 | schedule: 10 | - cron: '0 1 * * 4' 11 | 12 | jobs: 13 | analyze: 14 | name: Analyze 15 | runs-on: ubuntu-latest 16 | 17 | steps: 18 | - name: Checkout repository 19 | uses: actions/checkout@v4 20 | with: 21 | # We must fetch at least the immediate parents so that if this is 22 | # a pull request then we can checkout the head. 23 | fetch-depth: 2 24 | 25 | # If this run was triggered by a pull request event, then checkout 26 | # the head of the pull request instead of the merge commit. 27 | - run: git checkout HEAD^2 28 | if: ${{ github.event_name == 'pull_request' }} 29 | 30 | # Initializes the CodeQL tools for scanning. 31 | - name: Initialize CodeQL 32 | uses: github/codeql-action/init@v3 33 | # Override language selection by uncommenting this and choosing your languages 34 | # with: 35 | # languages: go, javascript, csharp, python, cpp, java 36 | 37 | # Autobuild attempts to build any compiled languages (C/C++, C#, or Java). 38 | # If this step fails, then you should remove it and run the build manually (see below) 39 | - name: Autobuild 40 | uses: github/codeql-action/autobuild@v3 41 | 42 | # ℹ️ Command-line programs to run using the OS shell. 43 | # 📚 https://git.io/JvXDl 44 | 45 | # ✏️ If the Autobuild fails above, remove it and uncomment the following three lines 46 | # and modify them (or add more) to build your code if your project 47 | # uses a compiled language 48 | 49 | #- run: | 50 | # make bootstrap 51 | # make release 52 | 53 | - name: Perform CodeQL Analysis 54 | uses: github/codeql-action/analyze@v3 55 | -------------------------------------------------------------------------------- /.github/workflows/codespell.yml: -------------------------------------------------------------------------------- 1 | # Codespell configuration is within pyproject.toml 2 | --- 3 | name: Codespell 4 | 5 | on: 6 | push: 7 | branches: [ master ] 8 | pull_request: 9 | branches: [ master ] 10 | 11 | permissions: 12 | contents: read 13 | 14 | jobs: 15 | codespell: 16 | name: Check for spelling errors 17 | runs-on: ubuntu-latest 18 | 19 | steps: 20 | - name: Checkout 21 | uses: actions/checkout@v4 22 | - name: Annotate locations with typos 23 | uses: codespell-project/codespell-problem-matcher@v1 24 | - name: Codespell 25 | uses: codespell-project/actions-codespell@v2 26 | with: 27 | ignore_words_file: .codespellignore 28 | skip: "*.svg" -------------------------------------------------------------------------------- /.github/workflows/dockerimage-latest.yml: -------------------------------------------------------------------------------- 1 | name: Docker Image latest 2 | 3 | on: 4 | push: 5 | branches: [ master ] 6 | 7 | jobs: 8 | build: 9 | runs-on: ubuntu-latest 10 | steps: 11 | - uses: actions/checkout@v4 12 | - name: Build the Docker image 13 | run: docker build . --file Dockerfile --tag markusressel/py-image-dedup:latest 14 | - name: Login to DockerHub Registry 15 | run: echo ${{ secrets.DOCKERHUB_PASSWORD }} | docker login -u ${{ secrets.DOCKERHUB_USERNAME }} --password-stdin 16 | - name: Push the Docker image 17 | if: github.ref_name == 'master' 18 | run: docker push markusressel/py-image-dedup:latest 19 | 20 | dockerHubDescription: 21 | runs-on: ubuntu-latest 22 | steps: 23 | - uses: actions/checkout@v4 24 | - name: Docker Hub Description 25 | uses: peter-evans/dockerhub-description@v4.0.2 26 | env: 27 | DOCKERHUB_USERNAME: ${{ secrets.DOCKERHUB_USERNAME }} 28 | DOCKERHUB_PASSWORD: ${{ secrets.DOCKERHUB_PASSWORD }} 29 | DOCKERHUB_REPOSITORY: markusressel/py-image-dedup 30 | -------------------------------------------------------------------------------- /.github/workflows/dockerimage-release.yml: -------------------------------------------------------------------------------- 1 | name: Docker Image Release 2 | 3 | on: 4 | push: 5 | tags: 6 | - "*.*.*" 7 | 8 | jobs: 9 | build: 10 | runs-on: ubuntu-latest 11 | steps: 12 | - uses: actions/checkout@v4 13 | # this writes the tag name into GIT_TAG_NAME 14 | - name: Get tag name 15 | uses: little-core-labs/get-git-tag@v3.0.2 16 | - name: Build the Docker image 17 | run: docker build . --file Dockerfile --tag markusressel/py-image-dedup:$GIT_TAG_NAME 18 | - name: Login to DockerHub Registry 19 | run: echo ${{ secrets.DOCKERHUB_PASSWORD }} | docker login -u ${{ secrets.DOCKERHUB_USERNAME }} --password-stdin 20 | - name: Push the Docker image 21 | run: docker push markusressel/py-image-dedup:$GIT_TAG_NAME 22 | -------------------------------------------------------------------------------- /.github/workflows/dockerimage-test.yml: -------------------------------------------------------------------------------- 1 | name: Test Docker Image 2 | 3 | on: 4 | pull_request: 5 | # The branches below must be a subset of the branches above 6 | branches: [ master ] 7 | 8 | jobs: 9 | test: 10 | runs-on: ubuntu-latest 11 | steps: 12 | - uses: actions/checkout@v4 13 | - name: Build the Docker image 14 | run: docker build . --file Dockerfile --tag markusressel/py-image-dedup:latest 15 | -------------------------------------------------------------------------------- /.gitignore: -------------------------------------------------------------------------------- 1 | # Created by .ignore support plugin (hsz.mobi) 2 | ### VirtualEnv template 3 | # Virtualenv 4 | # http://iamzed.com/2009/05/07/a-primer-on-virtualenv/ 5 | .Python 6 | [Bb]in 7 | [Ii]nclude 8 | [Ll]ib 9 | [Ll]ib64 10 | [Ll]ocal 11 | [Ss]cripts 12 | pyvenv.cfg 13 | .venv 14 | pip-selfcheck.json 15 | ### JetBrains template 16 | # Covers JetBrains IDEs: IntelliJ, RubyMine, PhpStorm, AppCode, PyCharm, CLion, Android Studio and Webstorm 17 | # Reference: https://intellij-support.jetbrains.com/hc/en-us/articles/206544839 18 | 19 | # User-specific stuff: 20 | .idea/**/workspace.xml 21 | .idea/**/tasks.xml 22 | .idea/dictionaries 23 | 24 | # Sensitive or high-churn files: 25 | .idea/**/dataSources/ 26 | .idea/**/dataSources.ids 27 | .idea/**/dataSources.xml 28 | .idea/**/dataSources.local.xml 29 | .idea/**/sqlDataSources.xml 30 | .idea/**/dynamic.xml 31 | .idea/**/uiDesigner.xml 32 | 33 | # Gradle: 34 | .idea/**/gradle.xml 35 | .idea/**/libraries 36 | 37 | # CMake 38 | cmake-build-debug/ 39 | cmake-build-release/ 40 | 41 | # Mongo Explorer plugin: 42 | .idea/**/mongoSettings.xml 43 | 44 | ## File-based project format: 45 | *.iws 46 | 47 | ## Plugin-specific files: 48 | 49 | # IntelliJ 50 | out/ 51 | 52 | # mpeltonen/sbt-idea plugin 53 | .idea_modules/ 54 | 55 | # JIRA plugin 56 | atlassian-ide-plugin.xml 57 | 58 | # Cursive Clojure plugin 59 | .idea/replstate.xml 60 | 61 | # Crashlytics plugin (for Android Studio and IntelliJ) 62 | com_crashlytics_export_strings.xml 63 | crashlytics.properties 64 | crashlytics-build.properties 65 | fabric.properties 66 | ### Python template 67 | # Byte-compiled / optimized / DLL files 68 | __pycache__/ 69 | *.py[cod] 70 | *$py.class 71 | 72 | # C extensions 73 | *.so 74 | 75 | # Distribution / packaging 76 | .Python 77 | build/ 78 | develop-eggs/ 79 | dist/ 80 | downloads/ 81 | eggs/ 82 | .eggs/ 83 | lib/ 84 | lib64/ 85 | parts/ 86 | sdist/ 87 | var/ 88 | wheels/ 89 | *.egg-info/ 90 | .installed.cfg 91 | *.egg 92 | MANIFEST 93 | 94 | # PyInstaller 95 | # Usually these files are written by a python script from a template 96 | # before PyInstaller builds the exe, so as to inject date/other infos into it. 97 | *.manifest 98 | *.spec 99 | 100 | # Installer logs 101 | pip-log.txt 102 | pip-delete-this-directory.txt 103 | 104 | # Unit test / coverage reports 105 | htmlcov/ 106 | .tox/ 107 | .coverage 108 | .coverage.* 109 | .cache 110 | nosetests.xml 111 | coverage.xml 112 | *.cover 113 | .hypothesis/ 114 | 115 | # Translations 116 | *.mo 117 | *.pot 118 | 119 | # Django stuff: 120 | *.log 121 | .static_storage/ 122 | .media/ 123 | local_settings.py 124 | 125 | # Flask stuff: 126 | instance/ 127 | .webassets-cache 128 | 129 | # Scrapy stuff: 130 | .scrapy 131 | 132 | # Sphinx documentation 133 | docs/_build/ 134 | 135 | # PyBuilder 136 | target/ 137 | 138 | # Jupyter Notebook 139 | .ipynb_checkpoints 140 | 141 | # pyenv 142 | .python-version 143 | 144 | # celery beat schedule file 145 | celerybeat-schedule 146 | 147 | # SageMath parsed files 148 | *.sage.py 149 | 150 | # Environments 151 | .env 152 | .venv 153 | env/ 154 | venv/ 155 | ENV/ 156 | env.bak/ 157 | venv.bak/ 158 | 159 | # Spyder project settings 160 | .spyderproject 161 | .spyproject 162 | 163 | # Rope project settings 164 | .ropeproject 165 | 166 | # mkdocs documentation 167 | /site 168 | 169 | # mypy 170 | .mypy_cache/ 171 | 172 | -------------------------------------------------------------------------------- /Dockerfile: -------------------------------------------------------------------------------- 1 | # dont use alpine for python builds: https://pythonspeed.com/articles/alpine-docker-python/ 2 | FROM python:3.11-slim-buster 3 | 4 | ENV PYTHONUNBUFFERED=1 5 | ENV POETRY_VERSION="2.1.2" 6 | ENV PIP_DISABLE_PIP_VERSION_CHECK=on 7 | 8 | RUN apt-get update \ 9 | && apt-get -y install sudo git python-skimage 10 | 11 | WORKDIR /app 12 | 13 | COPY . . 14 | 15 | COPY poetry.lock pyproject.toml ./ 16 | 17 | RUN apt-get update && \ 18 | apt-get install -y libatlas-base-dev gfortran && \ 19 | apt-get clean && rm -rf /var/lib/apt/lists/* 20 | 21 | RUN pip install "poetry==$POETRY_VERSION" \ 22 | && POETRY_VIRTUALENVS_CREATE=false poetry install --without dev \ 23 | && pip uninstall -y poetry 24 | 25 | ENV PUID=1000 PGID=1000 26 | 27 | ENTRYPOINT [ "docker/entrypoint.sh", "py-image-dedup" ] 28 | CMD [ "daemon" ] 29 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | GNU AFFERO GENERAL PUBLIC LICENSE 2 | Version 3, 19 November 2007 3 | 4 | Copyright (C) 2007 Free Software Foundation, Inc. 5 | Everyone is permitted to copy and distribute verbatim copies 6 | of this license document, but changing it is not allowed. 7 | 8 | Preamble 9 | 10 | The GNU Affero General Public License is a free, copyleft license for 11 | software and other kinds of works, specifically designed to ensure 12 | cooperation with the community in the case of network server software. 13 | 14 | The licenses for most software and other practical works are designed 15 | to take away your freedom to share and change the works. By contrast, 16 | our General Public Licenses are intended to guarantee your freedom to 17 | share and change all versions of a program--to make sure it remains free 18 | software for all its users. 19 | 20 | When we speak of free software, we are referring to freedom, not 21 | price. Our General Public Licenses are designed to make sure that you 22 | have the freedom to distribute copies of free software (and charge for 23 | them if you wish), that you receive source code or can get it if you 24 | want it, that you can change the software or use pieces of it in new 25 | free programs, and that you know you can do these things. 26 | 27 | Developers that use our General Public Licenses protect your rights 28 | with two steps: (1) assert copyright on the software, and (2) offer 29 | you this License which gives you legal permission to copy, distribute 30 | and/or modify the software. 31 | 32 | A secondary benefit of defending all users' freedom is that 33 | improvements made in alternate versions of the program, if they 34 | receive widespread use, become available for other developers to 35 | incorporate. Many developers of free software are heartened and 36 | encouraged by the resulting cooperation. However, in the case of 37 | software used on network servers, this result may fail to come about. 38 | The GNU General Public License permits making a modified version and 39 | letting the public access it on a server without ever releasing its 40 | source code to the public. 41 | 42 | The GNU Affero General Public License is designed specifically to 43 | ensure that, in such cases, the modified source code becomes available 44 | to the community. It requires the operator of a network server to 45 | provide the source code of the modified version running there to the 46 | users of that server. Therefore, public use of a modified version, on 47 | a publicly accessible server, gives the public access to the source 48 | code of the modified version. 49 | 50 | An older license, called the Affero General Public License and 51 | published by Affero, was designed to accomplish similar goals. This is 52 | a different license, not a version of the Affero GPL, but Affero has 53 | released a new version of the Affero GPL which permits relicensing under 54 | this license. 55 | 56 | The precise terms and conditions for copying, distribution and 57 | modification follow. 58 | 59 | TERMS AND CONDITIONS 60 | 61 | 0. Definitions. 62 | 63 | "This License" refers to version 3 of the GNU Affero General Public License. 64 | 65 | "Copyright" also means copyright-like laws that apply to other kinds of 66 | works, such as semiconductor masks. 67 | 68 | "The Program" refers to any copyrightable work licensed under this 69 | License. Each licensee is addressed as "you". "Licensees" and 70 | "recipients" may be individuals or organizations. 71 | 72 | To "modify" a work means to copy from or adapt all or part of the work 73 | in a fashion requiring copyright permission, other than the making of an 74 | exact copy. The resulting work is called a "modified version" of the 75 | earlier work or a work "based on" the earlier work. 76 | 77 | A "covered work" means either the unmodified Program or a work based 78 | on the Program. 79 | 80 | To "propagate" a work means to do anything with it that, without 81 | permission, would make you directly or secondarily liable for 82 | infringement under applicable copyright law, except executing it on a 83 | computer or modifying a private copy. Propagation includes copying, 84 | distribution (with or without modification), making available to the 85 | public, and in some countries other activities as well. 86 | 87 | To "convey" a work means any kind of propagation that enables other 88 | parties to make or receive copies. Mere interaction with a user through 89 | a computer network, with no transfer of a copy, is not conveying. 90 | 91 | An interactive user interface displays "Appropriate Legal Notices" 92 | to the extent that it includes a convenient and prominently visible 93 | feature that (1) displays an appropriate copyright notice, and (2) 94 | tells the user that there is no warranty for the work (except to the 95 | extent that warranties are provided), that licensees may convey the 96 | work under this License, and how to view a copy of this License. If 97 | the interface presents a list of user commands or options, such as a 98 | menu, a prominent item in the list meets this criterion. 99 | 100 | 1. Source Code. 101 | 102 | The "source code" for a work means the preferred form of the work 103 | for making modifications to it. "Object code" means any non-source 104 | form of a work. 105 | 106 | A "Standard Interface" means an interface that either is an official 107 | standard defined by a recognized standards body, or, in the case of 108 | interfaces specified for a particular programming language, one that 109 | is widely used among developers working in that language. 110 | 111 | The "System Libraries" of an executable work include anything, other 112 | than the work as a whole, that (a) is included in the normal form of 113 | packaging a Major Component, but which is not part of that Major 114 | Component, and (b) serves only to enable use of the work with that 115 | Major Component, or to implement a Standard Interface for which an 116 | implementation is available to the public in source code form. A 117 | "Major Component", in this context, means a major essential component 118 | (kernel, window system, and so on) of the specific operating system 119 | (if any) on which the executable work runs, or a compiler used to 120 | produce the work, or an object code interpreter used to run it. 121 | 122 | The "Corresponding Source" for a work in object code form means all 123 | the source code needed to generate, install, and (for an executable 124 | work) run the object code and to modify the work, including scripts to 125 | control those activities. However, it does not include the work's 126 | System Libraries, or general-purpose tools or generally available free 127 | programs which are used unmodified in performing those activities but 128 | which are not part of the work. For example, Corresponding Source 129 | includes interface definition files associated with source files for 130 | the work, and the source code for shared libraries and dynamically 131 | linked subprograms that the work is specifically designed to require, 132 | such as by intimate data communication or control flow between those 133 | subprograms and other parts of the work. 134 | 135 | The Corresponding Source need not include anything that users 136 | can regenerate automatically from other parts of the Corresponding 137 | Source. 138 | 139 | The Corresponding Source for a work in source code form is that 140 | same work. 141 | 142 | 2. Basic Permissions. 143 | 144 | All rights granted under this License are granted for the term of 145 | copyright on the Program, and are irrevocable provided the stated 146 | conditions are met. This License explicitly affirms your unlimited 147 | permission to run the unmodified Program. The output from running a 148 | covered work is covered by this License only if the output, given its 149 | content, constitutes a covered work. This License acknowledges your 150 | rights of fair use or other equivalent, as provided by copyright law. 151 | 152 | You may make, run and propagate covered works that you do not 153 | convey, without conditions so long as your license otherwise remains 154 | in force. You may convey covered works to others for the sole purpose 155 | of having them make modifications exclusively for you, or provide you 156 | with facilities for running those works, provided that you comply with 157 | the terms of this License in conveying all material for which you do 158 | not control copyright. Those thus making or running the covered works 159 | for you must do so exclusively on your behalf, under your direction 160 | and control, on terms that prohibit them from making any copies of 161 | your copyrighted material outside their relationship with you. 162 | 163 | Conveying under any other circumstances is permitted solely under 164 | the conditions stated below. Sublicensing is not allowed; section 10 165 | makes it unnecessary. 166 | 167 | 3. Protecting Users' Legal Rights From Anti-Circumvention Law. 168 | 169 | No covered work shall be deemed part of an effective technological 170 | measure under any applicable law fulfilling obligations under article 171 | 11 of the WIPO copyright treaty adopted on 20 December 1996, or 172 | similar laws prohibiting or restricting circumvention of such 173 | measures. 174 | 175 | When you convey a covered work, you waive any legal power to forbid 176 | circumvention of technological measures to the extent such circumvention 177 | is effected by exercising rights under this License with respect to 178 | the covered work, and you disclaim any intention to limit operation or 179 | modification of the work as a means of enforcing, against the work's 180 | users, your or third parties' legal rights to forbid circumvention of 181 | technological measures. 182 | 183 | 4. Conveying Verbatim Copies. 184 | 185 | You may convey verbatim copies of the Program's source code as you 186 | receive it, in any medium, provided that you conspicuously and 187 | appropriately publish on each copy an appropriate copyright notice; 188 | keep intact all notices stating that this License and any 189 | non-permissive terms added in accord with section 7 apply to the code; 190 | keep intact all notices of the absence of any warranty; and give all 191 | recipients a copy of this License along with the Program. 192 | 193 | You may charge any price or no price for each copy that you convey, 194 | and you may offer support or warranty protection for a fee. 195 | 196 | 5. Conveying Modified Source Versions. 197 | 198 | You may convey a work based on the Program, or the modifications to 199 | produce it from the Program, in the form of source code under the 200 | terms of section 4, provided that you also meet all of these conditions: 201 | 202 | a) The work must carry prominent notices stating that you modified 203 | it, and giving a relevant date. 204 | 205 | b) The work must carry prominent notices stating that it is 206 | released under this License and any conditions added under section 207 | 7. This requirement modifies the requirement in section 4 to 208 | "keep intact all notices". 209 | 210 | c) You must license the entire work, as a whole, under this 211 | License to anyone who comes into possession of a copy. This 212 | License will therefore apply, along with any applicable section 7 213 | additional terms, to the whole of the work, and all its parts, 214 | regardless of how they are packaged. This License gives no 215 | permission to license the work in any other way, but it does not 216 | invalidate such permission if you have separately received it. 217 | 218 | d) If the work has interactive user interfaces, each must display 219 | Appropriate Legal Notices; however, if the Program has interactive 220 | interfaces that do not display Appropriate Legal Notices, your 221 | work need not make them do so. 222 | 223 | A compilation of a covered work with other separate and independent 224 | works, which are not by their nature extensions of the covered work, 225 | and which are not combined with it such as to form a larger program, 226 | in or on a volume of a storage or distribution medium, is called an 227 | "aggregate" if the compilation and its resulting copyright are not 228 | used to limit the access or legal rights of the compilation's users 229 | beyond what the individual works permit. Inclusion of a covered work 230 | in an aggregate does not cause this License to apply to the other 231 | parts of the aggregate. 232 | 233 | 6. Conveying Non-Source Forms. 234 | 235 | You may convey a covered work in object code form under the terms 236 | of sections 4 and 5, provided that you also convey the 237 | machine-readable Corresponding Source under the terms of this License, 238 | in one of these ways: 239 | 240 | a) Convey the object code in, or embodied in, a physical product 241 | (including a physical distribution medium), accompanied by the 242 | Corresponding Source fixed on a durable physical medium 243 | customarily used for software interchange. 244 | 245 | b) Convey the object code in, or embodied in, a physical product 246 | (including a physical distribution medium), accompanied by a 247 | written offer, valid for at least three years and valid for as 248 | long as you offer spare parts or customer support for that product 249 | model, to give anyone who possesses the object code either (1) a 250 | copy of the Corresponding Source for all the software in the 251 | product that is covered by this License, on a durable physical 252 | medium customarily used for software interchange, for a price no 253 | more than your reasonable cost of physically performing this 254 | conveying of source, or (2) access to copy the 255 | Corresponding Source from a network server at no charge. 256 | 257 | c) Convey individual copies of the object code with a copy of the 258 | written offer to provide the Corresponding Source. This 259 | alternative is allowed only occasionally and noncommercially, and 260 | only if you received the object code with such an offer, in accord 261 | with subsection 6b. 262 | 263 | d) Convey the object code by offering access from a designated 264 | place (gratis or for a charge), and offer equivalent access to the 265 | Corresponding Source in the same way through the same place at no 266 | further charge. You need not require recipients to copy the 267 | Corresponding Source along with the object code. If the place to 268 | copy the object code is a network server, the Corresponding Source 269 | may be on a different server (operated by you or a third party) 270 | that supports equivalent copying facilities, provided you maintain 271 | clear directions next to the object code saying where to find the 272 | Corresponding Source. Regardless of what server hosts the 273 | Corresponding Source, you remain obligated to ensure that it is 274 | available for as long as needed to satisfy these requirements. 275 | 276 | e) Convey the object code using peer-to-peer transmission, provided 277 | you inform other peers where the object code and Corresponding 278 | Source of the work are being offered to the general public at no 279 | charge under subsection 6d. 280 | 281 | A separable portion of the object code, whose source code is excluded 282 | from the Corresponding Source as a System Library, need not be 283 | included in conveying the object code work. 284 | 285 | A "User Product" is either (1) a "consumer product", which means any 286 | tangible personal property which is normally used for personal, family, 287 | or household purposes, or (2) anything designed or sold for incorporation 288 | into a dwelling. In determining whether a product is a consumer product, 289 | doubtful cases shall be resolved in favor of coverage. For a particular 290 | product received by a particular user, "normally used" refers to a 291 | typical or common use of that class of product, regardless of the status 292 | of the particular user or of the way in which the particular user 293 | actually uses, or expects or is expected to use, the product. A product 294 | is a consumer product regardless of whether the product has substantial 295 | commercial, industrial or non-consumer uses, unless such uses represent 296 | the only significant mode of use of the product. 297 | 298 | "Installation Information" for a User Product means any methods, 299 | procedures, authorization keys, or other information required to install 300 | and execute modified versions of a covered work in that User Product from 301 | a modified version of its Corresponding Source. The information must 302 | suffice to ensure that the continued functioning of the modified object 303 | code is in no case prevented or interfered with solely because 304 | modification has been made. 305 | 306 | If you convey an object code work under this section in, or with, or 307 | specifically for use in, a User Product, and the conveying occurs as 308 | part of a transaction in which the right of possession and use of the 309 | User Product is transferred to the recipient in perpetuity or for a 310 | fixed term (regardless of how the transaction is characterized), the 311 | Corresponding Source conveyed under this section must be accompanied 312 | by the Installation Information. But this requirement does not apply 313 | if neither you nor any third party retains the ability to install 314 | modified object code on the User Product (for example, the work has 315 | been installed in ROM). 316 | 317 | The requirement to provide Installation Information does not include a 318 | requirement to continue to provide support service, warranty, or updates 319 | for a work that has been modified or installed by the recipient, or for 320 | the User Product in which it has been modified or installed. Access to a 321 | network may be denied when the modification itself materially and 322 | adversely affects the operation of the network or violates the rules and 323 | protocols for communication across the network. 324 | 325 | Corresponding Source conveyed, and Installation Information provided, 326 | in accord with this section must be in a format that is publicly 327 | documented (and with an implementation available to the public in 328 | source code form), and must require no special password or key for 329 | unpacking, reading or copying. 330 | 331 | 7. Additional Terms. 332 | 333 | "Additional permissions" are terms that supplement the terms of this 334 | License by making exceptions from one or more of its conditions. 335 | Additional permissions that are applicable to the entire Program shall 336 | be treated as though they were included in this License, to the extent 337 | that they are valid under applicable law. If additional permissions 338 | apply only to part of the Program, that part may be used separately 339 | under those permissions, but the entire Program remains governed by 340 | this License without regard to the additional permissions. 341 | 342 | When you convey a copy of a covered work, you may at your option 343 | remove any additional permissions from that copy, or from any part of 344 | it. (Additional permissions may be written to require their own 345 | removal in certain cases when you modify the work.) You may place 346 | additional permissions on material, added by you to a covered work, 347 | for which you have or can give appropriate copyright permission. 348 | 349 | Notwithstanding any other provision of this License, for material you 350 | add to a covered work, you may (if authorized by the copyright holders of 351 | that material) supplement the terms of this License with terms: 352 | 353 | a) Disclaiming warranty or limiting liability differently from the 354 | terms of sections 15 and 16 of this License; or 355 | 356 | b) Requiring preservation of specified reasonable legal notices or 357 | author attributions in that material or in the Appropriate Legal 358 | Notices displayed by works containing it; or 359 | 360 | c) Prohibiting misrepresentation of the origin of that material, or 361 | requiring that modified versions of such material be marked in 362 | reasonable ways as different from the original version; or 363 | 364 | d) Limiting the use for publicity purposes of names of licensors or 365 | authors of the material; or 366 | 367 | e) Declining to grant rights under trademark law for use of some 368 | trade names, trademarks, or service marks; or 369 | 370 | f) Requiring indemnification of licensors and authors of that 371 | material by anyone who conveys the material (or modified versions of 372 | it) with contractual assumptions of liability to the recipient, for 373 | any liability that these contractual assumptions directly impose on 374 | those licensors and authors. 375 | 376 | All other non-permissive additional terms are considered "further 377 | restrictions" within the meaning of section 10. If the Program as you 378 | received it, or any part of it, contains a notice stating that it is 379 | governed by this License along with a term that is a further 380 | restriction, you may remove that term. If a license document contains 381 | a further restriction but permits relicensing or conveying under this 382 | License, you may add to a covered work material governed by the terms 383 | of that license document, provided that the further restriction does 384 | not survive such relicensing or conveying. 385 | 386 | If you add terms to a covered work in accord with this section, you 387 | must place, in the relevant source files, a statement of the 388 | additional terms that apply to those files, or a notice indicating 389 | where to find the applicable terms. 390 | 391 | Additional terms, permissive or non-permissive, may be stated in the 392 | form of a separately written license, or stated as exceptions; 393 | the above requirements apply either way. 394 | 395 | 8. Termination. 396 | 397 | You may not propagate or modify a covered work except as expressly 398 | provided under this License. Any attempt otherwise to propagate or 399 | modify it is void, and will automatically terminate your rights under 400 | this License (including any patent licenses granted under the third 401 | paragraph of section 11). 402 | 403 | However, if you cease all violation of this License, then your 404 | license from a particular copyright holder is reinstated (a) 405 | provisionally, unless and until the copyright holder explicitly and 406 | finally terminates your license, and (b) permanently, if the copyright 407 | holder fails to notify you of the violation by some reasonable means 408 | prior to 60 days after the cessation. 409 | 410 | Moreover, your license from a particular copyright holder is 411 | reinstated permanently if the copyright holder notifies you of the 412 | violation by some reasonable means, this is the first time you have 413 | received notice of violation of this License (for any work) from that 414 | copyright holder, and you cure the violation prior to 30 days after 415 | your receipt of the notice. 416 | 417 | Termination of your rights under this section does not terminate the 418 | licenses of parties who have received copies or rights from you under 419 | this License. If your rights have been terminated and not permanently 420 | reinstated, you do not qualify to receive new licenses for the same 421 | material under section 10. 422 | 423 | 9. Acceptance Not Required for Having Copies. 424 | 425 | You are not required to accept this License in order to receive or 426 | run a copy of the Program. Ancillary propagation of a covered work 427 | occurring solely as a consequence of using peer-to-peer transmission 428 | to receive a copy likewise does not require acceptance. However, 429 | nothing other than this License grants you permission to propagate or 430 | modify any covered work. These actions infringe copyright if you do 431 | not accept this License. Therefore, by modifying or propagating a 432 | covered work, you indicate your acceptance of this License to do so. 433 | 434 | 10. Automatic Licensing of Downstream Recipients. 435 | 436 | Each time you convey a covered work, the recipient automatically 437 | receives a license from the original licensors, to run, modify and 438 | propagate that work, subject to this License. You are not responsible 439 | for enforcing compliance by third parties with this License. 440 | 441 | An "entity transaction" is a transaction transferring control of an 442 | organization, or substantially all assets of one, or subdividing an 443 | organization, or merging organizations. If propagation of a covered 444 | work results from an entity transaction, each party to that 445 | transaction who receives a copy of the work also receives whatever 446 | licenses to the work the party's predecessor in interest had or could 447 | give under the previous paragraph, plus a right to possession of the 448 | Corresponding Source of the work from the predecessor in interest, if 449 | the predecessor has it or can get it with reasonable efforts. 450 | 451 | You may not impose any further restrictions on the exercise of the 452 | rights granted or affirmed under this License. For example, you may 453 | not impose a license fee, royalty, or other charge for exercise of 454 | rights granted under this License, and you may not initiate litigation 455 | (including a cross-claim or counterclaim in a lawsuit) alleging that 456 | any patent claim is infringed by making, using, selling, offering for 457 | sale, or importing the Program or any portion of it. 458 | 459 | 11. Patents. 460 | 461 | A "contributor" is a copyright holder who authorizes use under this 462 | License of the Program or a work on which the Program is based. The 463 | work thus licensed is called the contributor's "contributor version". 464 | 465 | A contributor's "essential patent claims" are all patent claims 466 | owned or controlled by the contributor, whether already acquired or 467 | hereafter acquired, that would be infringed by some manner, permitted 468 | by this License, of making, using, or selling its contributor version, 469 | but do not include claims that would be infringed only as a 470 | consequence of further modification of the contributor version. For 471 | purposes of this definition, "control" includes the right to grant 472 | patent sublicenses in a manner consistent with the requirements of 473 | this License. 474 | 475 | Each contributor grants you a non-exclusive, worldwide, royalty-free 476 | patent license under the contributor's essential patent claims, to 477 | make, use, sell, offer for sale, import and otherwise run, modify and 478 | propagate the contents of its contributor version. 479 | 480 | In the following three paragraphs, a "patent license" is any express 481 | agreement or commitment, however denominated, not to enforce a patent 482 | (such as an express permission to practice a patent or covenant not to 483 | sue for patent infringement). To "grant" such a patent license to a 484 | party means to make such an agreement or commitment not to enforce a 485 | patent against the party. 486 | 487 | If you convey a covered work, knowingly relying on a patent license, 488 | and the Corresponding Source of the work is not available for anyone 489 | to copy, free of charge and under the terms of this License, through a 490 | publicly available network server or other readily accessible means, 491 | then you must either (1) cause the Corresponding Source to be so 492 | available, or (2) arrange to deprive yourself of the benefit of the 493 | patent license for this particular work, or (3) arrange, in a manner 494 | consistent with the requirements of this License, to extend the patent 495 | license to downstream recipients. "Knowingly relying" means you have 496 | actual knowledge that, but for the patent license, your conveying the 497 | covered work in a country, or your recipient's use of the covered work 498 | in a country, would infringe one or more identifiable patents in that 499 | country that you have reason to believe are valid. 500 | 501 | If, pursuant to or in connection with a single transaction or 502 | arrangement, you convey, or propagate by procuring conveyance of, a 503 | covered work, and grant a patent license to some of the parties 504 | receiving the covered work authorizing them to use, propagate, modify 505 | or convey a specific copy of the covered work, then the patent license 506 | you grant is automatically extended to all recipients of the covered 507 | work and works based on it. 508 | 509 | A patent license is "discriminatory" if it does not include within 510 | the scope of its coverage, prohibits the exercise of, or is 511 | conditioned on the non-exercise of one or more of the rights that are 512 | specifically granted under this License. You may not convey a covered 513 | work if you are a party to an arrangement with a third party that is 514 | in the business of distributing software, under which you make payment 515 | to the third party based on the extent of your activity of conveying 516 | the work, and under which the third party grants, to any of the 517 | parties who would receive the covered work from you, a discriminatory 518 | patent license (a) in connection with copies of the covered work 519 | conveyed by you (or copies made from those copies), or (b) primarily 520 | for and in connection with specific products or compilations that 521 | contain the covered work, unless you entered into that arrangement, 522 | or that patent license was granted, prior to 28 March 2007. 523 | 524 | Nothing in this License shall be construed as excluding or limiting 525 | any implied license or other defenses to infringement that may 526 | otherwise be available to you under applicable patent law. 527 | 528 | 12. No Surrender of Others' Freedom. 529 | 530 | If conditions are imposed on you (whether by court order, agreement or 531 | otherwise) that contradict the conditions of this License, they do not 532 | excuse you from the conditions of this License. If you cannot convey a 533 | covered work so as to satisfy simultaneously your obligations under this 534 | License and any other pertinent obligations, then as a consequence you may 535 | not convey it at all. For example, if you agree to terms that obligate you 536 | to collect a royalty for further conveying from those to whom you convey 537 | the Program, the only way you could satisfy both those terms and this 538 | License would be to refrain entirely from conveying the Program. 539 | 540 | 13. Remote Network Interaction; Use with the GNU General Public License. 541 | 542 | Notwithstanding any other provision of this License, if you modify the 543 | Program, your modified version must prominently offer all users 544 | interacting with it remotely through a computer network (if your version 545 | supports such interaction) an opportunity to receive the Corresponding 546 | Source of your version by providing access to the Corresponding Source 547 | from a network server at no charge, through some standard or customary 548 | means of facilitating copying of software. This Corresponding Source 549 | shall include the Corresponding Source for any work covered by version 3 550 | of the GNU General Public License that is incorporated pursuant to the 551 | following paragraph. 552 | 553 | Notwithstanding any other provision of this License, you have 554 | permission to link or combine any covered work with a work licensed 555 | under version 3 of the GNU General Public License into a single 556 | combined work, and to convey the resulting work. The terms of this 557 | License will continue to apply to the part which is the covered work, 558 | but the work with which it is combined will remain governed by version 559 | 3 of the GNU General Public License. 560 | 561 | 14. Revised Versions of this License. 562 | 563 | The Free Software Foundation may publish revised and/or new versions of 564 | the GNU Affero General Public License from time to time. Such new versions 565 | will be similar in spirit to the present version, but may differ in detail to 566 | address new problems or concerns. 567 | 568 | Each version is given a distinguishing version number. If the 569 | Program specifies that a certain numbered version of the GNU Affero General 570 | Public License "or any later version" applies to it, you have the 571 | option of following the terms and conditions either of that numbered 572 | version or of any later version published by the Free Software 573 | Foundation. If the Program does not specify a version number of the 574 | GNU Affero General Public License, you may choose any version ever published 575 | by the Free Software Foundation. 576 | 577 | If the Program specifies that a proxy can decide which future 578 | versions of the GNU Affero General Public License can be used, that proxy's 579 | public statement of acceptance of a version permanently authorizes you 580 | to choose that version for the Program. 581 | 582 | Later license versions may give you additional or different 583 | permissions. However, no additional obligations are imposed on any 584 | author or copyright holder as a result of your choosing to follow a 585 | later version. 586 | 587 | 15. Disclaimer of Warranty. 588 | 589 | THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY 590 | APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT 591 | HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY 592 | OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, 593 | THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR 594 | PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM 595 | IS WITH YOU. SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF 596 | ALL NECESSARY SERVICING, REPAIR OR CORRECTION. 597 | 598 | 16. Limitation of Liability. 599 | 600 | IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING 601 | WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS 602 | THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY 603 | GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE 604 | USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF 605 | DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD 606 | PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS), 607 | EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF 608 | SUCH DAMAGES. 609 | 610 | 17. Interpretation of Sections 15 and 16. 611 | 612 | If the disclaimer of warranty and limitation of liability provided 613 | above cannot be given local legal effect according to their terms, 614 | reviewing courts shall apply local law that most closely approximates 615 | an absolute waiver of all civil liability in connection with the 616 | Program, unless a warranty or assumption of liability accompanies a 617 | copy of the Program in return for a fee. 618 | 619 | END OF TERMS AND CONDITIONS 620 | 621 | How to Apply These Terms to Your New Programs 622 | 623 | If you develop a new program, and you want it to be of the greatest 624 | possible use to the public, the best way to achieve this is to make it 625 | free software which everyone can redistribute and change under these terms. 626 | 627 | To do so, attach the following notices to the program. It is safest 628 | to attach them to the start of each source file to most effectively 629 | state the exclusion of warranty; and each file should have at least 630 | the "copyright" line and a pointer to where the full notice is found. 631 | 632 | 633 | Copyright (C) 634 | 635 | This program is free software: you can redistribute it and/or modify 636 | it under the terms of the GNU Affero General Public License as published 637 | by the Free Software Foundation, either version 3 of the License, or 638 | (at your option) any later version. 639 | 640 | This program is distributed in the hope that it will be useful, 641 | but WITHOUT ANY WARRANTY; without even the implied warranty of 642 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 643 | GNU Affero General Public License for more details. 644 | 645 | You should have received a copy of the GNU Affero General Public License 646 | along with this program. If not, see . 647 | 648 | Also add information on how to contact you by electronic and paper mail. 649 | 650 | If your software can interact with users remotely through a computer 651 | network, you should also make sure that it provides a way for users to 652 | get its source. For example, if your program is a web application, its 653 | interface could display a "Source" link that leads users to an archive 654 | of the code. There are many ways you could offer source, and different 655 | solutions will be better for different programs; see section 13 for the 656 | specific requirements. 657 | 658 | You should also get your employer (if you work as a programmer) or school, 659 | if any, to sign a "copyright disclaimer" for the program, if necessary. 660 | For more information on this, and how to apply and follow the GNU AGPL, see 661 | . 662 | -------------------------------------------------------------------------------- /MANIFEST.in: -------------------------------------------------------------------------------- 1 | include Pipfile.lock -------------------------------------------------------------------------------- /Makefile: -------------------------------------------------------------------------------- 1 | .PHONY: all docker clean test 2 | 3 | docker: 4 | sudo docker build . --file Dockerfile --tag markusressel/py-image-dedup:latest 5 | 6 | test: 7 | cd tests; pytest -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | # py-image-dedup [![Build Status](https://img.shields.io/endpoint.svg?url=https%3A%2F%2Factions-badge.atrox.dev%2Fmarkusressel%2Fpy-image-dedup%2Fbadge%3Fref%3Dmaster&style=flat)](https://actions-badge.atrox.dev/markusressel/py-image-dedup/goto?ref=master) [![Code Climate](https://codeclimate.com/github/markusressel/py-image-dedup.svg)](https://codeclimate.com/github/markusressel/py-image-dedup) [![PyPI version](https://badge.fury.io/py/py-image-dedup.svg)](https://badge.fury.io/py/py-image-dedup) 2 | 3 | **py-image-dedup** is a tool to sort out or remove duplicates within a photo library. 4 | Unlike most other solutions, **py-image-dedup** 5 | intentionally uses an approximate image comparison to also detect 6 | duplicates of images that slightly differ in resolution, color or other minor details. 7 | 8 | It is build upon [Image-Match](https://github.com/ascribe/image-match) a very popular library to compute 9 | a pHash for an image and store the result in an ElasticSearch backend for very high scalability. 10 | 11 | [![asciicast](https://asciinema.org/a/3WbBxMXnZyT1QnuTP9fm37wkS.svg)](https://asciinema.org/a/3WbBxMXnZyT1QnuTP9fm37wkS) 12 | 13 | # How it works 14 | 15 | ### Phase 1 - Database cleanup 16 | 17 | In the first phase the elasticsearch backend is checked against the 18 | current filesystem state, cleaning up database entries of files that 19 | no longer exist. This will speed up queries made later on. 20 | 21 | ### Phase 2 - Counting files 22 | 23 | Although not necessary for the deduplication process it is very convenient 24 | to have some kind of progress indication while the deduplication process 25 | is at work. To be able to provide that, available files must be counted beforehand. 26 | 27 | ### Phase 3 - Analysing files 28 | 29 | In this phase every image file is analysed. This means generating a signature (pHash) 30 | to quickly compare it to other images and adding other metadata of the image 31 | to the elasticsearch backend that is used in the next phase. 32 | 33 | This phase is quite CPU intensive and the first run take take quite 34 | some time. Using as much threads as feasible (using the `-t` parameter) 35 | is advised to get the best performance. 36 | 37 | Since we might already have a previous version of this file in the database 38 | before analysing a given file the file modification time is compared to the 39 | given one. If the database content seems to be still correct the signature 40 | for this file will **not** be recalculated. Because of this, subsequent 41 | runs will be much faster. There still has to happen some file access though, 42 | so it is probably limited by that. 43 | 44 | ### Phase 4 - Finding duplicates 45 | 46 | Every file is now processed again - but only by means of querying the 47 | database backend for similar images (within the given `max_dist`). 48 | If there are images found that match the similarity criteria they are considered 49 | duplicate candidates. All candidates are then ordered according to the `prioritization_rules`, 50 | which you can specify yourself in the configuration, see [Configuration](#Configuration). 51 | 52 | If you do not specify `prioritization_rules` yourself, the following order will 53 | be used: 54 | 55 | 1. pixel count (more is better) 56 | 1. EXIF data (more exif data is better) 57 | 1. file size (bigger is better) 58 | 1. file modification time (newer is better) 59 | 1. distance (lower is better) 60 | 1. filename contains "copy" (False is better) 61 | 1. filename length (longer is better) - (for "edited" versions) 62 | 1. parent folder path length (shorter is better) 63 | 1. score (higher is better) 64 | 65 | The first candidate in the resulting list is considered to be the best 66 | available version of all candidates. 67 | 68 | ### Phase 5 - Moving/Deleting duplicates 69 | 70 | All but the best version of duplicate candidates identified in the previous 71 | phase are now deleted from the file system (if you didn't specify `--dry-run` of course). 72 | 73 | If `duplicates_target_directory` is set, the specified folder will be used as 74 | a root directory to move duplicates to, instead of deleting them, replicating their original 75 | folder structure. 76 | 77 | ### Phase 6 - Removing empty folders (Optional) 78 | 79 | In the last phase, folders that are empty due to the deduplication 80 | process are deleted, cleaning up the directory structure (if turned on in configuration). 81 | 82 | # How to use 83 | 84 | ## Install 85 | 86 | Install **py-image-dedup** using pip: 87 | 88 | ```shell 89 | pip3 install py-image-dedup 90 | ``` 91 | 92 | ## Configuration 93 | 94 | **py-image-dedup** uses [container-app-conf](https://github.com/markusressel/container-app-conf) 95 | to provide configuration via a YAML file as well as ENV variables which 96 | generates a reference config on startup. Have a look at the 97 | [documentation about it](https://github.com/markusressel/container-app-conf#generate-reference-config). 98 | 99 | See [py_image_dedup_reference.yaml](/py_image_dedup_reference.yaml) 100 | for an example in this repo. 101 | 102 | ## Setup elasticsearch backend 103 | 104 | Since this library is based on [Image-Match](https://github.com/ascribe/image-match) 105 | you need a running elasticsearch instance for efficient storing and 106 | querying of image signatures. 107 | 108 | ### Elasticsearch version 109 | 110 | This library requires elasticsearch version 5 or later. Sadly the 111 | [Image-Match](https://github.com/ascribe/image-match) library still 112 | specifies version 2, so [a fork of the original project](https://github.com/markusressel/image-match) 113 | is used instead. This fork is maintained by me, and any contributions 114 | are very much appreciated. 115 | 116 | ### Set up the index 117 | 118 | **py-image-dedup** uses a single index (called `images` by default). 119 | When configured, this index will be created automatically for you. 120 | 121 | ## Command line usage 122 | 123 | **py-image-dedup** can be used from the command line like this: 124 | 125 | ```shell 126 | py-image-dedup deduplicate --help 127 | ``` 128 | 129 | Have a look at the help output to see how you can customize it. 130 | 131 | ### Daemon 132 | 133 | **CAUTION!** This feature is still very much a work in progress. 134 | **Always** have a backup of your data! 135 | 136 | **py-image-dedup** has a built in daemon that allows you to continuously 137 | monitor your source directories and deduplicate them on the fly. 138 | 139 | When running the daemon (and enabled in configuration) a prometheus reporter 140 | is used to allow you to gather some statistical insights. 141 | 142 | ```shell 143 | py-image-dedup daemon 144 | ``` 145 | 146 | ## Dry run 147 | 148 | To analyze images and get an overview of what images would be deleted 149 | be sure to make a dry run first. 150 | 151 | ```shell 152 | py-image-dedup deduplicate --dry-run 153 | ``` 154 | 155 | 156 | ## FreeBSD 157 | 158 | If you want to run this on a FreeBSD host make sure you have an up 159 | to date release that is able to install ports. 160 | 161 | Since [Image-Match](https://github.com/ascribe/image-match) does a lot of 162 | math it relies on `numpy` and `scipy`. To get those working on FreeBSD 163 | you have to install them as a port: 164 | 165 | ```shell 166 | pkg install pkgconf 167 | pkg install py38-numpy 168 | pkg install py27-scipy 169 | ``` 170 | 171 | For `.png` support you also need to install 172 | ```shell 173 | pkg install png 174 | ``` 175 | 176 | I still ran into issues after installing all these and just threw those 177 | two in the mix and it finally worked: 178 | ```shell 179 | pkg install freetype 180 | pkg install py27-matplotlib # this has a LOT of dependencies 181 | ``` 182 | 183 | ### Encoding issues 184 | 185 | When using the python library `click` on FreeBSD you might run into 186 | encoding issues. To mitigate this change your locale from `ANSII` to `UTF-8` 187 | if possible. 188 | 189 | This can be achieved f.ex. by creating a file `~/.login_conf` with the following content: 190 | ```text 191 | me:\ 192 | :charset=ISO-8859-1:\ 193 | :lang=de_DE.UTF-8: 194 | ``` 195 | 196 | ## Docker 197 | 198 | To run **py-image-dedup** using docker you can use the [markusressel/py-image-dedup](https://hub.docker.com/r/markusressel/py-image-dedup) 199 | image from DockerHub: 200 | 201 | ``` 202 | sudo docker run -t \ 203 | -p 8000:8000 \ 204 | -v /where/the/original/photolibrary/is/located:/data/in \ 205 | -v /where/duplicates/should/be/moved/to:/data/out \ 206 | -e PY_IMAGE_DEDUP_DRY_RUN=False \ 207 | -e PY_IMAGE_DEDUP_ANALYSIS_SOURCE_DIRECTORIES=/data/in/ \ 208 | -e PY_IMAGE_DEDUP_ANALYSIS_RECURSIVE=True \ 209 | -e PY_IMAGE_DEDUP_ANALYSIS_ACROSS_DIRS=True \ 210 | -e PY_IMAGE_DEDUP_ANALYSIS_FILE_EXTENSIONS=.png,.jpg,.jpeg \ 211 | -e PY_IMAGE_DEDUP_ANALYSIS_THREADS=8 \ 212 | -e PY_IMAGE_DEDUP_ANALYSIS_USE_EXIF_DATA=True \ 213 | -e PY_IMAGE_DEDUP_DEDUPLICATION_DUPLICATES_TARGET_DIRECTORY=/data/out/ \ 214 | -e PY_IMAGE_DEDUP_ELASTICSEARCH_AUTO_CREATE_INDEX=True \ 215 | -e PY_IMAGE_DEDUP_ELASTICSEARCH_HOST=elasticsearch \ 216 | -e PY_IMAGE_DEDUP_ELASTICSEARCH_PORT=9200 \ 217 | -e PY_IMAGE_DEDUP_ELASTICSEARCH_INDEX=images \ 218 | -e PY_IMAGE_DEDUP_ELASTICSEARCH_AUTO_CREATE_INDEX=True \ 219 | -e PY_IMAGE_DEDUP_ELASTICSEARCH_MAX_DISTANCE=0.1 \ 220 | -e PY_IMAGE_DEDUP_REMOVE_EMPTY_FOLDERS=False \ 221 | -e PY_IMAGE_DEDUP_STATS_ENABLED=True \ 222 | -e PY_IMAGE_DEDUP_STATS_PORT=8000 \ 223 | markusressel/py-image-dedup:latest 224 | ``` 225 | 226 | Since an elasticsearch instance is required too, you can 227 | also use the `docker-compose.yml` file included in this repo which will 228 | set up a single-node elasticsearch cluster too: 229 | 230 | ```shell script 231 | sudo docker-compose up 232 | ``` 233 | 234 | ### UID and GID 235 | 236 | To run **py-image-dedup** inside the container using a specific user id 237 | and group id you can use the env variables `PUID=1000` and `PGID=1000`. 238 | 239 | # Contributing 240 | 241 | GitHub is for social coding: if you want to write code, I encourage contributions through pull requests from forks 242 | of this repository. Create GitHub tickets for bugs and new features and comment on the ones that you are interested in. 243 | 244 | # License 245 | 246 | ```text 247 | py-image-dedup by Markus Ressel 248 | Copyright (C) 2018 Markus Ressel 249 | 250 | This program is free software: you can redistribute it and/or modify 251 | it under the terms of the GNU General Public License as published by 252 | the Free Software Foundation, either version 3 of the License, or 253 | (at your option) any later version. 254 | 255 | This program is distributed in the hope that it will be useful, 256 | but WITHOUT ANY WARRANTY; without even the implied warranty of 257 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 258 | GNU General Public License for more details. 259 | 260 | You should have received a copy of the GNU General Public License 261 | along with this program. If not, see . 262 | ``` 263 | -------------------------------------------------------------------------------- /docker-compose.yml: -------------------------------------------------------------------------------- 1 | version: '3.7' 2 | 3 | services: 4 | elasticsearch: 5 | image: docker.elastic.co/elasticsearch/elasticsearch:7.9.2 6 | ports: 7 | - "9200:9200" 8 | - "9300:9300" 9 | environment: 10 | - discovery.type=single-node 11 | networks: 12 | - docker-elk 13 | restart: on-failure 14 | py-image-dedup: 15 | # build: . 16 | image: markusressel/py-image-dedup:latest 17 | environment: 18 | - PUID=1000 19 | - PGID=1000 20 | # change configuration to your liking 21 | - PY_IMAGE_DEDUP_DRY_RUN=True 22 | - PY_IMAGE_DEDUP_ANALYSIS_SOURCE_DIRECTORIES=/mnt/source/ 23 | - PY_IMAGE_DEDUP_ANALYSIS_RECURSIVE=True 24 | - PY_IMAGE_DEDUP_ANALYSIS_ACROSS_DIRS=True 25 | - PY_IMAGE_DEDUP_ANALYSIS_FILE_EXTENSIONS=.png,.jpg,.jpeg 26 | - PY_IMAGE_DEDUP_ANALYSIS_THREADS=8 27 | - PY_IMAGE_DEDUP_ANALYSIS_USE_EXIF_DATA=True 28 | - PY_IMAGE_DEDUP_DEDUPLICATION_DUPLICATES_TARGET_DIRECTORY=/mnt/duplicates/ 29 | - PY_IMAGE_DEDUP_ELASTICSEARCH_HOST=elasticsearch 30 | - PY_IMAGE_DEDUP_ELASTICSEARCH_PORT=9200 31 | - PY_IMAGE_DEDUP_ELASTICSEARCH_INDEX=images 32 | - PY_IMAGE_DEDUP_ELASTICSEARCH_AUTO_CREATE_INDEX=True 33 | - PY_IMAGE_DEDUP_ELASTICSEARCH_MAX_DISTANCE=0.1 34 | - PY_IMAGE_DEDUP_REMOVE_EMPTY_FOLDERS=False 35 | - PY_IMAGE_DEDUP_STATS_ENABLED=True 36 | - PY_IMAGE_DEDUP_STATS_PORT=8000 37 | volumes: 38 | # optionally mount a YAML configuration file 39 | # into /app/py_image_dedup.yaml instead of using environment: 40 | # - /mnt/data3/py_image_dedup.yaml:/app/py_image_dedup.yaml 41 | # change this to your local source directory: 42 | - /mnt/data3/py-image-dedup_testdata:/mnt/source 43 | # change this to your local duplicates directory: 44 | - /mnt/data3/py-image-dedup_duplicates:/mnt/duplicates 45 | links: 46 | - elasticsearch 47 | networks: 48 | - docker-elk 49 | ports: 50 | - "8000:8000" 51 | depends_on: 52 | - elasticsearch 53 | restart: on-failure 54 | networks: 55 | docker-elk: 56 | driver: bridge -------------------------------------------------------------------------------- /docker/entrypoint.sh: -------------------------------------------------------------------------------- 1 | #!/bin/sh 2 | 3 | set -eu 4 | sudo -E -u "#${PUID}" -g "#${PGID}" "$@" 5 | -------------------------------------------------------------------------------- /example.py: -------------------------------------------------------------------------------- 1 | from py_image_dedup.config import DeduplicatorConfig 2 | from py_image_dedup.library.deduplicator import ImageMatchDeduplicator 3 | 4 | config = DeduplicatorConfig() 5 | config.DRY_RUN.value = True 6 | # config.ELASTICSEARCH_HOST.value = "192.168.2.24" 7 | config.SOURCE_DIRECTORIES.value = [ 8 | # r'/home/markus/py-image-dedup/dir1/', 9 | # r'/home/markus/py-image-dedup/dir2/' 10 | # r'/mnt/data/py-dedup-test/Syncthing/', 11 | # r'/mnt/sdb2/Sample/', 12 | r'./tests/images/' 13 | ] 14 | config.SEARCH_ACROSS_ROOT_DIRS.value = True 15 | 16 | config.ANALYSIS_THREADS.value = 8 17 | config.ANALYSIS_USE_EXIF_DATA.value = False 18 | 19 | config.ELASTICSEARCH_MAX_DISTANCE.value = 0.30 20 | # config.MAX_FILE_MODIFICATION_TIME_DELTA.value = timedelta(minutes=5) 21 | config.DEDUPLICATOR_DUPLICATES_TARGET_DIRECTORY.value = "./duplicates/" 22 | config.REMOVE_EMPTY_FOLDERS.value = True 23 | 24 | deduplicator = ImageMatchDeduplicator() 25 | 26 | # max_file_modification_time_diff=1 * 1000 * 60 * 5, 27 | 28 | result = deduplicator.deduplicate_all( 29 | skip_analyze_phase=False, 30 | ) 31 | 32 | result.print_to_console() 33 | -------------------------------------------------------------------------------- /py_image_dedup/__init__.py: -------------------------------------------------------------------------------- 1 | import logging 2 | 3 | logging.basicConfig() 4 | -------------------------------------------------------------------------------- /py_image_dedup/cli.py: -------------------------------------------------------------------------------- 1 | import time 2 | 3 | import click 4 | 5 | from py_image_dedup.config import DeduplicatorConfig 6 | from py_image_dedup.library.deduplicator import ImageMatchDeduplicator 7 | from py_image_dedup.library.processing_manager import ProcessingManager 8 | from py_image_dedup.util import echo 9 | 10 | IMAGE_HASH_MAP = {} 11 | 12 | PARAM_SKIP_ANALYSE_PHASE = "skip-analyse-phase" 13 | PARAM_DRY_RUN = "dry-run" 14 | 15 | CMD_OPTION_NAMES = { 16 | PARAM_SKIP_ANALYSE_PHASE: ['--skip-analyse-phase', '-sap'], 17 | PARAM_DRY_RUN: ['--dry-run', '-dr'] 18 | } 19 | 20 | CONTEXT_SETTINGS = dict(help_option_names=['-h', '--help']) 21 | 22 | 23 | @click.group(context_settings=CONTEXT_SETTINGS) 24 | @click.version_option() 25 | def cli(): 26 | pass 27 | 28 | 29 | def get_option_names(parameter: str) -> list: 30 | """ 31 | Returns a list of all valid console parameter names for a given parameter 32 | :param parameter: the parameter to check 33 | :return: a list of all valid names to use this parameter 34 | """ 35 | return CMD_OPTION_NAMES[parameter] 36 | 37 | 38 | @cli.command(name="analyse") 39 | def c_analyse(): 40 | deduplicator = ImageMatchDeduplicator(interactive=True) 41 | deduplicator.analyse_all() 42 | 43 | 44 | @cli.command(name="deduplicate") 45 | @click.option(*get_option_names(PARAM_SKIP_ANALYSE_PHASE), required=False, default=False, is_flag=True, 46 | help='When set the image analysis phase will be skipped. Useful if you already did a dry-run.') 47 | @click.option(*get_option_names(PARAM_DRY_RUN), required=False, default=None, is_flag=True, 48 | help='When set no files or folders will actually be deleted but a preview of ' 49 | 'what WOULD be done will be printed.') 50 | def c_deduplicate(skip_analyse_phase: bool, 51 | dry_run: bool): 52 | config = DeduplicatorConfig() 53 | if dry_run is not None: 54 | config.DRY_RUN.value = dry_run 55 | deduplicator = ImageMatchDeduplicator(interactive=True) 56 | result = deduplicator.deduplicate_all( 57 | skip_analyze_phase=skip_analyse_phase, 58 | ) 59 | 60 | echo() 61 | result.print_to_console() 62 | 63 | 64 | @cli.command(name="daemon") 65 | @click.option(*get_option_names(PARAM_DRY_RUN), required=False, default=None, is_flag=True, 66 | help='When set no files or folders will actually be deleted but a preview of ' 67 | 'what WOULD be done will be printed.') 68 | def c_daemon(dry_run: bool): 69 | echo("Starting daemon...") 70 | 71 | config: DeduplicatorConfig = DeduplicatorConfig() 72 | if dry_run is not None: 73 | config.DRY_RUN.value = dry_run 74 | 75 | if config.STATS_ENABLED.value: 76 | from prometheus_client import start_http_server 77 | echo("Starting prometheus reporter...") 78 | start_http_server(config.STATS_PORT.value) 79 | 80 | deduplicator = ImageMatchDeduplicator(interactive=False) 81 | processing_manager = ProcessingManager(deduplicator) 82 | 83 | deduplicator.deduplicate_all() 84 | processing_manager.start() 85 | 86 | try: 87 | while True: 88 | time.sleep(1) 89 | except KeyboardInterrupt: 90 | processing_manager.stop() 91 | 92 | 93 | if __name__ == '__main__': 94 | cli() 95 | -------------------------------------------------------------------------------- /py_image_dedup/config.py: -------------------------------------------------------------------------------- 1 | import os 2 | from datetime import timedelta 3 | 4 | from container_app_conf import ConfigBase 5 | from container_app_conf.entry.bool import BoolConfigEntry 6 | from container_app_conf.entry.dict import DictConfigEntry 7 | from container_app_conf.entry.file import DirectoryConfigEntry 8 | from container_app_conf.entry.float import FloatConfigEntry 9 | from container_app_conf.entry.int import IntConfigEntry 10 | from container_app_conf.entry.list import ListConfigEntry 11 | from container_app_conf.entry.regex import RegexConfigEntry 12 | from container_app_conf.entry.string import StringConfigEntry 13 | from container_app_conf.entry.timedelta import TimeDeltaConfigEntry 14 | from container_app_conf.source.env_source import EnvSource 15 | from container_app_conf.source.yaml_source import YamlSource 16 | from py_range_parse import Range 17 | 18 | NODE_MAIN = "py_image_dedup" 19 | 20 | NODE_DAEMON = "daemon" 21 | NODE_PROCESSING_TIMEOUT = "processing_timeout" 22 | 23 | NODE_FILE_OBSERVER_TYPE = "file_observer" 24 | FILE_OBSERVER_TYPE_POLLING = "polling" 25 | FILE_OBSERVER_TYPE_INOTIFY = "inotify" 26 | 27 | NODE_DRY_RUN = "dry_run" 28 | 29 | NODE_ELASTICSEARCH = "elasticsearch" 30 | 31 | NODE_HOST = "host" 32 | NODE_MAX_DISTANCE = "max_distance" 33 | NODE_AUTO_CREATE_INDEX = "auto_create_index" 34 | NODE_INDEX = "index" 35 | 36 | NODE_ANALYSIS = "analysis" 37 | 38 | NODE_EXCLUSIONS = "exclusions" 39 | NODE_SOURCE_DIRECTORIES = "source_directories" 40 | NODE_RECURSIVE = "recursive" 41 | NODE_SEARCH_ACROSS_ROOT_DIRS = "across_dirs" 42 | NODE_FILE_EXTENSIONS = "file_extensions" 43 | NODE_USE_EXIF_DATA = "use_exif_data" 44 | NODE_THREADS = "threads" 45 | 46 | NODE_DEDUPLICATION = "deduplication" 47 | 48 | NODE_PRIORITIZATION_RULES = "prioritization_rules" 49 | NODE_MAX_FILE_MODIFICATION_TIME_DIFF = "max_file_modification_time_diff" 50 | NODE_REMOVE_EMPTY_FOLDERS = "remove_empty_folders" 51 | NODE_DUPLICATES_TARGET_DIRECTORY = "duplicates_target_directory" 52 | 53 | NODE_STATS = "stats" 54 | NODE_ENABLED = "enabled" 55 | NODE_PORT = "port" 56 | 57 | 58 | class DeduplicatorConfig(ConfigBase): 59 | 60 | def __new__(cls, *args, **kwargs): 61 | yaml_source = YamlSource("py_image_dedup") 62 | data_sources = [ 63 | EnvSource(), 64 | yaml_source 65 | ] 66 | return super(DeduplicatorConfig, cls).__new__(cls, *args, data_sources=data_sources, **kwargs) 67 | 68 | DRY_RUN = BoolConfigEntry( 69 | description="If enabled no source file will be touched", 70 | key_path=[ 71 | NODE_MAIN, 72 | NODE_DRY_RUN 73 | ], 74 | default=True 75 | ) 76 | 77 | ELASTICSEARCH_HOST = StringConfigEntry( 78 | description="Hostname of the elasticsearch backend instance to use.", 79 | key_path=[ 80 | NODE_MAIN, 81 | NODE_ELASTICSEARCH, 82 | NODE_HOST 83 | ], 84 | default="127.0.0.1" 85 | ) 86 | 87 | ELASTICSEARCH_PORT = IntConfigEntry( 88 | description="Port of the elasticsearch backend instance to use.", 89 | key_path=[ 90 | NODE_MAIN, 91 | NODE_ELASTICSEARCH, 92 | NODE_PORT 93 | ], 94 | range=Range(1, 65535), 95 | default=9200 96 | ) 97 | 98 | ELASTICSEARCH_MAX_DISTANCE = FloatConfigEntry( 99 | description="Maximum signature distance [0..1] to query from elasticsearch backend.", 100 | key_path=[ 101 | NODE_MAIN, 102 | NODE_ELASTICSEARCH, 103 | NODE_MAX_DISTANCE 104 | ], 105 | default=0.10 106 | ) 107 | 108 | ELASTICSEARCH_AUTO_CREATE_INDEX = BoolConfigEntry( 109 | description="Whether to automatically create an index in the target database.", 110 | key_path=[ 111 | NODE_MAIN, 112 | NODE_ELASTICSEARCH, 113 | NODE_AUTO_CREATE_INDEX 114 | ], 115 | default=True 116 | ) 117 | 118 | ELASTICSEARCH_INDEX = StringConfigEntry( 119 | description="The index name to use for storing and querying image analysis data.", 120 | key_path=[ 121 | NODE_MAIN, 122 | NODE_ELASTICSEARCH, 123 | NODE_INDEX 124 | ], 125 | default="images" 126 | ) 127 | 128 | ANALYSIS_USE_EXIF_DATA = BoolConfigEntry( 129 | description="Whether to scan for EXIF data or not.", 130 | key_path=[ 131 | NODE_MAIN, 132 | NODE_ANALYSIS, 133 | NODE_USE_EXIF_DATA 134 | ], 135 | default=True 136 | ) 137 | 138 | SOURCE_DIRECTORIES = ListConfigEntry( 139 | description="Comma separated list of source paths to analyse and deduplicate.", 140 | item_type=DirectoryConfigEntry, 141 | item_args={ 142 | "check_existence": True 143 | }, 144 | key_path=[ 145 | NODE_MAIN, 146 | NODE_ANALYSIS, 147 | NODE_SOURCE_DIRECTORIES 148 | ], 149 | required=True, 150 | example=[ 151 | "/home/myuser/pictures/" 152 | ] 153 | ) 154 | 155 | RECURSIVE = BoolConfigEntry( 156 | description="When set all directories will be recursively analyzed.", 157 | key_path=[ 158 | NODE_MAIN, 159 | NODE_ANALYSIS, 160 | NODE_RECURSIVE 161 | ], 162 | default=True 163 | ) 164 | 165 | SEARCH_ACROSS_ROOT_DIRS = BoolConfigEntry( 166 | description="When set duplicates will be found even if they are located in different root directories.", 167 | key_path=[ 168 | NODE_MAIN, 169 | NODE_ANALYSIS, 170 | NODE_SEARCH_ACROSS_ROOT_DIRS 171 | ], 172 | default=False 173 | ) 174 | 175 | FILE_EXTENSION_FILTER = ListConfigEntry( 176 | description="Comma separated list of file extensions.", 177 | item_type=StringConfigEntry, 178 | key_path=[ 179 | NODE_MAIN, 180 | NODE_ANALYSIS, 181 | NODE_FILE_EXTENSIONS 182 | ], 183 | required=True, 184 | default=[ 185 | ".png", 186 | ".jpg", 187 | ".jpeg" 188 | ] 189 | ) 190 | 191 | EXCLUSIONS = ListConfigEntry( 192 | description="Comma separated list of regular expression filters.", 193 | item_type=RegexConfigEntry, 194 | key_path=[ 195 | NODE_MAIN, 196 | NODE_ANALYSIS, 197 | NODE_EXCLUSIONS 198 | ], 199 | default=[] 200 | ) 201 | 202 | ANALYSIS_THREADS = IntConfigEntry( 203 | description="Number of threads to use for image analysis phase.", 204 | key_path=[ 205 | NODE_MAIN, 206 | NODE_ANALYSIS, 207 | NODE_THREADS 208 | ], 209 | default=os.cpu_count() 210 | ) 211 | 212 | MAX_FILE_MODIFICATION_TIME_DELTA = TimeDeltaConfigEntry( 213 | description="Maximum file modification date difference between multiple " 214 | "duplicates to be considered the same image", 215 | key_path=[ 216 | NODE_MAIN, 217 | NODE_DEDUPLICATION, 218 | NODE_MAX_FILE_MODIFICATION_TIME_DIFF 219 | ], 220 | default=None, 221 | example=timedelta(minutes=5) 222 | ) 223 | 224 | PRIORITIZATION_RULES = ListConfigEntry( 225 | description="Comma separated list of prioritization rules to use for ordering duplicate " 226 | "images before proceeding with the deduplication process.", 227 | item_type=DictConfigEntry, 228 | key_path=[ 229 | NODE_MAIN, 230 | NODE_DEDUPLICATION, 231 | NODE_PRIORITIZATION_RULES 232 | ], 233 | required=False, 234 | default=[ 235 | {"name": "higher-pixel-count"}, 236 | {"name": "more-exif-data"}, 237 | {"name": "bigger-file-size"}, 238 | {"name": "newer-file-modification-date"}, 239 | {"name": "smaller-distance"}, 240 | {"name": "doesnt-contain-copy-in-file-name"}, 241 | {"name": "longer-file-name"}, 242 | {"name": "shorter-folder-path"}, 243 | {"name": "higher-score"}, 244 | ] 245 | ) 246 | 247 | REMOVE_EMPTY_FOLDERS = BoolConfigEntry( 248 | description="Whether to remove empty folders or not.", 249 | key_path=[ 250 | NODE_MAIN, 251 | NODE_REMOVE_EMPTY_FOLDERS 252 | ], 253 | default=False 254 | ) 255 | 256 | DEDUPLICATOR_DUPLICATES_TARGET_DIRECTORY = DirectoryConfigEntry( 257 | description="Directory path to move duplicates to instead of deleting them.", 258 | key_path=[ 259 | NODE_MAIN, 260 | NODE_DEDUPLICATION, 261 | NODE_DUPLICATES_TARGET_DIRECTORY 262 | ], 263 | check_existence=True, 264 | default=None, 265 | example="/home/myuser/pictures/duplicates/" 266 | ) 267 | 268 | DAEMON_PROCESSING_TIMEOUT = TimeDeltaConfigEntry( 269 | description="Time to wait for filesystems changes to settle before analysing.", 270 | key_path=[ 271 | NODE_MAIN, 272 | NODE_DAEMON, 273 | NODE_PROCESSING_TIMEOUT 274 | ], 275 | default="30s" 276 | ) 277 | 278 | DAEMON_FILE_OBSERVER_TYPE = StringConfigEntry( 279 | description="Type of file observer to use.", 280 | key_path=[ 281 | NODE_MAIN, 282 | NODE_DAEMON, 283 | NODE_FILE_OBSERVER_TYPE 284 | ], 285 | regex="|".join([FILE_OBSERVER_TYPE_POLLING, FILE_OBSERVER_TYPE_INOTIFY]), 286 | default=FILE_OBSERVER_TYPE_POLLING, 287 | required=True 288 | ) 289 | 290 | STATS_ENABLED = BoolConfigEntry( 291 | description="Whether to enable prometheus statistics or not.", 292 | key_path=[ 293 | NODE_MAIN, 294 | NODE_STATS, 295 | NODE_ENABLED 296 | ], 297 | default=True 298 | ) 299 | 300 | STATS_PORT = IntConfigEntry( 301 | description="The port to expose statistics on.", 302 | key_path=[ 303 | NODE_MAIN, 304 | NODE_STATS, 305 | NODE_PORT 306 | ], 307 | default=8000 308 | ) 309 | -------------------------------------------------------------------------------- /py_image_dedup/library/__init__.py: -------------------------------------------------------------------------------- 1 | import logging 2 | import threading 3 | 4 | LOGGER = logging.getLogger(__name__) 5 | LOGGER.setLevel(logging.DEBUG) 6 | 7 | 8 | class Action: 9 | def __init__(self, name, color, ): 10 | self.name = name 11 | self.color = color 12 | 13 | 14 | class ActionEnum: 15 | NONE = Action("-", "green") 16 | DELETE = Action("delete", "red") 17 | MOVE = Action("move", "yellow") 18 | 19 | 20 | class RegularIntervalWorker: 21 | """ 22 | Base class for a worker that executes a specific task in a regular interval. 23 | """ 24 | 25 | def __init__(self, interval: float): 26 | self._interval = interval 27 | self._timer = None 28 | 29 | def start(self): 30 | """ 31 | Starts the worker 32 | """ 33 | if self._timer is None: 34 | LOGGER.debug(f"Starting worker: {self.__class__.__name__}") 35 | self._schedule_next_run() 36 | else: 37 | LOGGER.debug("Already running, ignoring start() call") 38 | 39 | def stop(self): 40 | """ 41 | Stops the worker 42 | """ 43 | if self._timer is not None: 44 | self._timer.cancel() 45 | self._timer = None 46 | 47 | def _schedule_next_run(self): 48 | """ 49 | Schedules the next run 50 | """ 51 | if self._timer is not None: 52 | self._timer.cancel() 53 | self._timer = threading.Timer(self._interval, self._worker_job) 54 | self._timer.start() 55 | 56 | def _worker_job(self): 57 | """ 58 | The regularly executed task. Override this method. 59 | """ 60 | try: 61 | self._run() 62 | except Exception as e: 63 | LOGGER.error(e, exc_info=True) 64 | finally: 65 | self._schedule_next_run() 66 | 67 | def _run(self): 68 | """ 69 | The regularly executed task. Override this method. 70 | """ 71 | raise NotImplementedError() 72 | -------------------------------------------------------------------------------- /py_image_dedup/library/deduplication_result.py: -------------------------------------------------------------------------------- 1 | from pathlib import Path 2 | from typing import List 3 | 4 | import click 5 | from tabulate import tabulate 6 | 7 | from py_image_dedup.library import ActionEnum 8 | from py_image_dedup.persistence import MetadataKey 9 | from py_image_dedup.util import echo 10 | 11 | BYTE_IN_A_MB = 1048576 12 | 13 | 14 | class DeduplicationResult: 15 | def __init__(self): 16 | self.item_actions = {} 17 | self._removed_folders = set() 18 | self._reference_files = {} 19 | self._file_duplicates = {} 20 | 21 | def add_file_action(self, file_path: Path, action: ActionEnum): 22 | if file_path in self.item_actions and self.item_actions[file_path] != action: 23 | raise ValueError("File path already in result " 24 | "but with different action: {}, {}, {}".format(file_path, 25 | self.item_actions[file_path], 26 | action)) 27 | self.item_actions[file_path] = action 28 | 29 | def get_file_with_action(self, action: ActionEnum) -> []: 30 | return list({k: v for k, v in self.item_actions.items() if v == action}.keys()) 31 | 32 | def get_duplicate_count(self) -> int: 33 | """ 34 | :return: amount of files that have at least one duplicate 35 | """ 36 | count = 0 37 | for key, value in self._file_duplicates.items(): 38 | if len(value) > 0: 39 | count += 1 40 | 41 | return count 42 | 43 | def get_removed_or_moved_files(self): 44 | return self.get_file_with_action(ActionEnum.MOVE) + self.get_file_with_action(ActionEnum.DELETE) 45 | 46 | def get_removed_empty_folders(self) -> []: 47 | """ 48 | :return: a list of empty folders that have been deleted 49 | """ 50 | return self._removed_folders 51 | 52 | def add_removed_empty_folder(self, folder: Path): 53 | """ 54 | Adds a folder to the list of removed empty folders 55 | :param folder: the folder to add 56 | """ 57 | self._removed_folders.add(folder) 58 | 59 | def set_file_duplicates(self, reference_files: List[dict], duplicate_files: []): 60 | """ 61 | Set a list of files that are duplicates of the reference file 62 | :param reference_files: the file that is used as a baseline 63 | :param duplicate_files: duplicates of the reference_file 64 | """ 65 | reference_file = reference_files[0] 66 | reference_file_path = Path(reference_file[MetadataKey.PATH.value]) 67 | self._reference_files[reference_file_path] = reference_file 68 | self._file_duplicates[reference_file_path] = reference_files[1:] + duplicate_files 69 | 70 | def get_file_duplicates(self) -> {}: 71 | """ 72 | Get a list of files that are duplicates of other files 73 | """ 74 | return self._file_duplicates 75 | 76 | def print_to_console(self): 77 | title = "" * 7 + "Summary" 78 | echo(title, color='cyan') 79 | echo('=' * 21, color='cyan') 80 | echo(f"Files with duplicates: {self.get_duplicate_count()}") 81 | echo(f"Files moved: {len(self.get_file_with_action(ActionEnum.MOVE))}") 82 | echo(f"Files deleted: {len(self.get_file_with_action(ActionEnum.DELETE))}") 83 | 84 | headers = ("Action", "File path", "Dist", "Filesize", "Pixels") 85 | 86 | for reference_file_path, folder in self.get_file_duplicates().items(): 87 | duplicate_count = len(folder) 88 | if duplicate_count > 0: 89 | columns = [] 90 | echo() 91 | 92 | for item in [self._reference_files[reference_file_path]] + folder: 93 | file_path = Path(item[MetadataKey.PATH.value]) 94 | distance = item[MetadataKey.DISTANCE.value] 95 | distance_rounded = round(distance, 3) 96 | file_size = item[MetadataKey.METADATA.value][MetadataKey.FILE_SIZE.value] 97 | file_size_mb = round(file_size / BYTE_IN_A_MB, 3) 98 | pixel_count = item[MetadataKey.METADATA.value][MetadataKey.PIXELCOUNT.value] 99 | 100 | action = self.item_actions.get(file_path, ActionEnum.NONE) 101 | row = [ 102 | action.name, 103 | file_path, 104 | distance_rounded, 105 | file_size_mb, 106 | pixel_count 107 | ] 108 | 109 | # apply action style 110 | row = list(map(lambda x: str(click.style(str(x), action.color)), row)) 111 | columns.append(row) 112 | 113 | self._echo_table( 114 | tabulate(columns, headers=headers, colalign=['center', 'left', 'left', 'right', 'right'])) 115 | 116 | echo() 117 | echo(f"Removed (empty) folders ({len(self.get_removed_empty_folders())}):") 118 | for folder in self.get_removed_empty_folders(): 119 | echo(f"{folder}", color='red') 120 | 121 | @staticmethod 122 | def _echo_table(table: str): 123 | lines = table.splitlines() 124 | 125 | for line in lines[:2]: 126 | echo(line, color='cyan') 127 | 128 | for line in lines[2:]: 129 | echo(line) 130 | -------------------------------------------------------------------------------- /py_image_dedup/library/deduplicator.py: -------------------------------------------------------------------------------- 1 | import datetime 2 | import filecmp 3 | import logging 4 | import os 5 | import shutil 6 | import sys 7 | from concurrent.futures import ThreadPoolExecutor 8 | from pathlib import Path 9 | from typing import List 10 | 11 | import click 12 | from ordered_set import OrderedSet 13 | 14 | from py_image_dedup import util 15 | from py_image_dedup.config import DeduplicatorConfig 16 | from py_image_dedup.library import ActionEnum 17 | from py_image_dedup.library.deduplication_result import DeduplicationResult 18 | from py_image_dedup.library.progress_manager import ProgressManager 19 | from py_image_dedup.persistence import ImageSignatureStore 20 | from py_image_dedup.persistence.elasticsearchstorebackend import ElasticSearchStoreBackend 21 | from py_image_dedup.persistence.metadata_key import MetadataKey 22 | from py_image_dedup.stats import DUPLICATE_ACTION_MOVE_COUNT, DUPLICATE_ACTION_DELETE_COUNT, ANALYSIS_TIME, \ 23 | FIND_DUPLICATES_TIME 24 | from py_image_dedup.util import file, echo 25 | from py_image_dedup.util.file import get_files_count, file_has_extension 26 | 27 | LOGGER = logging.getLogger(__name__) 28 | LOGGER.setLevel(logging.DEBUG) 29 | 30 | 31 | class ImageMatchDeduplicator: 32 | EXECUTOR = ThreadPoolExecutor() 33 | 34 | _config: DeduplicatorConfig 35 | _progress_manager: ProgressManager 36 | 37 | _processed_files: dict = {} 38 | _deduplication_result: DeduplicationResult = None 39 | 40 | def __init__(self, interactive: bool): 41 | """ 42 | 43 | :param interactive: whether cli output should be interactive or not 44 | """ 45 | self.interactive = interactive 46 | 47 | self._progress_manager = ProgressManager() 48 | self._config = DeduplicatorConfig() 49 | self._persistence: ImageSignatureStore = ElasticSearchStoreBackend( 50 | host=self._config.ELASTICSEARCH_HOST.value, 51 | port=self._config.ELASTICSEARCH_PORT.value, 52 | connections_per_node=self._config.ANALYSIS_THREADS.value, 53 | el_index=self._config.ELASTICSEARCH_INDEX.value, 54 | use_exif_data=self._config.ANALYSIS_USE_EXIF_DATA.value, 55 | max_dist=self._config.ELASTICSEARCH_MAX_DISTANCE.value, 56 | setup_database=self._config.ELASTICSEARCH_AUTO_CREATE_INDEX.value 57 | ) 58 | 59 | def reset_result(self): 60 | self._deduplication_result = DeduplicationResult() 61 | self._processed_files = {} 62 | 63 | def analyse_all(self): 64 | """ 65 | Runs the analysis phase independently. 66 | """ 67 | directories = self._config.SOURCE_DIRECTORIES.value 68 | 69 | echo("Phase 1/2: Counting files ...", color='cyan') 70 | directory_map = self._count_files(directories) 71 | 72 | echo("Phase 2/2: Analyzing files ...", color='cyan') 73 | self.analyze_directories(directory_map) 74 | 75 | def deduplicate_all(self, skip_analyze_phase: bool = False) -> DeduplicationResult: 76 | """ 77 | Runs the full 6 deduplication phases. 78 | :param skip_analyze_phase: useful if you already did a dry run and want to do a real run afterwards 79 | :return: result of the operation 80 | """ 81 | # see: https://stackoverflow.com/questions/14861891/runtimewarning-invalid-value-encountered-in-divide 82 | # and: https://stackoverflow.com/questions/29347987/why-cant-i-suppress-numpy-warnings 83 | import warnings 84 | warnings.filterwarnings('ignore') 85 | 86 | directories = self._config.SOURCE_DIRECTORIES.value 87 | if len(directories) <= 0: 88 | raise ValueError("No root directories to scan") 89 | 90 | if self._config.DRY_RUN.value: 91 | echo("==> DRY RUN! No files or folders will actually be deleted! <==", color='yellow') 92 | 93 | echo("Phase 1/6: Cleaning up database ...", color='cyan') 94 | self.cleanup_database(directories) 95 | 96 | echo("Phase 2/6: Counting files ...", color='cyan') 97 | directory_map = self._count_files(directories) 98 | 99 | phase_3_text = "Phase 3/6: Analyzing files" 100 | if skip_analyze_phase: 101 | echo(phase_3_text + " - Skipping", color='yellow') 102 | else: 103 | echo(phase_3_text, color='cyan') 104 | self.analyze_directories(directory_map) 105 | 106 | echo("Phase 4/6: Finding duplicate files ...", color='cyan') 107 | self.find_duplicates_in_directories(directory_map) 108 | 109 | # Phase 5/6: Move or Delete duplicate files 110 | self.process_duplicates() 111 | 112 | self.remove_empty_folders() 113 | 114 | return self._deduplication_result 115 | 116 | def analyze_directories(self, directory_map: dict): 117 | """ 118 | Analyzes all files, generates identifiers (if necessary) and stores them for later access 119 | """ 120 | threads = self._config.ANALYSIS_THREADS.value 121 | 122 | # load truncated images too 123 | # TODO: this causes an infinite loop on some (truncated) images 124 | # ImageFile.LOAD_TRUNCATED_IMAGES = True 125 | 126 | for directory, file_count in directory_map.items(): 127 | self._progress_manager.start(f"Analyzing files in '{directory}'", file_count, "Files", self.interactive) 128 | self.__walk_directory_files( 129 | root_directory=directory, 130 | threads=threads, 131 | command=lambda root_dir, file_dir, file_path: self.analyze_file(file_path)) 132 | self._progress_manager.clear() 133 | 134 | def find_duplicates_in_directories(self, directory_map: dict): 135 | """ 136 | Finds duplicates in the given directories 137 | :param directory_map: map of directory path -> file count 138 | """ 139 | self.reset_result() 140 | 141 | for directory, file_count in directory_map.items(): 142 | self._progress_manager.start(f"Finding duplicates in '{directory}' ...", file_count, "Files", 143 | self.interactive) 144 | self.__walk_directory_files( 145 | root_directory=directory, 146 | threads=1, # there seems to be no performance advantage in using multiple threads here 147 | command=lambda root_dir, _, file_path: self.find_duplicates_of_file( 148 | root_directories=self._config.SOURCE_DIRECTORIES.value, 149 | root_directory=root_dir, 150 | reference_file_path=file_path 151 | ) 152 | ) 153 | self._progress_manager.clear() 154 | 155 | def cleanup_database(self, directories: List[Path]): 156 | """ 157 | Removes database entries of files that don't exist on disk. 158 | Note that this cleanup will only consider files within one 159 | of the root directories specified in constructor, as other file paths 160 | might have been added on other machines. 161 | :param directories: directories in this run 162 | """ 163 | # TODO: This iterates through all db entries - even the ones we are ignoring. 164 | # The db query should be improved to speed this up 165 | 166 | count, entries = self._persistence.get_all() 167 | if count <= 0: 168 | return 169 | 170 | self._progress_manager.start(f"Cleanup database", count, "entries", self.interactive) 171 | for entry in entries: 172 | try: 173 | image_entry = entry['_source'] 174 | metadata = image_entry.get(MetadataKey.METADATA.value, {}) 175 | 176 | file_path = Path(image_entry[MetadataKey.PATH.value]) 177 | self._progress_manager.set_postfix(self._truncate_middle(str(file_path))) 178 | 179 | if MetadataKey.DATAMODEL_VERSION.value not in metadata: 180 | echo(f"Removing db entry with missing db model version number: {file_path}") 181 | self._persistence.remove(str(file_path)) 182 | continue 183 | 184 | data_version = metadata.get(MetadataKey.DATAMODEL_VERSION.value, -1) 185 | if data_version != self._persistence.DATAMODEL_VERSION: 186 | echo(f"Removing db entry with old db model version: {file_path}") 187 | self._persistence.remove(str(file_path)) 188 | continue 189 | 190 | # filter by files in at least one of the specified root directories 191 | # this is necessary because the database might hold items for other paths already 192 | # and those are not interesting to us 193 | if not any(root_dir in file_path.parents for root_dir in directories): 194 | continue 195 | 196 | if not file_path.exists(): 197 | echo(f"Removing db entry for missing file: {file_path}") 198 | self._persistence.remove(str(file_path)) 199 | except Exception as e: 200 | logging.exception(e) 201 | echo(f"Error while cleaning up database entry {entry}: {e}") 202 | try: 203 | image_entry = entry['_source'] 204 | file_path = Path(image_entry[MetadataKey.PATH.value]) 205 | self._persistence.remove(str(file_path)) 206 | except Exception as e: 207 | logging.exception(e) 208 | echo(f"Error removing db entry: {e}") 209 | finally: 210 | self._progress_manager.inc() 211 | self._progress_manager.clear() 212 | 213 | def _remove_empty_folders(self, directories: List[Path], recursive: bool): 214 | """ 215 | Searches for empty folders and removes them 216 | :param directories: directories to scan 217 | """ 218 | dry_run = self._config.DRY_RUN.value 219 | 220 | # remove empty folders 221 | for directory in directories: 222 | empty_folders = self._find_empty_folders(directory, recursive, dry_run) 223 | self._remove_folders(directory, empty_folders, dry_run) 224 | 225 | def _count_files(self, directories: List[Path]) -> dict: 226 | """ 227 | Counts the amount of files to analyze (used in progress) and stores them in a map 228 | :return map "directory path" -> "directory file count" 229 | """ 230 | directory_map = {} 231 | 232 | self._progress_manager.start(f"Counting files", len(directories), "Dirs", self.interactive) 233 | for directory in directories: 234 | self._progress_manager.set_postfix(self._truncate_middle(directory)) 235 | 236 | file_count = get_files_count( 237 | directory, 238 | self._config.RECURSIVE.value, 239 | self._config.FILE_EXTENSION_FILTER.value, 240 | self._config.EXCLUSIONS.value 241 | ) 242 | directory_map[directory] = file_count 243 | 244 | self._progress_manager.inc() 245 | self._progress_manager.clear() 246 | 247 | return directory_map 248 | 249 | def __walk_directory_files(self, root_directory: Path, command, threads: int): 250 | """ 251 | Walks through the files of the given directory 252 | :param root_directory: the directory to start with 253 | :param command: the method to execute for every file found 254 | :return: file_path -> identifier 255 | """ 256 | with ThreadPoolExecutor(max_workers=threads, thread_name_prefix="py-image-dedup-walker") as self.EXECUTOR: 257 | for (root, dirs, files) in os.walk(str(root_directory)): 258 | # root is the place you're listing 259 | # dirs is a list of directories directly under root 260 | # files is a list of files directly under root 261 | root = Path(root) 262 | 263 | for file in files: 264 | file_path = Path(root, file) 265 | 266 | # skip file in exclusion 267 | if any(list(map(lambda x: x.search(str(file_path.absolute())), self._config.EXCLUSIONS.value))): 268 | continue 269 | 270 | # skip file with unwanted file extension 271 | if not file_has_extension(file_path, self._config.FILE_EXTENSION_FILTER.value): 272 | continue 273 | 274 | # skip if not existent (probably already deleted) 275 | if not file_path.exists(): 276 | self._progress_manager.inc() 277 | continue 278 | 279 | try: 280 | self.EXECUTOR.submit(util.reraise_with_stack(command), root_directory, root, file_path) 281 | except Exception as e: 282 | click.echo(e, err=True) 283 | sys.exit(1) 284 | 285 | if not self._config.RECURSIVE.value: 286 | return 287 | 288 | @ANALYSIS_TIME.time() 289 | def analyze_file(self, file_path: Path): 290 | """ 291 | Analyzes a single file 292 | :param file_path: the file path 293 | """ 294 | self._progress_manager.set_postfix(self._truncate_middle(file_path)) 295 | 296 | try: 297 | self._persistence.add(str(file_path)) 298 | except Exception as e: 299 | logging.exception(e) 300 | echo(f"Error analyzing file '{file_path}': {e}") 301 | finally: 302 | self._progress_manager.inc() 303 | 304 | @FIND_DUPLICATES_TIME.time() 305 | def find_duplicates_of_file(self, root_directories: List[Path], root_directory: Path, reference_file_path: Path): 306 | """ 307 | Finds duplicates and marks all but the best copy as "to-be-deleted". 308 | :param root_directories: valid root directories 309 | :param root_directory: root directory of reference_file_path 310 | :param reference_file_path: the file to check for duplicates 311 | """ 312 | self._progress_manager.inc() 313 | self._progress_manager.set_postfix(self._truncate_middle(reference_file_path)) 314 | 315 | # remember processed files to prevent processing files in multiple directions 316 | if reference_file_path in self._processed_files: 317 | # already found a better candidate for this file 318 | return 319 | 320 | duplicate_candidates = self._persistence.find_similar(str(reference_file_path)) 321 | 322 | if self._config.SEARCH_ACROSS_ROOT_DIRS.value: 323 | # filter by files in at least one of the specified root directories 324 | # this is necessary because the database might hold items for other paths already 325 | # and those are not interesting to us 326 | duplicate_candidates = [ 327 | candidate for candidate in duplicate_candidates if 328 | any(root_dir in Path(candidate[MetadataKey.PATH.value]).parents for root_dir in root_directories) 329 | ] 330 | else: 331 | # filter by files in the same root directory 332 | duplicate_candidates = [ 333 | candidate for candidate in duplicate_candidates if 334 | root_directory in Path(candidate[MetadataKey.PATH.value]).parents 335 | ] 336 | 337 | if len(duplicate_candidates) <= 0: 338 | echo(f"No duplication candidates found in database for '{reference_file_path}'. " 339 | "This is an indication that the file has not been analysed yet or " 340 | "there was an issue analysing it.", 341 | color='yellow') 342 | 343 | if len(duplicate_candidates) <= 1: 344 | for candidate in duplicate_candidates: 345 | candidate_path = Path(candidate[MetadataKey.PATH.value]) 346 | 347 | if candidate_path != reference_file_path: 348 | echo(f"Unexpected unique duplication candidate '{candidate_path}' for " 349 | f"reference file '{reference_file_path}'", color='yellow') 350 | 351 | self._processed_files[candidate_path] = True 352 | 353 | # nothing to do here since the result is unique 354 | return 355 | 356 | # sort by quality criteria and redo the search to use the best candidate as the reference image 357 | sorted_duplicate_candidates = self._sort_by_quality_descending(duplicate_candidates) 358 | new_reference_file_path = sorted_duplicate_candidates[0][MetadataKey.PATH.value] 359 | duplicate_candidates = self._persistence.find_similar(new_reference_file_path) 360 | 361 | candidates_to_keep, candidates_to_delete = self._select_images_to_delete(duplicate_candidates) 362 | self._save_duplicates_for_result(candidates_to_keep, candidates_to_delete) 363 | 364 | def _save_duplicates_for_result(self, files_to_keep: List[dict], duplicates: List[dict]) -> None: 365 | """ 366 | Saves the comparison result for the final summary 367 | 368 | :param files_to_keep: list of image that shall be kept 369 | :param duplicates: less good duplicates 370 | """ 371 | self._deduplication_result.set_file_duplicates(files_to_keep, duplicates) 372 | 373 | for file_to_keep in files_to_keep: 374 | file_path = Path(file_to_keep[MetadataKey.PATH.value]) 375 | self._deduplication_result.add_file_action(file_path, ActionEnum.NONE) 376 | 377 | if self._config.DEDUPLICATOR_DUPLICATES_TARGET_DIRECTORY.value is None: 378 | action = ActionEnum.DELETE 379 | else: 380 | action = ActionEnum.MOVE 381 | for duplicate in duplicates: 382 | file_path = Path(duplicate[MetadataKey.PATH.value]) 383 | self._deduplication_result.add_file_action(file_path, action) 384 | 385 | def _select_images_to_delete(self, duplicate_candidates: [{}]) -> tuple: 386 | """ 387 | Selects which image to keep and which to remove 388 | :return: tuple (image to keep, list of images to remove) 389 | """ 390 | duplicate_candidates = self._sort_by_quality_descending(duplicate_candidates) 391 | 392 | # keep first and mark others for removal 393 | keep = [duplicate_candidates[0]] 394 | dont_keep = duplicate_candidates[1:] 395 | 396 | # move files that don't fit criteria to "keep" list 397 | max_mod_time_diff = self._config.MAX_FILE_MODIFICATION_TIME_DELTA.value 398 | if max_mod_time_diff is not None: 399 | # filter files that don't match max mod time diff criteria 400 | best_candidate = keep[0] 401 | best_match_mod_timestamp = best_candidate[MetadataKey.METADATA.value][ 402 | MetadataKey.FILE_MODIFICATION_DATE.value] 403 | 404 | for c in dont_keep: 405 | c_timestamp = c[MetadataKey.METADATA.value][MetadataKey.FILE_MODIFICATION_DATE.value] 406 | timestamp_diff = abs(c_timestamp - best_match_mod_timestamp) 407 | difference = datetime.timedelta(seconds=timestamp_diff) 408 | if difference > max_mod_time_diff: 409 | keep.append(c) 410 | dont_keep = list(filter(lambda x: x not in keep, dont_keep)) 411 | 412 | # remember that we have processed these files 413 | for candidate in duplicate_candidates: 414 | self._processed_files[candidate[MetadataKey.PATH.value]] = True 415 | 416 | return keep, dont_keep 417 | 418 | @staticmethod 419 | def _sort_by_quality_descending(duplicate_candidates) -> []: 420 | """ 421 | Sorts images according to the desired priorities. 422 | The first item in the list will be the most preferred one of all found duplicates. 423 | 424 | :param duplicate_candidates: the images to analyze 425 | :return: duplicate candidates sorted by given criteria 426 | """ 427 | 428 | def sort_criteria(candidate: dict) -> (): 429 | criteria = [] 430 | 431 | for rule in DeduplicatorConfig.PRIORITIZATION_RULES.value: 432 | rule_name = rule.get("name") 433 | if rule_name == "more-exif-data": 434 | if MetadataKey.EXIF_DATA.value in candidate[MetadataKey.METADATA.value]: 435 | # more exif data is better 436 | criteria.append(len(candidate[MetadataKey.METADATA.value][MetadataKey.EXIF_DATA.value]) * -1) 437 | elif rule_name == "less-exif-data": 438 | if MetadataKey.EXIF_DATA.value in candidate[MetadataKey.METADATA.value]: 439 | # more exif data is better 440 | criteria.append(len(candidate[MetadataKey.METADATA.value][MetadataKey.EXIF_DATA.value]) * 1) 441 | elif rule_name == "bigger-file-size": 442 | # reverse, bigger is better 443 | criteria.append(candidate[MetadataKey.METADATA.value][MetadataKey.FILE_SIZE.value] * -1) 444 | elif rule_name == "smaller-file-size": 445 | # smaller is better 446 | criteria.append(candidate[MetadataKey.METADATA.value][MetadataKey.FILE_SIZE.value] * 1) 447 | elif rule_name == "newer-file-modification-date": 448 | # reverse, bigger (later time) is better 449 | criteria.append( 450 | candidate[MetadataKey.METADATA.value][MetadataKey.FILE_MODIFICATION_DATE.value] * -1) 451 | elif rule_name == "older-file-modification-date": 452 | # smaller (earlier time) is better 453 | criteria.append( 454 | candidate[MetadataKey.METADATA.value][MetadataKey.FILE_MODIFICATION_DATE.value] * 1) 455 | elif rule_name == "smaller-distance": 456 | # smaller distance is better 457 | criteria.append(candidate[MetadataKey.DISTANCE.value]) 458 | elif rule_name == "bigger-distance": 459 | # bigger distance is better 460 | criteria.append(candidate[MetadataKey.DISTANCE.value] * -1) 461 | # elif rule_name == "longer-path": 462 | # elif rule_name == "shorter-path": 463 | elif rule_name == "contains-copy-in-file-name": 464 | # if the filename contains "copy" it is less good 465 | criteria.append("copy" in file.get_file_name(candidate[MetadataKey.PATH.value]).lower()) 466 | elif rule_name == "doesnt-contain-copy-in-file-name": 467 | # if the filename contains "copy" it is better 468 | criteria.append("copy" not in file.get_file_name(candidate[MetadataKey.PATH.value]).lower()) 469 | elif rule_name == "longer-file-name": 470 | # longer filename is better (for "edited" versions) 471 | criteria.append(len(file.get_file_name(candidate[MetadataKey.PATH.value])) * -1) 472 | 473 | elif rule_name == "shorter-file-name": 474 | # shorter filename is better (for "edited" versions) 475 | criteria.append(len(file.get_file_name(candidate[MetadataKey.PATH.value])) * 1) 476 | 477 | elif rule_name == "longer-folder-path": 478 | # shorter folder path is better 479 | criteria.append(len(file.get_containing_folder(candidate[MetadataKey.PATH.value])) * -1) 480 | elif rule_name == "shorter-folder-path": 481 | # shorter folder path is better 482 | criteria.append(len(file.get_containing_folder(candidate[MetadataKey.PATH.value]))) 483 | elif rule_name == "higher-score": 484 | # reverse, bigger is better 485 | criteria.append(candidate[MetadataKey.SCORE.value] * -1) 486 | elif rule_name == "lower-score": 487 | # lower is better 488 | criteria.append(candidate[MetadataKey.SCORE.value] * 1) 489 | elif rule_name == "higher-pixel-count": 490 | # higher pixel count is better 491 | criteria.append(candidate[MetadataKey.METADATA.value][MetadataKey.PIXELCOUNT.value] * -1) 492 | elif rule_name == "lower-pixel-count": 493 | # lower pixel count is better 494 | criteria.append(candidate[MetadataKey.METADATA.value][MetadataKey.PIXELCOUNT.value] * 1) 495 | 496 | # just to assure the order in the result is the same 497 | # if all other criteria (above) are equal 498 | # and recurring runs will result in the same order 499 | # (although they shouldn't be compared twice to begin with) 500 | criteria.append(candidate[MetadataKey.PATH.value]) 501 | 502 | return tuple(criteria) 503 | 504 | duplicate_candidates = sorted(duplicate_candidates, key=sort_criteria) 505 | 506 | return duplicate_candidates 507 | 508 | def process_duplicates(self): 509 | """ 510 | Moves or removes duplicates based on the configuration 511 | """ 512 | dry_run = self._config.DRY_RUN.value 513 | duplicate_target_directory = self._config.DEDUPLICATOR_DUPLICATES_TARGET_DIRECTORY.value 514 | if duplicate_target_directory: 515 | echo("Phase 5/6: Moving duplicates ...", color='cyan') 516 | self._move_files_marked_as_delete(duplicate_target_directory, dry_run) 517 | else: 518 | echo("Phase 5/6: Removing duplicates ...", color='cyan') 519 | self._remove_files_marked_as_delete(dry_run) 520 | 521 | def _find_empty_folders(self, root_path: Path, recursive: bool, dry_run: bool) -> [str]: 522 | """ 523 | Finds empty folders within the given root_path 524 | :param root_path: folder to search in 525 | """ 526 | result = OrderedSet() 527 | 528 | # traverse bottom-up to remove folders that are empty due to file removal 529 | for root, directories, files in os.walk(str(root_path), topdown=False): 530 | # get absolute paths of all files and folders in the current root directory 531 | abs_file_paths = list(map(lambda x: os.path.abspath(os.path.join(root, x)), files)) 532 | abs_folder_paths = list(map(lambda x: os.path.abspath(os.path.join(root, x)), directories)) 533 | 534 | # find out which of those files were deleted by the deduplication process 535 | files_deleted = list( 536 | map(lambda x: Path(x), filter( 537 | lambda x: Path(x) in self._deduplication_result.get_removed_or_moved_files(), 538 | abs_file_paths))) 539 | files_deleted = list(set(files_deleted + list( 540 | filter(lambda x: x.parent == Path(root), self._deduplication_result.get_removed_or_moved_files())))) 541 | 542 | folders_deleted = list(filter(lambda x: x in result, abs_folder_paths)) 543 | filtered_directories = list(filter(lambda x: x not in folders_deleted, abs_folder_paths)) 544 | 545 | if dry_run: 546 | if len(files_deleted) > 0 and len(files_deleted) == len(files) and len(folders_deleted) == len( 547 | directories): 548 | result.append(root) 549 | else: 550 | if len(files_deleted) > 0 and len(files) <= 0 and len(directories) <= 0: 551 | result.append(root) 552 | 553 | if not recursive: 554 | break 555 | 556 | return result 557 | 558 | def _remove_folders(self, root_path: Path, folders: [str], dry_run: bool): 559 | """ 560 | Function to remove empty folders 561 | :param root_path: 562 | """ 563 | echo(f"Removing empty folders ({len(folders)}) in: '{root_path}' ...") 564 | 565 | if len(folders) == 0: 566 | return 567 | 568 | self._progress_manager.start("Removing empty folders", len(folders), "Folder", self.interactive) 569 | for folder in folders: 570 | self._progress_manager.set_postfix(self._truncate_middle(folder)) 571 | 572 | if not dry_run: 573 | os.rmdir(folder) 574 | 575 | self._deduplication_result.add_removed_empty_folder(folder) 576 | self._progress_manager.inc() 577 | self._progress_manager.clear() 578 | 579 | def _remove_files_marked_as_delete(self, dry_run: bool): 580 | """ 581 | Removes files that were marked to be deleted in previous deduplication step 582 | :param dry_run: set to true to simulate this action 583 | """ 584 | items_to_remove = self._deduplication_result.get_file_with_action(ActionEnum.DELETE) 585 | marked_files_count = len(items_to_remove) 586 | if marked_files_count == 0: 587 | return 588 | 589 | self._progress_manager.start("Removing files", marked_files_count, "File", self.interactive) 590 | self._delete_files(items_to_remove, dry_run) 591 | self._progress_manager.clear() 592 | 593 | def _move_files_marked_as_delete(self, target_dir: Path, dry_run: bool): 594 | """ 595 | Moves files that were marked to be deleted in previous deduplication step to the target directory 596 | :param target_dir: the directory to move duplicates to 597 | :param dry_run: set to true to simulate this action 598 | """ 599 | items_to_move = self._deduplication_result.get_file_with_action(ActionEnum.MOVE) 600 | marked_files_count = len(items_to_move) 601 | if marked_files_count == 0: 602 | return 603 | 604 | self._progress_manager.start("Moving files", marked_files_count, "File", self.interactive) 605 | self._move_files(items_to_move, target_dir, dry_run) 606 | self._progress_manager.clear() 607 | 608 | def _delete_files(self, files_to_delete: [str], dry_run: bool): 609 | """ 610 | Deletes files on disk 611 | :param files_to_delete: list of absolute file paths 612 | :param dry_run: set to true to simulate this action 613 | """ 614 | for file_path in files_to_delete: 615 | self._progress_manager.set_postfix(self._truncate_middle(file_path)) 616 | 617 | if dry_run: 618 | pass 619 | else: 620 | # remove from file system 621 | if os.path.exists(file_path): 622 | os.remove(file_path) 623 | 624 | # remove from persistence 625 | self._persistence.remove(file_path) 626 | 627 | DUPLICATE_ACTION_DELETE_COUNT.inc() 628 | 629 | self._progress_manager.inc() 630 | 631 | def _move_files(self, files_to_move: List[Path], target_dir: Path, dry_run: bool): 632 | """ 633 | Moves files on disk 634 | :param files_to_move: list of absolute file paths 635 | :param target_dir: directory to move files to 636 | """ 637 | for file_path in files_to_move: 638 | self._progress_manager.set_postfix(self._truncate_middle(file_path)) 639 | 640 | try: 641 | if dry_run: 642 | continue 643 | 644 | # move file 645 | if not file_path.exists(): 646 | continue 647 | 648 | target_file = Path(str(target_dir), *file_path.parts[1:]) 649 | if target_file.exists(): 650 | if filecmp.cmp(file_path, target_file, shallow=False): 651 | os.remove(file_path) 652 | else: 653 | raise ValueError(f"Can't move duplicate file because the target already exists: {target_file}") 654 | else: 655 | target_file.parent.mkdir(parents=True, exist_ok=True) 656 | shutil.move(file_path, target_file) 657 | 658 | # remove from persistence 659 | self._persistence.remove(str(file_path)) 660 | 661 | DUPLICATE_ACTION_MOVE_COUNT.inc() 662 | except Exception as ex: 663 | logging.exception(ex) 664 | # LOGGER.log(ex) 665 | finally: 666 | self._progress_manager.inc() 667 | 668 | @staticmethod 669 | def _truncate_middle(text: any, max_length: int = 50): 670 | text = str(text) 671 | if len(text) <= max_length: 672 | # string is already short-enough, fill up with spaces 673 | return text + ((max_length - len(text)) * " ") 674 | # half of the size, minus the 3 .'s 675 | n_2 = int(max_length / 2) - 3 676 | # whatever's left 677 | n_1 = max_length - n_2 - 3 678 | return '{0}...{1}'.format(text[:n_1], text[-n_2:]) 679 | 680 | def remove_empty_folders(self): 681 | phase_6_text = "Phase 6/6: Removing empty folders" 682 | if not self._config.REMOVE_EMPTY_FOLDERS.value: 683 | echo(phase_6_text + " - Skipping", color='yellow') 684 | else: 685 | echo(phase_6_text, color='cyan') 686 | self._remove_empty_folders(self._config.SOURCE_DIRECTORIES.value, self._config.RECURSIVE.value) 687 | -------------------------------------------------------------------------------- /py_image_dedup/library/file_watch.py: -------------------------------------------------------------------------------- 1 | import re 2 | from pathlib import Path 3 | 4 | from watchdog.events import FileSystemEventHandler, EVENT_TYPE_MODIFIED, EVENT_TYPE_MOVED, EVENT_TYPE_CREATED, \ 5 | EVENT_TYPE_DELETED 6 | 7 | from py_image_dedup.config import DeduplicatorConfig 8 | from py_image_dedup.stats import FILE_EVENT_COUNT 9 | from py_image_dedup.util import echo 10 | 11 | 12 | class EventHandler(FileSystemEventHandler): 13 | 14 | def __init__(self, processing_manager): 15 | super().__init__() 16 | self.processing_manager = processing_manager 17 | 18 | self.config = DeduplicatorConfig() 19 | 20 | self.directory_regex = re.compile(rf"^({'|'.join(list(map(str, self.config.SOURCE_DIRECTORIES.value)))}).*$") 21 | self.file_regex = re.compile(rf"^.*({'|'.join(self.config.FILE_EXTENSION_FILTER.value)})$", re.IGNORECASE) 22 | 23 | def on_any_event(self, event): 24 | if not self._event_matches_filter(event): 25 | return 26 | 27 | FILE_EVENT_COUNT.labels(type=event.event_type).inc() 28 | 29 | echo("FileSystemEvent: {} {} {}".format(event.event_type, 30 | "directory" if event.is_directory else "file", 31 | event.src_path)) 32 | 33 | _actions = { 34 | EVENT_TYPE_CREATED: self.created, 35 | EVENT_TYPE_MODIFIED: self.modified, 36 | EVENT_TYPE_MOVED: self.moved, 37 | EVENT_TYPE_DELETED: self.deleted, 38 | } 39 | _actions[event.event_type](event) 40 | 41 | def created(self, event): 42 | self._process(event.src_path) 43 | 44 | def modified(self, event): 45 | self._process(event.src_path) 46 | 47 | def moved(self, event): 48 | self._cleanup(event.src_path) 49 | self._process(event.dest_path) 50 | 51 | def deleted(self, event): 52 | self._cleanup(event.src_path) 53 | 54 | def _process(self, path: str): 55 | self.processing_manager.add(Path(path)) 56 | 57 | def _cleanup(self, path: str): 58 | self.processing_manager.remove(Path(path)) 59 | 60 | def _event_matches_filter(self, event) -> bool: 61 | if event.is_directory: 62 | return False 63 | else: 64 | result = bool(self.directory_regex.match(event.src_path)) 65 | result &= bool(self.file_regex.match(event.src_path)) 66 | return result 67 | -------------------------------------------------------------------------------- /py_image_dedup/library/processing_manager.py: -------------------------------------------------------------------------------- 1 | from collections import OrderedDict 2 | from datetime import datetime, timedelta 3 | from pathlib import Path 4 | from threading import Lock 5 | from typing import List 6 | 7 | from watchdog.observers.inotify import InotifyObserver 8 | from watchdog.observers.polling import PollingObserver 9 | 10 | from py_image_dedup.config import DeduplicatorConfig, FILE_OBSERVER_TYPE_INOTIFY, FILE_OBSERVER_TYPE_POLLING 11 | from py_image_dedup.library import ActionEnum, RegularIntervalWorker 12 | from py_image_dedup.library.file_watch import EventHandler 13 | from py_image_dedup.library.progress_manager import ProgressManager 14 | from py_image_dedup.util.file import get_files_count 15 | 16 | 17 | class ProcessingManager(RegularIntervalWorker): 18 | lock = Lock() 19 | queue = OrderedDict() 20 | 21 | progress_manager: ProgressManager 22 | 23 | latest_event_time = None 24 | 25 | def __init__(self, deduplicator): 26 | self.config = DeduplicatorConfig() 27 | timeout = self.config.DAEMON_PROCESSING_TIMEOUT.value 28 | interval = timeout.total_seconds() 29 | super().__init__(interval) 30 | self.progress_manager = ProgressManager() 31 | self.deduplicator = deduplicator 32 | self.event_handler = EventHandler(self) 33 | self.observers = [] 34 | 35 | def start(self): 36 | observer_type = self.config.DAEMON_FILE_OBSERVER_TYPE.value 37 | directories = self.config.SOURCE_DIRECTORIES.value 38 | self.observers = self._setup_file_observers(observer_type, directories) 39 | super().start() 40 | 41 | def stop(self): 42 | for observer in self.observers: 43 | observer.stop() 44 | observer.join() 45 | 46 | self.observers.clear() 47 | 48 | def _setup_file_observers(self, observer_type: str, source_directories: List[Path]): 49 | observers = [] 50 | 51 | for directory in source_directories: 52 | if observer_type == FILE_OBSERVER_TYPE_INOTIFY: 53 | observer = InotifyObserver() 54 | elif observer_type == FILE_OBSERVER_TYPE_POLLING: 55 | observer = PollingObserver() 56 | else: 57 | raise ValueError(f"Unexpected file observer type {observer_type}") 58 | 59 | observer.schedule(self.event_handler, str(directory), recursive=True) 60 | observer.start() 61 | observers.append(observer) 62 | 63 | return observers 64 | 65 | def add(self, path: Path): 66 | with self.lock: 67 | self.latest_event_time = datetime.now() 68 | if path not in self.queue: 69 | self.queue[path] = path 70 | 71 | def remove(self, path: Path): 72 | if path in self.queue: 73 | self.queue.pop(path) 74 | self.deduplicator._persistence.remove(str(path)) 75 | 76 | def _should_process(self): 77 | return len(self.queue) > 0 and ( 78 | self.latest_event_time is None or 79 | (datetime.now() - timedelta(seconds=self._interval) > self.latest_event_time) 80 | ) 81 | 82 | def _run(self): 83 | with self.lock: 84 | self.process_queue() 85 | 86 | def process_queue(self): 87 | if not self._should_process(): 88 | return 89 | 90 | self.progress_manager.start("Processing", len(self.queue), "Files", False) 91 | while True: 92 | try: 93 | path, value = self.queue.popitem() 94 | self._process_queue_item(path, value) 95 | self.progress_manager.inc() 96 | except KeyError: 97 | break 98 | self.progress_manager.clear() 99 | 100 | def _process_queue_item(self, path, value): 101 | self.deduplicator.reset_result() 102 | 103 | # TODO: only a workaround until files can be processed too 104 | if path.is_file(): 105 | path = path.parent 106 | 107 | if path.is_dir(): 108 | files_count = get_files_count( 109 | path, 110 | self.config.RECURSIVE.value, 111 | self.config.FILE_EXTENSION_FILTER.value, 112 | self.config.EXCLUSIONS.value 113 | ) 114 | directory_map = { 115 | path: files_count 116 | } 117 | 118 | self.deduplicator.analyze_directories(directory_map) 119 | self.deduplicator.find_duplicates_in_directories(directory_map) 120 | 121 | # TODO: allow processing individual files 122 | # if path.is_file(): 123 | # self.deduplicator.analyze_file(path) 124 | # root_dir = Path(os.path.commonpath([path] + self.config.SOURCE_DIRECTORIES.value)) 125 | # self.deduplicator.find_duplicates_of_file(self.config.SOURCE_DIRECTORIES.value, root_dir, path) 126 | 127 | self.deduplicator.process_duplicates() 128 | 129 | # TODO: this needs rethinking 130 | # remove items that have been (re-)moved already from the event queue 131 | removed_items = self.deduplicator._deduplication_result.get_file_with_action(ActionEnum.DELETE) 132 | moved_items = self.deduplicator._deduplication_result.get_file_with_action(ActionEnum.MOVE) 133 | for item in set(removed_items + moved_items): 134 | if item in self.queue: 135 | self.queue.pop(item) 136 | self.progress_manager.inc() 137 | -------------------------------------------------------------------------------- /py_image_dedup/library/progress_manager.py: -------------------------------------------------------------------------------- 1 | import logging 2 | 3 | from tqdm import tqdm 4 | 5 | LOGGER = logging.getLogger(__name__) 6 | LOGGER.setLevel(logging.DEBUG) 7 | 8 | 9 | class ProgressManager: 10 | 11 | def __init__(self): 12 | self._progress_bar = None 13 | self._task = None 14 | self._n = None 15 | self._total = None 16 | self._unit = None 17 | self._last_percentage = None 18 | 19 | def start(self, task: str, total: int, unit: str, interactive: bool): 20 | if self._task is not None: 21 | LOGGER.warning(f"Starting new progress without explicitly closing the current one '{self._task}'") 22 | self.clear() 23 | 24 | self._task = task 25 | self._total = total 26 | self._unit = unit 27 | if interactive: 28 | self._progress_bar = self._create_progressbar(total, unit) 29 | 30 | def set_postfix(self, postfix: str): 31 | if self._progress_bar is not None and postfix is not None: 32 | self._progress_bar.set_postfix_str(postfix) 33 | 34 | def inc(self, n: int = 1): 35 | if self._task is None: 36 | raise AssertionError( 37 | "Can't increase before start. " 38 | "Please start a new task progress using start() before incrementing it.") 39 | 40 | if self._n is None: 41 | self._n = n 42 | else: 43 | self._n += n 44 | 45 | if self._progress_bar is not None: 46 | self._progress_bar.update(n) 47 | 48 | new_percentage = int((self._n / self._total) * 100) 49 | if self._last_percentage is None or self._last_percentage != new_percentage: 50 | self._last_percentage = new_percentage 51 | LOGGER.info(f"{self._task}: {new_percentage}% ({self._n}/{self._total})") 52 | 53 | def clear(self): 54 | if self._progress_bar is not None: 55 | self._progress_bar.close() 56 | self._progress_bar = None 57 | self._last_percentage = None 58 | self._n = None 59 | self._total = None 60 | self._task = None 61 | self._unit = None 62 | 63 | def _create_progressbar(self, total_count: int, unit: str) -> tqdm: 64 | """ 65 | Creates a new progress bar 66 | :param total_count: target for 100% 67 | :param unit: "Things" that are counted 68 | :return: progress bar 69 | """ 70 | self._progress_bar = tqdm(total=total_count, unit=unit, unit_scale=True, mininterval=1) 71 | return self._progress_bar 72 | -------------------------------------------------------------------------------- /py_image_dedup/persistence/__init__.py: -------------------------------------------------------------------------------- 1 | import logging 2 | import os 3 | 4 | from PIL import TiffImagePlugin 5 | 6 | from py_image_dedup.persistence.metadata_key import MetadataKey 7 | 8 | 9 | class ImageSignatureStore: 10 | """ 11 | Base class for Persistence implementations 12 | """ 13 | 14 | DATAMODEL_VERSION = 5 15 | 16 | def __init__(self, use_exif_data: bool = True): 17 | self._use_exif_data = use_exif_data 18 | 19 | def add(self, image_file_path: str): 20 | """ 21 | Analyze an image file and add it to the store 22 | 23 | :param image_file_path: path to the image file 24 | """ 25 | image_data = self._create_metadata_dict(image_file_path) 26 | 27 | # check if the file has already been analyzed (and didn't change in the meantime) 28 | existing_entity = self.get(image_file_path) 29 | if existing_entity is not None: 30 | is_data_version_ok = False 31 | try: 32 | if MetadataKey.DATAMODEL_VERSION.value in existing_entity[MetadataKey.METADATA.value]: 33 | is_data_version_ok = existing_entity[MetadataKey.METADATA.value][ 34 | MetadataKey.DATAMODEL_VERSION.value] == self.DATAMODEL_VERSION 35 | 36 | if is_data_version_ok and \ 37 | existing_entity[MetadataKey.METADATA.value][MetadataKey.FILE_SIZE.value] == image_data[ 38 | MetadataKey.FILE_SIZE.value] and \ 39 | existing_entity[MetadataKey.METADATA.value][ 40 | MetadataKey.FILE_MODIFICATION_DATE.value] == image_data[ 41 | MetadataKey.FILE_MODIFICATION_DATE.value]: 42 | # print("File is the same, not adding again") 43 | return 44 | except Exception as ex: 45 | logging.exception(ex) 46 | try: 47 | self.remove(image_file_path) 48 | except Exception as ex: 49 | logging.exception(ex) 50 | return 51 | 52 | self._add(image_file_path, image_data) 53 | 54 | def _create_metadata_dict(self, image_file_path: str) -> dict: 55 | """ 56 | Creates a dictionary that should be stored in persistence 57 | 58 | :param image_file_path: path to the image file 59 | :return: dictionary containing all relevant information 60 | """ 61 | from py_image_dedup.util import image 62 | 63 | image_data = {} 64 | image_data[MetadataKey.PATH.value] = image_file_path 65 | 66 | # get some metadata 67 | file_size = os.stat(image_file_path).st_size 68 | file_modification_date = os.path.getmtime(image_file_path) 69 | 70 | image_data[MetadataKey.DATAMODEL_VERSION.value] = self.DATAMODEL_VERSION 71 | image_data[MetadataKey.FILE_SIZE.value] = file_size 72 | image_data[MetadataKey.FILE_MODIFICATION_DATE.value] = file_modification_date 73 | 74 | image_data[MetadataKey.PIXELCOUNT.value] = image.get_pixel_count(image_file_path) 75 | 76 | if self._use_exif_data: 77 | exif_data = image.get_exif_data(image_file_path) 78 | exif_data = self._normalize_meta_data_for_db(exif_data) 79 | image_data[MetadataKey.EXIF_DATA.value] = exif_data 80 | 81 | return image_data 82 | 83 | def _normalize_meta_data_for_db(self, dictionary: dict) -> dict: 84 | """ 85 | :param dictionary: 86 | :return: 87 | """ 88 | result = {} 89 | for k, v in dictionary.items(): 90 | if isinstance(v, dict): 91 | result[k] = self._normalize_meta_data_for_db(v) 92 | continue 93 | 94 | normalized_value = v 95 | if isinstance(v, bytes) or isinstance(v, tuple): 96 | normalized_value = str(v) 97 | elif isinstance(v, TiffImagePlugin.IFDRational): 98 | if v._denominator != 0: 99 | normalized_value = v._numerator / v._denominator 100 | else: 101 | normalized_value = float(v._numerator) 102 | 103 | result[k] = normalized_value 104 | 105 | return result 106 | 107 | def _add(self, image_file_path: str, image_data: dict) -> None: 108 | """ 109 | Saves image data for the specified image file path 110 | 111 | :param image_file_path: image file path 112 | :param image_data: metadata for the image 113 | """ 114 | raise NotImplementedError() 115 | 116 | def get(self, image_file_path: str) -> dict or None: 117 | """ 118 | Get a store entry by it's file_path 119 | :param image_file_path: file path to search for 120 | :return: store entry or None 121 | """ 122 | raise NotImplementedError() 123 | 124 | def get_all(self) -> (int, object): 125 | """ 126 | :return: item count, stored entries as a generator function 127 | """ 128 | raise NotImplementedError() 129 | 130 | def find_similar(self, reference_image_file_path: str) -> []: 131 | """ 132 | Search for similar images to the specified one 133 | 134 | :param reference_image_file_path: the reference image file 135 | :return: list of images that are similar to the reference file 136 | """ 137 | raise NotImplementedError() 138 | 139 | def remove(self, image_file_path: str) -> None: 140 | """ 141 | Remove all entries with the given file path 142 | 143 | :param image_file_path: the path of an image file 144 | """ 145 | raise NotImplementedError() 146 | 147 | def remove_entries_of_missing_files(self): 148 | """ 149 | Remove all entries with files that don't exist 150 | """ 151 | entries = self.get_all() 152 | for entry in entries: 153 | file_path = entry['path'] 154 | if not os.path.exists(file_path): 155 | self.remove(file_path) 156 | 157 | def remove_all(self) -> None: 158 | """ 159 | Remove all entries from Database 160 | """ 161 | raise NotImplementedError() 162 | -------------------------------------------------------------------------------- /py_image_dedup/persistence/elasticsearchstorebackend.py: -------------------------------------------------------------------------------- 1 | import logging 2 | import time 3 | 4 | import requests 5 | from elasticsearch import Elasticsearch 6 | from image_match.elasticsearch_driver import SignatureES 7 | 8 | from py_image_dedup.persistence import ImageSignatureStore 9 | from py_image_dedup.util import echo 10 | 11 | 12 | class ElasticSearchStoreBackend(ImageSignatureStore): 13 | DEFAULT_EL_DOC_TYPE_EL_6 = 'image' 14 | DEFAULT_EL_DOC_TYPE_EL_7 = '_doc' 15 | 16 | def __init__(self, 17 | host: str, 18 | port: int, 19 | connections_per_node: int, 20 | el_index: str, 21 | el_version: int = None, 22 | el_doctype: str = None, 23 | max_dist: float = 0.03, 24 | use_exif_data: bool = True, 25 | setup_database: bool = True, 26 | ): 27 | """ 28 | Image signature persistence backed by image_match and elasticsearch 29 | 30 | :param host: host address of the elasticsearch server 31 | :param port: port of the elasticsearch server 32 | :param el_version: elasticsearch version 33 | :param el_index: elasticsearch index where the data is stored 34 | :param el_doctype: elasticsearch document type of the stored data 35 | :param max_dist: maximum "difference" allowed, ranging from [0 .. 1] where 0.2 is still a pretty similar image 36 | """ 37 | super().__init__(use_exif_data) 38 | 39 | self.host = host 40 | self.port = port 41 | self._connections_per_node = connections_per_node 42 | 43 | detected_version = None 44 | while detected_version is None: 45 | time.sleep(2) 46 | detected_version = self._detect_db_version() 47 | 48 | self._el_version = el_version 49 | if self._el_version is not None and detected_version is not None and self._el_version != detected_version: 50 | raise AssertionError( 51 | "Detected database version ({}) does not match expected version ({})".format(detected_version, 52 | self._el_version)) 53 | 54 | if detected_version is not None: 55 | self._el_version = detected_version 56 | elif self._el_version is None: 57 | # assume version 6 by default 58 | self._el_version = 6 59 | 60 | self._el_index = el_index 61 | if el_doctype is not None: 62 | self._el_doctype = el_doctype 63 | else: 64 | self._el_doctype = self.DEFAULT_EL_DOC_TYPE_EL_6 if self._el_version < 7 else self.DEFAULT_EL_DOC_TYPE_EL_7 65 | 66 | self.setup_database = setup_database 67 | if setup_database: 68 | try: 69 | # self._clear_database() 70 | self._setup_database() 71 | except Exception as e: 72 | logging.exception(e) 73 | raise AssertionError("Could not setup database") 74 | 75 | # noinspection PyTypeChecker 76 | self._store = SignatureES( 77 | es=Elasticsearch( 78 | hosts=[ 79 | {'host': self.host, 'port': self.port} 80 | ], 81 | maxsize=self._connections_per_node, 82 | ), 83 | # el_version=self._el_version, 84 | index=self._el_index, 85 | doc_type=self._el_doctype, 86 | distance_cutoff=max_dist, 87 | ) 88 | 89 | def _detect_db_version(self) -> int or None: 90 | try: 91 | response = requests.get('http://{}:{}'.format(self.host, self.port)) 92 | response.raise_for_status() 93 | return int(str(response.json()["version"]['number']).split(".")[0]) 94 | except Exception as ex: 95 | logging.exception(ex) 96 | return None 97 | 98 | def _setup_database(self): 99 | """ 100 | Creates the expected index, if it does not exist 101 | """ 102 | response = requests.get('http://{}:{}/{}'.format(self.host, self.port, self._el_index)) 103 | if response.status_code == 200: 104 | return 105 | elif response.status_code == 404: 106 | 107 | properties = { 108 | "properties": { 109 | "path": { 110 | "type": "keyword", 111 | "ignore_above": 256 112 | } 113 | } 114 | } 115 | 116 | if self._el_version == 7: 117 | json_data = { 118 | "mappings": properties 119 | } 120 | else: 121 | json_data = { 122 | "mappings": { 123 | self._el_doctype: properties 124 | } 125 | } 126 | 127 | response = requests.put( 128 | url='http://{}:{}/{}'.format(self.host, self.port, self._el_index), 129 | json=json_data 130 | ) 131 | 132 | response.raise_for_status() 133 | else: 134 | response.raise_for_status() 135 | 136 | def _clear_database(self): 137 | """ 138 | Removes the index and all data it contains 139 | """ 140 | requests.delete('http://{}:{}/{}'.format(self.host, self.port, self._el_index)) 141 | 142 | def _add(self, image_file_path: str, image_data: dict) -> None: 143 | # remove existing entries 144 | self.remove(image_file_path) 145 | self._store.add_image(image_file_path, metadata=image_data) 146 | 147 | def get(self, image_file_path: str) -> dict or None: 148 | """ 149 | Get a store entry by it's file_path 150 | :param image_file_path: file path to search for 151 | :return: 152 | """ 153 | db_entity = self._get(image_file_path) 154 | return db_entity 155 | 156 | def _get(self, image_file_path: str) -> dict or None: 157 | """ 158 | Get a store entry by it's file_path 159 | :param image_file_path: file path to search for 160 | :return: elasticsearch result dictionary 161 | """ 162 | es_query = { 163 | 'query': { 164 | "constant_score": { 165 | "filter": { 166 | "term": {'path': image_file_path} 167 | } 168 | } 169 | } 170 | } 171 | 172 | query_result = self._store.es.search(index=self._el_index, body=es_query) 173 | 174 | hits = query_result['hits']['hits'] 175 | 176 | if len(hits) > 1: 177 | echo(f"WARNING: More than a single entry for a file, cleaning up: {image_file_path}", color='yellow') 178 | self.remove(image_file_path) 179 | self.add(image_file_path) 180 | 181 | if len(hits) == 0: 182 | return None 183 | else: 184 | return hits[0]['_source'] 185 | 186 | def get_all(self) -> (int, object): 187 | es_query = { 188 | "track_total_hits": True, 189 | 'query': {'match_all': {}} 190 | } 191 | 192 | item_count = self._store.es.search(index=self._el_index, body=es_query, size=0)['hits']['total'] 193 | if self._el_version >= 7: 194 | item_count = item_count['value'] 195 | 196 | from elasticsearch.helpers import scan 197 | 198 | el6_params = { 199 | "doc_type": self._el_doctype 200 | } 201 | return item_count, scan( 202 | self._store.es, 203 | index=self._el_index, 204 | preserve_order=True, 205 | query=es_query, 206 | **(el6_params if self._el_version < 7 else {}) 207 | ) 208 | 209 | def find_similar(self, reference_image_file_path: str) -> []: 210 | try: 211 | entry = self._get(reference_image_file_path) 212 | if entry is not None: 213 | result = [] 214 | rec = self._store.search_single_record(entry) 215 | result.extend(rec) 216 | 217 | return result 218 | else: 219 | return self._store.search_image(reference_image_file_path, all_orientations=True) 220 | except Exception as e: 221 | echo(f"Error querying database for similar images of '{reference_image_file_path}': {e}", color="red") 222 | return [] 223 | 224 | def search_metadata(self, metadata: dict) -> []: 225 | """ 226 | Search for images with metadata properties. 227 | 228 | Note: Metadata will be empty if you did not provide it when adding an image 229 | :param metadata: 230 | :return: 231 | """ 232 | search_dict = {} 233 | for key, value in metadata.items(): 234 | search_dict[f"metadata.{key}"] = value 235 | 236 | es_query = { 237 | 'query': {'match': search_dict} 238 | } 239 | 240 | return self._store.es.search(index=self._el_index, body=es_query) 241 | 242 | def remove(self, image_file_path: str) -> None: 243 | # NOTE: this query will only work if the index has been created 244 | # with a custom mapping for the path property: 245 | 246 | # # remove existing index 247 | # curl -X DELETE "192.168.2.24:9200/images" 248 | # 249 | # # create index with custom mapping for "path" 250 | # curl -X PUT "192.168.2.24:9200/images?pretty" -H "Content-Type: application/json" -d 251 | # " 252 | # { 253 | # "mappings": { 254 | # "image": { 255 | # "properties": { 256 | # "path": { 257 | # "type": "keyword", 258 | # "ignore_above": 256 259 | # } 260 | # } 261 | # } 262 | # } 263 | # } 264 | # " 265 | 266 | es_query = { 267 | 'query': { 268 | "constant_score": { 269 | "filter": { 270 | "term": {'path': image_file_path} 271 | } 272 | } 273 | } 274 | } 275 | 276 | self._remove_by_query(es_query) 277 | 278 | def remove_all(self) -> None: 279 | es_query = { 280 | 'query': {'match_all': {}} 281 | } 282 | 283 | self._remove_by_query(es_query) 284 | 285 | def _remove_by_query(self, es_query: dict): 286 | el6_params = { 287 | "doc_type": self._el_doctype 288 | } 289 | 290 | return self._store.es.delete_by_query( 291 | index=self._el_index, 292 | body=es_query, 293 | conflicts="proceed", 294 | **(el6_params if self._el_version < 7 else {}) 295 | ) 296 | -------------------------------------------------------------------------------- /py_image_dedup/persistence/metadata_key.py: -------------------------------------------------------------------------------- 1 | from enum import Enum 2 | 3 | 4 | class MetadataKey(Enum): 5 | METADATA = "metadata" 6 | 7 | DATAMODEL_VERSION = "py-image-dedup_datamodel-version" 8 | 9 | PATH = "path" 10 | DISTANCE = "dist" 11 | SCORE = "score" 12 | 13 | FILE_SIZE = "filesize" 14 | FILE_MODIFICATION_DATE = "file_modification_date" 15 | 16 | PIXELCOUNT = "pixelcount" 17 | EXIF_DATA = "exif_data" 18 | -------------------------------------------------------------------------------- /py_image_dedup/stats.py: -------------------------------------------------------------------------------- 1 | from prometheus_client import Gauge, Summary 2 | 3 | DUPLICATE_ACTION_COUNT = Gauge( 4 | 'duplicate_action_total', 5 | 'Number of images per action', 6 | ['action'] 7 | ) 8 | DUPLICATE_ACTION_NONE_COUNT = DUPLICATE_ACTION_COUNT.labels(action="none") 9 | DUPLICATE_ACTION_MOVE_COUNT = DUPLICATE_ACTION_COUNT.labels(action="move") 10 | DUPLICATE_ACTION_DELETE_COUNT = DUPLICATE_ACTION_COUNT.labels(action="delete") 11 | 12 | FILE_EVENT_COUNT = Gauge( 13 | 'file_event_total', 14 | 'Number of file events per event type', 15 | ['type'] 16 | ) 17 | 18 | ANALYSIS_TIME = Summary('analyse_file_summary', 'Time spent analysing a file') 19 | 20 | FIND_DUPLICATES_TIME = Summary('find_duplicates_summary', 'Time spent finding duplicates of a file') 21 | -------------------------------------------------------------------------------- /py_image_dedup/util/__init__.py: -------------------------------------------------------------------------------- 1 | import functools 2 | import logging 3 | import traceback 4 | 5 | import click 6 | 7 | LOGGER = logging.getLogger(__name__) 8 | LOGGER.setLevel(logging.DEBUG) 9 | 10 | 11 | def echo(text: str = "", color=None): 12 | """ 13 | Prints a text to the console 14 | :param text: the text 15 | :param color: an optional color 16 | """ 17 | if text is not click.termui and text is not str: 18 | text = str(text) 19 | if color: 20 | text = click.style(text, fg=color) 21 | if len(text) > 0: 22 | LOGGER.debug(text) 23 | click.echo(text) 24 | 25 | 26 | def reraise_with_stack(func): 27 | """ 28 | Decorator used to reraise exceptions occurring within a future. 29 | 30 | :param func: function to decorate 31 | :return: decorated function 32 | """ 33 | 34 | @functools.wraps(func) 35 | def wrapped(*args, **kwargs): 36 | try: 37 | return func(*args, **kwargs) 38 | except Exception as e: 39 | traceback_str = traceback.format_exc() 40 | raise ValueError("Error occurred. Original traceback is\n%s\n" % traceback_str) 41 | 42 | return wrapped 43 | -------------------------------------------------------------------------------- /py_image_dedup/util/file.py: -------------------------------------------------------------------------------- 1 | import os 2 | from pathlib import Path 3 | from typing import List 4 | 5 | 6 | def get_file_name(file_path: str) -> str: 7 | folder, file = os.path.split(file_path) 8 | return file 9 | 10 | 11 | def get_containing_folder(file_path: str) -> str: 12 | folder, file = os.path.split(file_path) 13 | return folder 14 | 15 | 16 | def file_has_extension(file: Path, extensions: List[str] or None) -> bool: 17 | """ 18 | Checks if a file has one of the given extensions 19 | :param file: the file to check 20 | :param extensions: allowed extensions 21 | :return: true if it matches (case insensitive), false otherwise 22 | """ 23 | if not isinstance(extensions, List): 24 | extensions = [extensions] 25 | if not extensions: 26 | return True 27 | 28 | if file.suffix.lower() not in (ext.lower() for ext in extensions): 29 | # skip file with unwanted file extension 30 | return False 31 | else: 32 | return True 33 | 34 | 35 | def get_files_count(directory: Path, recursive: bool, file_extensions: List[str] or None, exclusions: List) -> int: 36 | """ 37 | :param directory: the directory to analyze 38 | :param recursive: whether to search the directory recursively 39 | :param file_extensions: file extensions to include 40 | :return: number of files in the given directory that match the currently set file filter 41 | """ 42 | files_count = 0 43 | for r, d, files in os.walk(str(directory)): 44 | for file in files: 45 | file = Path(file) 46 | if any(list(map(lambda x: x.search(str(file.absolute())), exclusions))): 47 | continue 48 | if not file_has_extension(file, file_extensions): 49 | continue 50 | files_count += 1 51 | if not recursive: 52 | break 53 | 54 | return files_count 55 | -------------------------------------------------------------------------------- /py_image_dedup/util/image.py: -------------------------------------------------------------------------------- 1 | import PIL.ExifTags 2 | from PIL import Image 3 | 4 | 5 | def get_exif_data(image_file_path: str) -> {}: 6 | """ 7 | Tries to extract all exif data from the given image file 8 | :param image_file_path: path of the image file 9 | :return: dictionary containing all available exif data entries and their values 10 | """ 11 | 12 | result = {} 13 | try: 14 | img = Image.open(image_file_path) 15 | 16 | exif_data = img._getexif() 17 | if not exif_data: 18 | return result 19 | 20 | for k, v in exif_data.items(): 21 | if k in PIL.ExifTags.TAGS: 22 | tag_name = PIL.ExifTags.TAGS[k] 23 | result[tag_name] = v 24 | except Exception as e: 25 | pass 26 | return result 27 | 28 | 29 | def get_pixel_count(image_file_path: str) -> int: 30 | try: 31 | img = Image.open(image_file_path) 32 | width, height = img.size 33 | return width * height 34 | except Exception as e: 35 | pass 36 | return 0 37 | -------------------------------------------------------------------------------- /py_image_dedup_reference.yaml: -------------------------------------------------------------------------------- 1 | # This is a reference configuration file explaining all the options 2 | # of py-image-dedup. 3 | 4 | py_image_dedup: 5 | # Configuration for the analysis phase, see README.md 6 | analysis: 7 | # Whether to search for duplicates across directories when 8 | # specifying more than one image source directory 9 | across_dirs: false 10 | # A filter for the file extensions to analyse 11 | file_extensions: 12 | - .png 13 | - .jpg 14 | - .jpeg 15 | # Whether to search recursively in each of the source directories 16 | recursive: true 17 | # A list of source directories to analyse 18 | source_directories: 19 | - /home/myuser/pictures/ 20 | # A list of regex patterns to ignore when traversing any of the 21 | # source directories 22 | exclusions: 23 | - ".*/excluded/.*" 24 | # The number of threads to use for image analysis. 25 | # If unset, this will default to `os.cpu_count()`. 26 | threads: 1 27 | # Whether to include EXIF data of images in the analysis 28 | use_exif_data: true 29 | 30 | # Deduplication phase specific configuration options, see README.md 31 | deduplication: 32 | # The target directory to move duplicate images to 33 | duplicates_target_directory: /home/myuser/pictures/duplicates/ 34 | # Upper limit on the modification date difference between 35 | # two duplicate images to be considered the same image. 36 | max_file_modification_time_diff: 0:05:00 37 | # Specifies the criteria and their order for ordering the list of duplicates 38 | # to select the best copy. 39 | prioritization_rules: 40 | - name: "more-exif-data" 41 | - name: "less-exif-data" 42 | - name: "bigger-file-size" 43 | - name: "smaller-file-size" 44 | - name: "newer-file-modification-date" 45 | - name: "older-file-modification-date" 46 | - name: "smaller-distance" 47 | - name: "bigger-distance" 48 | - name: "longer-path" 49 | - name: "shorter-path" 50 | - name: "contains-copy-in-file-name" 51 | - name: "longer-file-name" 52 | - name: "shorter-file-name" 53 | - name: "longer-folder-path" 54 | - name: "shorter-folder-path" 55 | - name: "higher-score" 56 | - name: "lower-score" 57 | 58 | # Daemon specific configuration options, see README.md 59 | daemon: 60 | # Time for waiting on filesystems changes to settle before analysing. 61 | timeout: 30s 62 | # The type of file observer to use. 63 | # One of: polling, inotify 64 | file_observer: polling 65 | # A dry run can be used to validate the log output of a specific configuration 66 | # before actually deleting or removing any images in any of the source 67 | # directories. 68 | dry_run: true 69 | 70 | # Elasticsearch specific configuration options, see README.md 71 | elasticsearch: 72 | # Whether to automatically create an index in the target database. 73 | auto_create_index: true 74 | # Hostname of the elasticsearch backend instance to use 75 | host: 127.0.0.1 76 | # Port of the elasticsearch backend instance to use. 77 | port: 9200 78 | # The index name to use for storing and querying image analysis data. 79 | index: images 80 | # Maximum signature distance [0..1] to query from elasticsearch backend. 81 | max_distance: 0.1 82 | # Whether to remove empty folders or not. 83 | remove_empty_folders: false 84 | 85 | # Prometheus exporter specific configuration options, see README.md 86 | stats: 87 | # Whether to enable prometheus statistics or not. 88 | enabled: true 89 | # The port to expose statistics on. 90 | port: 8000 91 | -------------------------------------------------------------------------------- /pyproject.toml: -------------------------------------------------------------------------------- 1 | [tool.poetry] 2 | name = "py-image-dedup" 3 | version = "2.0.1" 4 | description = "A library to find duplicate images and delete unwanted ones" 5 | 6 | license = "AGPLv3+" 7 | 8 | authors = [ 9 | "Markus Ressel ", 10 | ] 11 | 12 | readme = 'README.md' 13 | 14 | repository = "https://github.com/markusressel/py-image-dedup" 15 | homepage = "https://github.com/markusressel/py-image-dedup" 16 | 17 | keywords = ['deduplication', 'py-image-dedup'] 18 | 19 | classifiers = [ 20 | "License :: OSI Approved :: GNU Affero General Public License v3 or later (AGPLv3+)", 21 | "Programming Language :: Python :: 3 :: Only", 22 | "Programming Language :: Python :: 3", 23 | "Programming Language :: Python :: 3.8", 24 | "Programming Language :: Python :: 3.9", 25 | "Development Status :: 5 - Production/Stable" 26 | ] 27 | 28 | [build-system] 29 | requires = ["poetry-core>=1.0.0"] 30 | build-backend = "poetry.core.masonry.api" 31 | 32 | [tool.poetry.dependencies] 33 | python = "^3.11" # Compatible python versions must be declared here 34 | 35 | Pillow = "*" 36 | ordered-set = "*" 37 | watchdog = ">=0.10.2,<6.1.0" 38 | elasticsearch = "^7" 39 | scipy = "*" 40 | numpy = "*" 41 | container-app-conf = "^5.0.0" 42 | requests = "^2.20.0" 43 | click = "*" 44 | tabulate = ">=0.8.3,<0.10.0" 45 | tqdm = "*" 46 | prometheus-client = "*" 47 | image-match = { git = "https://github.com/markusressel/image-match.git", tag = "3.0.0" } 48 | 49 | [tool.poetry.group.dev.dependencies] 50 | pytest = "*" 51 | 52 | [tool.poetry.scripts] 53 | py-image-dedup = 'py_image_dedup.cli:cli' 54 | -------------------------------------------------------------------------------- /pytest.ini: -------------------------------------------------------------------------------- 1 | [pytest] 2 | testpaths = tests 3 | -------------------------------------------------------------------------------- /tests/__init__.py: -------------------------------------------------------------------------------- 1 | import unittest 2 | 3 | from py_image_dedup.config import DeduplicatorConfig 4 | 5 | 6 | class TestBase(unittest.TestCase): 7 | 8 | def setUp(self): 9 | self.config = DeduplicatorConfig() 10 | from py_image_dedup.library.deduplicator import ImageMatchDeduplicator 11 | self.under_test = ImageMatchDeduplicator(interactive=False) 12 | 13 | def tearDown(self): 14 | pass 15 | 16 | if __name__ == '__main__': 17 | unittest.main() 18 | -------------------------------------------------------------------------------- /tests/images/bottles/IMG_20190903_193151-edited.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/bottles/IMG_20190903_193151-edited.jpg -------------------------------------------------------------------------------- /tests/images/bottles/IMG_20190903_193151-grayscale.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/bottles/IMG_20190903_193151-grayscale.jpg -------------------------------------------------------------------------------- /tests/images/bottles/IMG_20190903_193151-telegram-compression.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/bottles/IMG_20190903_193151-telegram-compression.jpg -------------------------------------------------------------------------------- /tests/images/bottles/IMG_20190903_193151.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/bottles/IMG_20190903_193151.jpg -------------------------------------------------------------------------------- /tests/images/building/IMG_20190903_193508-edited.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/building/IMG_20190903_193508-edited.jpg -------------------------------------------------------------------------------- /tests/images/building/IMG_20190903_193508-grayscale.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/building/IMG_20190903_193508-grayscale.jpg -------------------------------------------------------------------------------- /tests/images/building/IMG_20190903_193508-telegram-compression.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/building/IMG_20190903_193508-telegram-compression.jpg -------------------------------------------------------------------------------- /tests/images/building/IMG_20190903_193508.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/building/IMG_20190903_193508.jpg -------------------------------------------------------------------------------- /tests/images/clouds/IMG_20190903_193537-edited.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/clouds/IMG_20190903_193537-edited.jpg -------------------------------------------------------------------------------- /tests/images/clouds/IMG_20190903_193537-grayscale.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/clouds/IMG_20190903_193537-grayscale.jpg -------------------------------------------------------------------------------- /tests/images/clouds/IMG_20190903_193537-telegram-compression.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/clouds/IMG_20190903_193537-telegram-compression.jpg -------------------------------------------------------------------------------- /tests/images/clouds/IMG_20190903_193537.jpg: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/markusressel/py-image-dedup/c24afc64eb339ae339f15ed03b9ddd8e83a799ce/tests/images/clouds/IMG_20190903_193537.jpg -------------------------------------------------------------------------------- /tests/py_image_dedup.yaml: -------------------------------------------------------------------------------- 1 | py_image_dedup: 2 | dry_run: true 3 | analysis: 4 | across_dirs: true 5 | file_extensions: 6 | - .png 7 | - .jpg 8 | - .jpeg 9 | recursive: true 10 | source_directories: 11 | - ./images/ 12 | threads: 8 13 | use_exif_data: true 14 | deduplication: 15 | # duplicates_target_directory: 16 | max_file_modification_time_diff: 0:01:40 17 | prioritization_rules: 18 | - name: "more-exif-data" 19 | - name: "less-exif-data" 20 | - name: "bigger-file-size" 21 | - name: "smaller-file-size" 22 | - name: "newer-file-modification-date" 23 | - name: "older-file-modification-date" 24 | - name: "smaller-distance" 25 | - name: "bigger-distance" 26 | - name: "longer-path" 27 | - name: "shorter-path" 28 | - name: "contains-copy-in-file-name" 29 | - name: "longer-file-name" 30 | - name: "shorter-file-name" 31 | - name: "longer-folder-path" 32 | - name: "shorter-folder-path" 33 | - name: "higher-score" 34 | - name: "lower-score" 35 | 36 | elasticsearch: 37 | auto_create_index: true 38 | host: 127.0.0.1 39 | max_distance: 0.1 40 | remove_empty_folders: false 41 | 42 | -------------------------------------------------------------------------------- /tests/test_file_extension.py: -------------------------------------------------------------------------------- 1 | from pathlib import Path 2 | 3 | from py_image_dedup.util.file import file_has_extension 4 | from tests import TestBase 5 | 6 | 7 | class FileExtensionTest(TestBase): 8 | 9 | def test_png(self): 10 | paths = [ 11 | "file.png", 12 | "file.PNG" 13 | ] 14 | for path in paths: 15 | path = Path(path) 16 | self.assertTrue(file_has_extension(path, [".png"])) 17 | -------------------------------------------------------------------------------- /tests/test_select_images_to_delete.py: -------------------------------------------------------------------------------- 1 | import random 2 | import unittest 3 | from random import shuffle 4 | from random import uniform 5 | 6 | from py_image_dedup.persistence.metadata_key import MetadataKey 7 | from tests import TestBase 8 | 9 | 10 | class SelectImagesToDeleteTest(TestBase): 11 | 12 | def test_select_images_to_delete__filter_max_mod_time_diff(self): 13 | keep = [ 14 | self._create_default_candidate(modification_date=1500), 15 | # file modification time is too far apart from the optimal candidate for these 16 | self._create_default_candidate(modification_date=1001), 17 | ] 18 | 19 | max_delta_seconds = int(self.config.MAX_FILE_MODIFICATION_TIME_DELTA.value.total_seconds()) 20 | for i in range(50): 21 | c = self._create_default_candidate(modification_date=random.choice(range(0, 1500 - max_delta_seconds))) 22 | keep.append(c) 23 | 24 | dont_keep = [ 25 | self._create_default_candidate(modification_date=1450) 26 | ] 27 | 28 | self._run_test(keep, dont_keep) 29 | 30 | def test_select_images_to_delete__contains_copy(self): 31 | keep = [self._create_default_candidate(path="C:/1.jpg")] 32 | 33 | dont_keep = [] 34 | for i in range(50): 35 | c = self._create_default_candidate(path=f"C:/1{i}-Copy.jpg") 36 | dont_keep.append(c) 37 | 38 | self._run_test(keep, dont_keep) 39 | 40 | def test_select_images_to_delete__newer_and_bigger(self): 41 | keep = [self._create_default_candidate(filesize=100, modification_date=100)] 42 | 43 | dont_keep = [] 44 | for i in range(50): 45 | c = self._create_default_candidate(filesize=i, modification_date=i) 46 | dont_keep.append(c) 47 | 48 | self._run_test(keep, dont_keep) 49 | 50 | def test_select_images_to_delete__newer(self): 51 | keep = [self._create_default_candidate(modification_date=100)] 52 | 53 | dont_keep = [] 54 | for i in range(50): 55 | c = self._create_default_candidate(modification_date=i) 56 | dont_keep.append(c) 57 | 58 | self._run_test(keep, dont_keep) 59 | 60 | def test_select_images_to_delete__bigger(self): 61 | keep = [self._create_default_candidate(filesize=100)] 62 | 63 | dont_keep = [] 64 | for i in range(50): 65 | c = self._create_default_candidate(filesize=i) 66 | dont_keep.append(c) 67 | 68 | self._run_test(keep, dont_keep) 69 | 70 | def test_select_images_to_delete__all_the_same(self): 71 | keep = [self._create_default_candidate(path="C:/00000.jpg")] 72 | 73 | dont_keep = [] 74 | for i in range(50): 75 | c = self._create_default_candidate(path=f"C:/1{i}.jpg") 76 | dont_keep.append(c) 77 | 78 | self._run_test(keep, dont_keep) 79 | 80 | def test_select_images_to_delete__all_the_same_2(self): 81 | keep = [self._create_default_candidate(path="C:/50-edited.jpg")] 82 | 83 | dont_keep = [] 84 | for i in range(50): 85 | c = self._create_default_candidate(path=f"C:/{i}.jpg") 86 | dont_keep.append(c) 87 | 88 | self._run_test(keep, dont_keep) 89 | 90 | def test_select_images_to_delete__higher_score(self): 91 | keep = [self._create_default_candidate(score=100)] 92 | 93 | dont_keep = [] 94 | for i in range(50): 95 | c = self._create_default_candidate() 96 | dont_keep.append(c) 97 | 98 | self._run_test(keep, dont_keep) 99 | 100 | def test_select_images_to_delete__lower_dist(self): 101 | keep = [self._create_default_candidate(dist=0)] 102 | 103 | dont_keep = [] 104 | for i in range(50): 105 | c = self._create_default_candidate(dist=uniform(0.1, 1.0)) 106 | dont_keep.append(c) 107 | 108 | self._run_test(keep, dont_keep) 109 | 110 | def test_select_images_to_delete__real_example(self): 111 | keep = [self._create_default_candidate( 112 | path=r"M:\Fotos\Markus\Google Photos Archiv\Takeout\Google Photos\2017-06-17\20170617_153437.jpg", 113 | filesize=10000000, modification_date=1)] 114 | 115 | dont_keep = [] 116 | for i in range(50): 117 | c = self._create_default_candidate( 118 | path=r"M:\Fotos\Iris\Syncthing\Telegram Empfangen\223023133_644761%i.jpg" % i, 119 | filesize=270000, modification_date=2) 120 | dont_keep.append(c) 121 | 122 | self._run_test(keep, dont_keep) 123 | 124 | def _run_test( 125 | self, keep: [{}], dont_keep: [{}], test_reversed_order: bool = True, 126 | test_random_input_order: bool = True 127 | ): 128 | candidates = keep + dont_keep 129 | 130 | kept, not_kept = self.under_test._select_images_to_delete(candidates) 131 | self._test_result_outcome(kept, not_kept, keep, dont_keep) 132 | 133 | if test_reversed_order: 134 | kept, not_kept = self.under_test._select_images_to_delete(reversed(candidates)) 135 | self._test_result_outcome(kept, not_kept, keep, dont_keep) 136 | 137 | if test_random_input_order: 138 | # test random sort orders of input just to be sure 139 | for i in range(50): 140 | shuffle(candidates) 141 | kept, not_kept = self.under_test._select_images_to_delete(candidates) 142 | self._test_result_outcome(kept, not_kept, keep, dont_keep) 143 | 144 | def _test_result_outcome(self, kept, not_kept, keep: [{}], dont_keep: [{}]): 145 | for c in keep: 146 | self.assertIn(c, kept) 147 | for c in dont_keep: 148 | self.assertIn(c, not_kept) 149 | 150 | @staticmethod 151 | def _create_default_candidate( 152 | path: str = "C:/test", dist: float = 0.05, filesize: int = 100, 153 | modification_date: int = 1, pixel_count: int = 10000, exif_tags: {} = {}, 154 | score: int = 64 155 | ) -> {}: 156 | return { 157 | MetadataKey.PATH.value: path, 158 | MetadataKey.DISTANCE.value: dist, 159 | MetadataKey.METADATA.value: { 160 | MetadataKey.FILE_SIZE.value: filesize, 161 | MetadataKey.FILE_MODIFICATION_DATE.value: modification_date, 162 | MetadataKey.PIXELCOUNT.value: pixel_count, 163 | MetadataKey.EXIF_DATA.value: exif_tags 164 | }, 165 | MetadataKey.SCORE.value: score 166 | } 167 | 168 | 169 | if __name__ == '__main__': 170 | unittest.main() 171 | --------------------------------------------------------------------------------